MG-Cars.net

Welcome to our resource for MG Car Information.

Recommendations

Parts

MG parts spares and accessories are available for MG T Series (TA, MG TB, MG TC, MG TD, MG TF), Magnette, MGA, Twin cam, MGB, MGBGT, MGC, MGC GT, MG Midget, Sprite and other MG models from British car spares company LBCarCo.

MG Midget and Sprite Technical - Diff rebuild 2023. Can this be saved?

I need a diff rebuild. I'll farm the job out to
Hardy engineering in Reigate. But before I do, I'd like to garner a bit more knowledge.

I've got two 3.9 diffs, and one 4.2.

The 4.2 is pretty good and quiet, but I don't use it.

One 3.9 is pretty good if you're deaf, reasonably smooth, but howls a little like a banshee having a bad temper after a night on the beer. Esp' on the over-run.

The other 3.9 is pretty quiet, but rumbles, and makes the rear end feel as if it has a bent halfshaft. Rrrrr - rrrrr- rrrr -, which increases frequency as speed increases.

So I stripped that diff down today and think I've found the cause.

I took the carrier out of the diff casing. I went to pick up the CW and P by holding the bearing, and the bearing just pulled off the carrier on one side. The other is still fixed TIGHT.

Here are some pictures.

If tempted to answer, give me a chance to post the full set of pics before answering.









anamnesis

More pictures. Hard to see, but I can feel the grooves with me finger nail in places.

And see the state of the shims that came from behind the bearing.






anamnesis

The odd thing is, the inside facei of the inner race of the bearing isn't marked.

Which makes me think it was like this when it was assembled. It was SUPPOSED to be a BL refurbished diff. I bought it from MOSS many years ago on special offer.

It sat unused for a couple of years. When I pit it in, it lasted about 2 months, when it seized solid about a hundred yards from home one day. Then it 'freed up' just enough for me to get it home. I took it apart and found a chunk of metal down by the pinion bearings. Having removed the metal, the diff seemed ok. but Rrrrr -- rrrr- as I described earlier.

The question is, is the crown wheel and pinion salvageable? I've read here in the past about swapping the CW, to a different carrier, but the carriers aren't all compatible. Would my 4.2 carrier be compatible?

Any ideas?




anamnesis

It looks like the inner bearing race was slipping on the carrier. IIRC that bearing should be a press fit. I would take it in to the people who do this every day and see what they say. It will probably be cheaper in the long run since they have all of the proper tools and knowledge to fix these things. I did that with mine and all I had to do was pay a bunch of money and re-install it. No more howling.
Martin

Anam

When you to take the diff to Hardy talk to Bill McDonough the owner. He is very helpful (owns two Frogeyes!) and will advise what diff is the most appropriate to restore. I had two 3.7 diffs one looked fine and was reasonably quiet whereas the other was even loud and clunked on the overrun. Ironically it was that diff that was restored as the cw and pinion were in good shape. The restoration included all new bearings including the sun/planet thrusts. £250 but that was 4 years ago. Ring Bill before you go as he can be out.

Bob Beaumont

Yes, you should be able to swap the CW & P to a different differential/housing. It just needs a bit of care to shim things up correctly.

I did it when I built up my LSD. I had a 3.9 (obliterated pinion gear) and a 4.2 (noisy a.f). I can't remember why I did it the way I did... but I build the new 3.55 CW & P and LSD into the 4.2 pumpkin. Then the best of the various remaining parts, the diff from the 3.9 and the 4.2 CW&P got built into the 3.9 pumpkin.

Best of luck.
Malc.
Malcolm

Thanks Bob. That's just about the plan. I've been meaning to do this for some time. I think it was Alan who told me about Hardy's a while ago, not sure if I knew the name of Bill though, or if I've just forgotten it, so thanks for that.

Only posting here to gain some better prior understanding, so I can talk a little more intelligently to Hardy when I call/visit them.

Hi Malcolm. Sorry, I may be/probably I am, mixing my terms.

By carrier, I meant the bit the CW bolts to. But I think now that's actually the 'cage'.

It's my CAGE journal that's damaged/scored. And what I'm trying to establish is, can a CW be unbolted from one cage, and bolted onto another.

As I have an unused 4.2 diff, with a good undamaged cage, it'd be useful if I could use that cage.

Are the 4.2 diff CAGES, the same as the 3.9 diff cages?

anamnesis

Yeah, the "cages" should be the same. What I referred to as the "diff" is the cage + internal sun and planet gears as a unit.

Potato, tomato! ;-)

Malc.
Malcolm

Terminology clarification.

Yes, you should be able to mix and match what you have :-)



Malcolm

Excellent, thanks Malcolm.

Bottom line, money. If I can use my spare 4.2 cage, I don't have to pay for a new one. 🙂.
anamnesis

A friend had a diff that whined on overrun, gradually getting worse but then it all went quiet, followed shortly afterwards by the rear axle locking solid.

The pinion bearing had been wearing and allowing the input shaft to move, eventually letting go such that the pinion went backwards and jammed in the cage.

Not a midget though, the diff is one of the few bits of my car that I haven't had apart, but then it doesn't have the mileage on it that your does.
AdrianR

Anam,

New I had this somewhere but took a bit of finding - its from the Special Tuning pages - from the early edition but its also in the 1275 edition,
HTH

R.
richard b

Hi Richard. Thanks for that. Funnily enough I searched the archives and found your reference to that a little earlier.

I'll have a close look at my spare 4.2 cage, and see if it matches the numbers.

If it does, the bearings on my spare cage are good, but I assume they'd have to be removed to set the shims for it to go back into my 3.9 diff case. Which is beyond my capabilities, but hopefully can be done at Hardy's.

anamnesis

If you fit the bearings up to the crownwheel carrier without shims, then it's a lot easier to fit shims to the outside of the bearing instead of the inside---they're 2 3/4" od
Minimania carry them in .002"steps

Unusual that your carrier is loose on that side, it's usually on the shorter/loaded side

William Revit

Yep - I did the outside shimming thing. So much easier.

Depending on how bored I get this winter, I am considering pulling the diff to see how it is settling in after 12 months abuse.

Best of luck with yours Anam,
Malc.
Malcolm

And for amusement on the topic of diffs... this was my pinion gear. I found two teeth in the "sump" and another was cracked so far through that it broke off with a light tap with a hammer.

It was still working, and I don't ever drive it gently! It was getting a bit clunky mind... ha ha ha! :-D

(Note, the cuts in the shaft were me destructively removing the bearing!)

Malc.


Malcolm

Looks a familiar sight Malc. we towed to a race meeting about 1000Klm away once and had mistakingly left the car in gear, It was tied down on the trailer but just the rocking was enough to pluck a tooth off the crownwheel-----quick dismantle, pulled one side off the housing and removed the tooth , back together with one missing, just a bit clunky moving off but otherwise went ok---spent all weekend with the fingers crossed though, waiting for the bang, normally would have packed up and gone home but it was a big dollar race and the last one for the year anyway
William Revit

Hi Willy, I can only guess that the rouge bit of metal that was loose in there, and caused the diff to seize momentarily and then free up, must have locked that bearing I've taken off (fell off with just my finger pressure), causing it's inner race to rotate on the cage/carrier journal, as per pictures.

Maybe the bearing the other side is worn too. But either way, the cage is knackered.

Pity our diffs aren't adjusted by threaded adjusters, like my Capri English diff is.

Very nice Malc. Don't do things by half do you. 😂. Amazing how things can keep going even though they look well knackered.

My bearing was loose enough to just pull off with my fingers, but hasn't caused THAT much of a problem in use.

Hmmm, this suggestion of shimming the outsides is tempting me to have a go at fixing this myself.

But if I can get my head round the numbers, I still don't think I have to tools to make accurate enough measurements. And I don't have access to a press to fit new bearings to either the carrier or the pinion.

But still, having rebuilt engines and gearboxes, I've yet to ever rebuild a diff. This could be my last chance. On the other hand, if I have a go and it doesn't work out, I waste the cost of a set of bearings.

I'll have to investigate the cost of those on bearing boys/simply bearings, and see how much I have to lose.

The most 'sensible' thing, has GOT to be, ring Bill at Hardy's. But I do like doing something I've never done before, and it'll be a buzz if I do it and it works out.







anamnesis

A question on shimming, Malc and Willy, and anyone.

I've now got my head around the whys and wherefores and hows, of selecting shims. It's actually simpler than I thought, now I understand it.

But, --- the bmc manual says you must also compensate for bearing thickness variation.

In practice, do the bearing thicknesses vary from nominal values?




anamnesis

new bearings are usually spot on but there can be a little when comparing used to new but then you wouldn't wind as much preload into used bearings so basically the shims would be the same
William Revit

Anam,

There are published tables available which give the tolerances for various features of rolling element bearings such as OD, ID, race width etc. I have a SKF book which has the details and I'm sure it's available online at the SKF site but it's a matter of finding the right tables as there are so many. Maybe I'll have a look tomorrow.
David Billington

Ok, thanks Willy and David.
anamnesis

Yeah, diffs seem complicated at first glance, but actually they are pretty simple. Especially if you aren't messing around with the pinion depth... which you won't have to if you are using the same CW&P in the same pumpkin as the spacing will already be set.

All you need is basic socket set, a dial gauge and a couple of brain cells! :-) we can help you through the rest. Oh... a decent bearing puller too!

I can direct you towards a company that does the correct size shims for external shimming if you fancy it.

All the parts you need for a diff rebuild can be had for less than £100 (4 x bearings, shims, pinion seal, crush spacer, sun and planet thrusts).

Malc.
Malcolm

Thanks Malc. More and more tempting.

Yep will use the same pumpkin. And since I have two 3.9's that need doing, maybe then I can have both smooth and quiet for circa 200 quid; -- I always love saving money 😂. plus the huge additional satisfaction of doing it myself, -- which is in tune with almost everything that's been done to my Sprite since I got it in 1977.

As I have the 4.2 diff for the cage donation, I'll see if my puller is up to the job of getting the bearings off its cage and pinion. It may not be, it looks a bit too big.

If not, what puller do I need that isn't an arm and 'leg'? (pun pun 😁).



anamnesis

The worst wear point in my diff was the thrusts behind the bevel gears. I don't think new ones are available but Hardy have second hand ones.

Jan T
J Targosz

What brand of bearings would you get in a kit like that Malc?
anamnesis

New thrusts are readily available in both std and oversize.
2A7062 (cupped)
ATA7039 (flat)
Paul Walbran

Paul,

Are those flat thrusts tufnol though as fitted to later diffs as in my experience they don't last well compared to the phosphor bronze or steel ones.
David Billington

Agree on that.
Just checked stock, std are tufnol and oversize (+005, +013) are yellow metal, brass or bronze.
Paul Walbran

On the other hand, it is fair to say that in all the diffs we repair the tufnol flat thrusts usually have outlasted the cupped thrusts.
Paul Walbran

My apologies I checked my invoices and it was the thrusts from the gearbox that were scarce. The ones I used were oversized tufnol.

Jan T
J Targosz

I've only just understood (slow on the uptake lol ) that it isn't only the pinion bearings that are preloaded. The diff cage bearings are preloaded too. I thought before, that all the shimming did was position the crown wheel correctly relative to the pinion. But now I realise that being angular contacts, there's a bit of loading too, to stop the cage, hence crown wheel, 'wobbling'.

Back to brands. A full rebuild kit for circa 100 quid. Not likely to be top brand bearings is it? And does that matter? Since they are tapers, and 40 degree angular contacts, the tolerances aren't critical because you adjust them to fit. So cheaper brands will be just as good. Is that right?

I looked at the spec's and can't find the tolerence variations for the width of 17mm. And anyway, as Willy said the shims take care of that.

https://www.bearingboys.co.uk/search.cgi?search=7207&display=list

I've seen Dunlops on ebay for as little as around 10/12 quid.

https://www.ebay.co.uk/itm/223736213227







anamnesis

The majority apart from say SKF probably come from the same factory in China so its unlikely to matter!
Bob Beaumont

I used KSM bearings for both the gear box and diff. They are genuine Japanese items and supposed to be of absolutely top quality. Unfortunately some of the names us "oldies" remember as being British made are now produced in China.

Jan T
J Targosz

I think KSM are a 'premium' brand. Can't find them for sale uk. But are circa $50 each. So 2 of them plus 2 tapers, shims and crush sleeve, way more than 100 quid.

Pulled the bearing off the other side of the cage. Crown wheel side. Willy was right, it's worse. It was tight to pull off, but it looks like it was spinning on the shaft. And the shims are in the same damaged condition as those fromthe other side.

Looking at the cage for numbers for shimming, there aren't any. It looks like they've been removed.

I haven't got the right kind of puller to get the tapers off the pinion though. I need they type that clamps round the bearing.







anamnesis

Anam,

How do the bearing feel when turned?
David Billington

If you find some SKF or other premium brand bearings that are a very good price, BEWARE! If the price looks too good to be true, it probably is. The market is awash with fake bearings, many of which are very convincing, even down to high-quality repro packaging. However, the bearings will be made of lower-quality steels and the tolerances won't be as good. Avoid!
Jonathan Severn

Thanks Johnathan.

The bearings feel quite smooth when spun by hand David. But maybe when loaded they are rough. If they were easier to get at, and fit, and take off again, I'd be tempted to use them on the other cage and see how they perform.

anamnesis

Looking at a decent puller to pull the inner(front - teeth end) pinion bearing, as I likely won't need it again, I reckon I might as well cut/grind the bearing off.
anamnesis

I bought my KSM bearings from Heathrow Transmissions. I have always maintained that if dismantling has taken a long time e.g. gearbox or diff only use the very best bearings. If it is a simple job you can risk cheaper ones.

Jan T
J Targosz

As Jonathan said---beware of bargains
We've had a run of NGK sparkplugs here--really hard to pick, the packaging is perfect, the only way to pickem is the end of the sparkplug near where the earth tang is welded on--On the originals the end of the threaded body part of the plug is machined with a square end and corner whereas the copy looks more moulded looking with a rounded corner that gives a sort of sandblasted appearance ---silly thing is they seem to work ok--but
William Revit

Addressing the comment re NGK spark plugs I recall that back in the late 60s/early 70s we had sponsorship from NGK for our racing motorcycles. Plugs were usually supplied in 100s but unfortunately we found that we were lucky if 1 in 100 actually worked so finished up reverting to Champion !!!
S G KEIL

Diffs to spark plugs. How about breakfast cereals? 🤣🤣
anamnesis

I have recently discovered the same issue with NGK plugs as Willy mentioned. They seem to work okay in a fourstroke (pressure washer/salvage pump - both Hondas) but can foul up quickly in twostrokes. I've learned not to use them in the LifeBoat engines, but the familiar square finished plugs are still fine.
Just a matter of being selective.

On the bearings question, different manufacturers have different tolerances in respect of width. In general the bore and OD will not vary much but the width might. Hence the need for care in shim selection.
ETA: the variation quite possibly results from some manufacturers using the same shell size for unsealed, half sealed and double sealed races, and other manufacturers taking the trouble to make dedicated sizes for different applications.
You pays yer money and you takes yer pick...
Greybeard

Autogear Transmissions Limited

95 quid inc post.

https://www.ebay.co.uk/itm/110998944307

I wonder what brand the bearings are.

I'll ring them tommorow and ask.








anamnesis

I can't remember what brand they were, but I bought my bearings from simply bearings and used their "premium budget" range.

I don't know if it is still the case, but their budget ranges always used to be EU rather than far eastern sourced.

Although for some people EU might be even worse ;-)

Best of luck,
Malc.
Malcolm

As an aside, how much whine is too much whine? Is any acceptable? I put a 3.9 in the Frogeye and it whines just a bit. I've only changed the input shaft oil seal and re-set the pre-load.
Bill Bretherton

I'm usually ok if I stop at just the one bottle Bill. 😁
anamnesis

🤣🤣🤣
Bill Bretherton

Thanks Malc. Eu to eu too. 😂😂

Rang autogear, they use ntn and koyo.

anamnesis

I tried to check on reviews of various makes of bearings but all respondents commented on the seller and not the bearings themselves. I was suspicious of "high quality LOW PRICES" I know my KSMs are supposed to be made in Japan and are premium quality but as others have said it is easy to print up authentic looking boxes and packaging and stuff them with mild steel races and oval balls. Heathrow Transmissions state in their adverts that they only supply/use KSMs so I trusted them.

Jan T
J Targosz

Malc, or Jan of you did the diff too.

What puller did you use to get the pinion bearing off?
anamnesis

Malcolm,

Where in the uk did you find the diff external shims?
anamnesis

Willy mentioned minimania further up the page like here https://www.minimania.com/part/2K7779/Differential-Bearing-Shim-002-275-Od--Sprite-Midget-Minor-Sold-Each
David Billington

I know David. That's not in the uk though.
anamnesis

It might be worth asking Minimania what shipment service they use as if USPS flat rate international it can be very reasonable IME.
David Billington

This place has been mentioned before for 72mm OD x 56mm ID shims but only down to 0.1mm (0.0039")

https://www.thebearingcompany.co.uk/category/quality-bearing-for-sale/bearing-shims-washer/
David Billington

Thanks David.
anamnesis

Been looking at bearing pullers/splitters.

Local car spares shop has sets for 100 quid and above. Somewhat overkill for me as a one off use.

Ebay has 'similar' kits for as little as circa 23 quid. I can't help thinking that's too cheap and not strong enough to pull a pinion bearing.

https://www.ebay.co.uk/itm/183101678490

But I have a cunning plan. Any old iron, any old iron. Using angle iron, I'm going to see if I can knock something up by way of a 'splitter' type. A couple of bits, shaped and profiled with a grinder, and bolted together.

Nothing to lose but time and scrap iron. 🙂.






anamnesis

Yes that's what I did. I made up a puller from angle iron and lengths of studding. You will need to bolt the two "halves", which fit under the bearing, together and grind a taper on the edges.

Jan T
J Targosz

Did you happen to take pics of your home made tool Jan? No point me reinventing the wheel.
anamnesis

Don't worry about pics Jan.

Back to plan A. Give the diff's to Hardys to rebuild.

Having achieved the original objective, and more, of just wanting to understand how to do it, I've decided I can live the rest of my life without the satisfaction of rebuilding a diff. 🤣.

Even if I could guarantee getting a good smooth quiet result, which isn't guaranteed at all, if I do it myself, I've decided I'd rather spend a hundred quid or so more per diff, and get an actual guarantee. 😁 ; And no doubt get back driving my Sprite much sooner than if I diy it.

Really MUST be getting old. Lol.





anamnesis

Chatting to Bill at Hardy's.

Here's the trick to the diff cage shims.

Have a pair of spare new bearings. Bore out the centres a tad, so they slip on and off the cage with only finger pressure. Set up the diff using internal shims as usual, remove the dummies, fit the permanent bearings.
anamnesis

Anam,

Rebuilding the diff is the easy bit as it only comprises the 4 bevel gears in the cage, the more complicated part is getting the CWP set-up correctly.

Interesting the trick mentioned by Hardy and I would expect use of the same make bearing to consistent spec would be in order as well.
David Billington

Same method with outside shims Anam.--well similar, but we haven't progressed that far with your education yet.
---
Press the bearings on the carrier without shims and then assemble the diff but use 2 screwdrivers on each side on the outside of the bearing instead of shims--The carrier can be moved about by adjusting the screwdrivers in and out until the backlash and tooth pattern is good then it's a simple, measure the gaps, add a couple of thou. and that's the shim thickness needed
William Revit

If you go back to plan B, yes I bought shims down to 0.1mm from the above link.

I got lucky on one of the diffs and 0.1mm resolution was OK. On the other I bought a roll of 0.05mm shim steel to make up thinner shims. The first ones I hand cut, then a fellow enthusiast (sorry... I forget who it was off the top of my head!) cut me a bunch "properly".

Cheers,
Malc.
Malcolm

Ah clever Willy. But still nah. I think I'll quit education and leave school early. Lol.

I agree David, now I understand it, on paper at least it's easy, and if I already had the right pullers, and if all the shims were local (cheaper than shipping from minimania), then I'd have a go, maybe, -- esp' as Willy has now explained how to do it with externals.

But I'll be happy with my original plan.

Maybe I'll look for another cheap 3.9 English diff to rebuild for my Capri, to satisfy my need to have a go. Much easier with threaded adjuster rings on the cage.

But a useful thread I reckon, -- at least for me. Or as Brucie would say, "good game, good game". 😄

If you do yours Malcolm, you can keep it updated.




anamnesis

Anam
At least this thread has got me interested in looking at my original 4.2 diff to see if I can identify what needs renewal/ adjustment then maybe do the 3.9 sometime. All I've ever done is the pinion shaft seal and the pre-load. I've regarded the rest as a dark art!

So, thanks 🙂
Bill Bretherton

Bill-
-Always a good idea when doing a diff if the same C/W and P is hopefully going to be used, is before you pull it apart take a measurement of the backlash and take a tooth marking. As well as giving you somewhere to check/set it to if it's just a bearing replacement job, Doing this will also tell you if something is out of whack causing the noise.

Anam
You could do what i ended up doing with my midget--
105e Anglia(English) axle housing and axles with an Escort LS centre, You can get any ratio you want for them from down in the 5s right through to ,i think 3.3 now, and as you know a bit easier to set up and dare i say way stronger-

willy
William Revit

Hi Willy, I'm happy to keep my Sprite as a bmc axle/diff. Even if I wanted to swap to a ford axle and english diff, it wouldn't be that easy. I looked on ebay today to see if by chance there was a used 3.9 cortina/capri/corsair diff for sale for my Capri. Zilch, nada, now't. Almost hen's teeth. I've looked before, hard to find. When they do come up, they are pricey.

Took my diffs to Hardy's. Goood news is both diffs cw/p are in great shape. So new bearings for both, and pinions, and the cage used from the 4.2 to replace the ruined 3.9 cage.

While I wait, I'll fiddle with the Capri instead. I need to pull the rear screen out and fix a leak in the seal again. Last done 7 years ago, but the sealant I used, Arbomast Autograde, has dried out (even though it claims it won't). Club does new rubber seals, but they don't fit as well as the old originals. This time I'll use a black pu sealant.

anamnesis

I've been playing with a few diffs over the past months and agree with Malc that external shims offer the best approach as the bearings can be secured using chemical bond.
This negates the suspected cause of the bearings losing their interference fit by repeated removal/replacement which requires a specific puller if the bearings themselves are not to become disassembled.
For the input shaft I have reverted to the same approach as the early diffs and made some solid spacers that would permit subsequent removal/replacement without having to use a new crush spacer each time.
However with the limitations of ready availability, cost and thickness limitations I decided to make my own shims covering 0.001" to 0.020" where the centres of the larger size carrier shims were used to make those for the input shaft.
Fortunately I already had a reasonable selection of shim stock collected over the years which was supplemented by a donation from Malc in return for my providing him with a number of completed shims.
Having elected to make my own shims it should be noted there is no additional effort in making dozens as opposed to just a few.
I did similar for the steering rack inner ball joints to effect their refurb.
Unfortunately my shim stock has now been totally consumed !
S G KEIL

SG Keil,

How did you make your shims?
David Billington

I followed information gleaned from internet which requires use of a lathe to enable outside diameter to be turned and inside bored as two separate operations.
A pack of shim blanks are rough cut to size (e.g. oversize squares) before clamping (e.g. 4 bolts) between solid steel blocks.
They are then bored to size before dismantling and fitting onto a mandrel where they are compressed and turned to diameter.
Hence the a actual number of shims is almost irrelevant as the complete stack is usually only a fraction of an inch/few millimetres.
S G KEIL

It's good news week.

Took my diffs's to Hardy's last Thursday.

Wasn't expecting it so soon, they're both ready to collect.

Now THAT'S service. With a smile, and a comensurate price too of course. Ha ha. But, done and dusted. Might blast out there today and get them. 😁

anamnesis

Yes I have always had good service from them. Glad your sorted.
Bob Beaumont

How can you open out the bore of the crown wheel bearings so that you can remove them easily when sorting out the shims. You could lock balls with araldite or molten lead so the outer race can be gripped in a three jaw chuck but how do you cut into the hardened inner race.

Jan T
J Targosz

I'd probably make an expanding mandrel like in the attached image and do half the bore at a time using a toolpost grinder. I made a mount for a Dremel for this sort of job and have similar to mount my other die grinders if they'd be more suitable. The expanding mandrel I would rough turn it then cut the slots and tap, fit the countersunk screw and then finish turn the diameter and face that hold the bore and leave it in the lathe to ensure those features are true to the lathe axis when you grind the bore. If using a self centring chuck mark the master key position before you remove it so if used in future it will still go back quite true.

Another way if taking light cuts would be to put strong PSA tape across the inner and outer race faces.



David Billington

Bearings can be filled with molten wax before grinding od or id as required using previously mentioned set ups.
Machining even using carbide tooling will be problematic and require higher forces than grinding.
Fortunately less than a thou would need to be typically removed.
Unfortunately whilst bearings are made to extremely tight tolerances they are not directly interchangeable. This is especially the case with angular contact and taper roller that need to be set up on the actual hardware typically using shims if no specific adjustment features are included as part of the equipment design.
Chemically bond the bearing inner tracks to the carrier and use external shims.
I've been playing today with a tool to enable the carrier to be moved laterally across the casing to hopefully aid set up with shims as I tend to work by trial and error rather than calculated values.
S G KEIL

Yep, cheers Bob.

External/internal shims. Ask Hardy's. I asked about external shims.

The answer, it's much easier with internal shims and removable dummy bearings.

They've been doing it for yonks with loads of happy customers, including racers, so I guess it must work and be accurate.

I'll let you know when I put one of my diff's back in. 🙂

So is external accurate it seems, so just comes down to choice really.

I agree, lateral adjusters on the cage bearings is simpler, as per the ford english diff adjuster pictures I posted a ways back. Why aren't they all done that way? Cost?

anamnesis

Anam,

A mate has worked for Ford in the past as his employer was contracted to do development on various occasions and he said in his experience if Ford did it that way it was because it was cheaper. Likely the additional machining cost involved was offset by it taking less time to set-up correctly so saving money overall. He told me to look up 'Ford baked potato' as he said it was all to true in his and others he knows experience having worked with Ford, see https://gamefaqs.gamespot.com/boards/271-cars-and-trucks/67390850 .
David Billington

Lol. Brilliant David. Never heard of that before.
anamnesis

DIFF OIL.

Bob, et al.

Bill at hardy's recommends ep140 instead of the usual ep90. I haven't seen that here before.

Do you run ep140 in your diff Bob?
anamnesis

No I run the normal 90. Bill never mentioned this to me the last time he rebuilt a 3.7 diff for me 3 years ago.
Bob Beaumont

David,

That process reminds me of my time at Boeing (aka The Lazy B) not to mention other jobs dealing with government contracts. People wonder why a hammer costs $700.00. It's not the hammer, it's the paperwork tracing the materials all the way back to the mine.

Did you know that the US government has military specs for pizza sauce, canned peas and even condoms? I won't go into the gory details of coordinating a new MIL-SPEC.

Anam,

I'm sure someone in the military has done a study of which oil is best suited for a Spridget diff. And they didn't ask the specialists who rebuild them.
Martin

Anam
Castrol SAFXA 80/140 synthetic, I run it in everything
Best stuff going, it was developed for use in axles with hob cut gears and has all the good additives already in it--Make sure you give the bottle a good shakeup though --If it's been sitting on the shelf for ages the goodies can settle in the bottom of the bottle

willy
William Revit

also sold as -
Castrol Syntrax 80/140 synthetic axle oil
exactly the same as SAFXA 80/140
typical castrol, changing the name all the time for marketing-you can get it here under either name but SAFXA was the original and they still sell it under that name as well as the new---bit weird, I try to find the original SAFXA when i go looking--just me being me

Tell you how good it is-
Ford here recommend it for 'all' their high performance and commercial axles
Falcon cars with hob cut gears(high performance models) use it and if you unknowingly fill one with normal ep90 as soon as they get nice and hot out the road, like 100 klm run they start to whine under load but with the SAFXA they're as quiet as a mouse

It's good oil--get some
William Revit

Military spec pizza. Lol. Of course, there has to be. 😁

Thanks Bob, Willy.

I think he said straight ep140.

So if a 140 oil isn't too thick, I wonder why it wasn't used to begin with?








anamnesis

Here's a bit of information worth knowing.

Not all 3.9 duff cases are the same.

Both diffs rebuilt. Same cages, same bolts used, to bolt cw to cage.

1 diff bolts hit case. The other case they don't.

See pictures. The internal oicture is diff ONE.






anamnesis

The internal pictures are diff ONE, and diff TWO.

BOLT HEADS, Completely clear on diff TWO.





anamnesis

Close up internal diff TWO. Bolt heads clear.

On diff ONE, only one head actually hits. The others are about 1 thou inch clear. One bolt head a gnatt's cock higher.

What a difference a case makes.






anamnesis

If I may be so blunt... how the fudge did they rebuild it and give it back to you like that then?! Surely Hardy should have noticed a collision like that?

Malc.
Malcolm

was thinking the same--bit dodgey--and how could they check for runout and propper tooth markings if you can't turn the thing a full turn---and giving it back to a customer like that-???
Whoever did that wouldn't be working for me for long--

any chance of a pic of the tooth markings of both An.
William Revit

Being fair, you can turn the cw by turning the flange, and not really feel it, because of the gearing. Now I know it's there, you feel a 'notch' as that one particular bolt head goes past. At first I thought it was a bit of fluff or something in a bearing. So setting the lash etc, wasn't a problem I think.

I only noticed it when turning the cw over, and just happened to have that bolt head in that position as it hits/passes the case at that point. All the other heads clear, albeit by a small margin.

The marker has worn off a bit. I haven't got any. Is this clear enough? DIFF ONE.











anamnesis

These two here and the next picture following, are the other diff.







anamnesis

Same diff as previous two pictures.


Apart from the bolt head interference they both feel smooth to turn.

I'm going to 'grind' a tad of the corners of the bolt heads that run close to the casing.

These are slightly higher bolt heads than were in it previously. It must have always run close.

I'm wondering if the rrrr rrrr rrrr I described at the beginning of this thread, was, actually this bolt head/case interference. But it was also a bit whiney and rumbling.

As long as the new bearings fixes that, I'll be happy.



anamnesis

I imagine there's some wondering why I haven't taken it back.

At first I thought it was either teeth binding, or a bearing. I rang Hardy's and they said don't try and fix it bring it back.

I was going to, but then realised it was the bolt heads too close to the casing, and why. So I rang and spoke to Bill, and he was happy for me to grind a bit off the heads. It suits me as I don't have to do the m25 on a Friday (I want the diff in today too), and it saves me 15/20 quid in petrol if I diy.

Anyway. It's done. All bolt heads shaved, and running clear now.









anamnesis

1st diff in. Test, take out, 2nd goes in.

If 2nd good too, no reason not to be, it stays in and 1st goes on the shelf as a spare I'll possibly never need.

Still, nice to have a decent, quite, smooth, spare 3.9 kicking around, instead of a noisey rough one, as I have for over 20 years. Lol


anamnesis

A few queries.
No tab washers shown so presumably chemical thread lock used !
Bolt head thickness should be common for standard bolts !
Unable to read bolt strength grade in photos but should I believe be 'T'.
Length of bolts appears to be less than original !

My preference is for chemical thread lock and cross drilled bolt heads for lock wiring with bolts fitted with self made hardened plain washers of same thickness as original tabs.
S G KEIL

Yep, thread lock used. I asked why not bent tabs, and he said he favoured thread lock.

Said he uses bolts with a deeper head than originals.

These are the bolt head markings on the rebuilt diff's.

As for bolt lengths -- Before I took the diff's for a rebuild, I happened to take pic's that show the bolt lengths too. As far as I know these are factory original bolts. I got one of the diff's new, in a BMC box, yonks and yonks ago. So he's used the same lengths. Must be ok, as they didn't come loose/out before.

They build diff's for a lot of people. Well known. Mine was just an unlucky one off. I'll only be unhappy if they make whine noises and rumble. 😉.









anamnesis

Anam,

Here are the matching details from my unused NOS 3.9 final drive unit. Screw depth looks to match and the hex screw looks to be a W grade made by GKN.





David Billington

Thanks David.

Yep shallower bolt head. GKN were the bolts in mine before the rebuild.

I wonder why they didn't use a longer bolt thread.

Was that for sale David?
anamnesis

SG KEIL.

"I've been playing today with a tool to enable the carrier to be moved laterally across the casing to hopefully aid set up with shims -"

Are you making your own tool for this? It sounds interesting.

Anyway, I've got a twinge of buyers regret I think. I think I wish I'd bought the tools I needed, and done the rebuilds myself.

I found my old thread from 2021 in the 2021 technical archives.

"Diff questions. Various."

https://mg-cars.org.uk/cgi-bin/or17?runprog=mgbbs&mode=archiveth&archiveyear=97_2021.dat&access=&subject=97&subjectar=97&source=T&thread=2021070519314368298

I'd forgotten just how much I had done previously, and what my decision was then.

I'd already stripped my 4.2 diff, and used the very good planet and sun gears, and the bronze thrusts from that cage, and bought 2 new cup washers, and used it all, to remove 'lash' from the cage of the 3.9 diff I was actually using at the time. The intention was to see if it quietened it down. And to have the other (spare 'new'seized') 3.9 diff rebuilt.

Then I forgot I'd done that and my decision.

Fast forwrd to 2023. My currently in use 3.9 diff whose cage I rebuilt, hadn't been rendered quiet, but the 'lash' at the wheel rims was greatly improved. And I'd unseized my other 3.9 diff, and found it to be a bit noisey, plus rough and running with a rrr rrr rrr, that increased with speed.

So I figured the cost of me doing one diff, -- buying the tools and bearings etc, and having one rebuilt, was a false economy, as I wasn't sure I could do a good job anyway.

Now, I've perhaps changed my mind, -- somewhat too late to catch the horse that bolted. Lol. I should have costed it better. Doing two diff's myself, I could have bought the tools, and all the bearings etc, and still saved money, and had the satisfaction I thought I wanted earlier.

My Ford diff needs attention. If I keep the Capri this year instead of selling it, instead of trusting the rebuild to someone, as the Ford diff has cage bearing side adjusters (no need for shims and repeated removal of the cage bearings), all I need is a decent pinion bearing puller, which will be paid for in not paying someone else to do the job.

Event, after, easy, wise, could be rearranged to suit me right now.😅







anamnesis

I think we've probably all paid to have something done that we might have done ourselves. With the Frogeye rebuild I used a partial new shell (the "tub") which wasn't perfectly aligned, despite going back twice to the manufacturer. But the price was right (at the time, not now) and it saved me considerable time. Also I wasn't confident at making safety critical welds.

But you make a decision and move on. No point "wishing I'd done that myself".

I guess you do what you think you're good enough at. I always thought a diff needed very precise measurement and alignment which I wouldn't be confident enough to do (although I've rebuilt gearboxes and engines). Or is it just about the tools cost for you?
Bill Bretherton

Anam,

Why were the bolts changed by Hardy, is it a requirement to use new or did they think it a good idea to change them. Not sure of the new bolt grade in yours, maybe P which is somewhat lower grade than the W in mine.

"Was that for sale David?" Yes which is why I bought it ;), it's not for sale now. It's the 2nd I bought from the local MM parts centre when they had a batch of NOS ones in stock some years back. The 1st is in my car and has a few thousand miles on it before I took it off the road. When I pull it I expect I'll pull the CW and cage and check if the sun gear thrusts are tufnol and change to PB if they are.

Regarding pulling the pinion race I used one of the cheap 2 part plate type you listed earlier in the thread to remove one although I put it in my hydraulic bench press rather than use the threaded extensions not because they wouldn't work just I have the press and it was easier. When my neighbour replaced the CWP in his Morgan he used one of those cheap ones and it worked fine although he did have to fettle the edges to get behind the race in that case.
David Billington

David. I don't know if the bolt change is requirement or just a choice. Maybe becuse they build for racing, and standardise on heavy duty?

Yep, partly cost of tools Bill, and also because I don't have the tools to make precise measurements. But I'm beginning to wonder, just how 'precise' you need to be.

Bill.
A little while back, you asked about whine. How much is too much?

Did you get a serious answer? I can't find it.



anamnesis

I checked a 4.22 I have on the bench and those bolts have the makers name 'NEWALLOY' but no grade marking letter so I don't know what that indicates. Apparently the circular depression indicates UNF.
David Billington

Anam
No answer so far....
Bill Bretherton

Thanks David.

Seems not Bill.

And if you try searching, you come up with a myriad of answers, but nothing useful on how much you should be able to hear the diff.

Do ALL diff's whine, to some extent -- even when new/rebuilt?

Not much sound proofing in a Spridget, so maybe you always hear a bit of diff whine.

How 'noisey' is yours Bill?

Has anyone here got a totally silent diff in their Spridget?

Meanwhile on oil. Willy there's a TRANSMAX Axle EPX 85W-140, but it's a GL5 oil.

Haven't so far found a multigrade 140 at GL4.


anamnesis

Here's an 85-140 GL4 EP Oil.

Motul HD 85w-140 Extreme Pressure Transmission & Differential Oil

https://www.opieoils.co.uk/p-92079-motul-hd-85w-140-extreme-pressure-transmission-differential-oil.aspx

anamnesis

Mine's not very noisy, nothing to trouble me really but there IS a whine and, as you suggest, maybe all diffs whine a bit (well classic cars at least). Surely that fraction of movement called backlash produces some noise, despite the EP oil.
Bill Bretherton

Any particular speed that it's worse or better Bill?
anamnesis

If you want to live in the past there's always the Castrol Vintage/Classic 140 gear oil it's a mineral oil in GL4, but why would you bother when you can get a real oil
SAFXA 80/140, yep it's GL5 but what's your issue with that---it's a diff isn't it
GL4 's a gearbox oil, the differance being GL4 is designed to keep the syncro rings clean where GL5 not so much , but the newer synthetic 'clean gear' oils work ok with syncro's anyway
William Revit

Not really Anam - it's very slight actually, maybe mostly around 50mph, but hard to hear above the exhaust note so nothing to worry about I presume.
Bill Bretherton

Hi Willy, I can't find that oil on sale here.

GL5, is hard on bronze/yellow metals. The cup and thrust washers in my Diffs are 'yellow metal, not fibre. Hence GL4.

Hardy's use Smith Allen oil. I spoke with them and they too said not to use their 85/140, because it's a GL5.

The motul is GL4
Motul HD 85w-140 Extreme Pressure Transmission & Differential Oil.

Never used Motul. Are their oils good?



anamnesis

I've posted it before that I've seen online that the yellow metal attack problem is only an issue at elevated temperatures that won't be seen in our cars. IIRC there's a test spec for the test to determine the effects but what the spec is I don't know.
David Billington

I've never used Motul but it is used a lot in racing so should be OK. I must admit I had never heard of them until they started sponsoring at Goodwood a few years ago.

Trev
T Mason

Anam---you worry too much
SAFXA 80/140 is the best choice in my view but if you want to run ancient oils then go the Castrol classic EP140----you should be able to get it there ,it's probably made there----I tried uk ebay and it's there, but be carefull selecting as there's also non EP 140



https://www.ebay.co.uk/itm/181731602840?epid=8017010563&hash=item2a500c3198:g:-SwAAOSwI9Zgfp6P&amdata=enc%3AAQAHAAAA4GhhVcezHEV%2Fj4PcblSWVAkHYNYZtsAWLs25cRnaFQs4SH56OrxV2aElRE7wFompQ%2FOhgnF6yjWHi8DbOVuqK7WwvCKCDsAeKLXhi0u%2FNt23ocyXscEwkBTW%2F8kp2LarvC5l6LPwQEApZY%2B1aaQOA3qqnXNDmL4w%2Bi5vVyby6ax2pnYt%2BLhAfE%2F3s2wuabV3ZX%2F6KVxQFBK6adQXmuYzwITqD4kAMedM2XCTFV7D2qVilVCt7sBmuP3pGgtBtI2pOWGOyIPDgbVkVflrg8HjNQxEB%2B98WpeP3tFbnYF2kkZt%7Ctkp%3ABFBM4qKGvs1h

William Revit

Just checked Millers and they suggest GL5 EP90 and note OK for brass etc.

https://www.millersoils.co.uk/products/classic-differential-oil-ep-90-gl5-3/

R.
richard b

The Motul oil mentioned above at Opie oils is listed as GL4 and GL5. Having looked earlier there seemed to be quite a bit about this online and much of it 20 years ago and the consensus was modern buffered GL5 was fine for yellow metals but check the specs and tests were mentioned for the effect on copper and therefore copper based alloys. Much of the questions were about gearboxes and the effect on synchros but some on use in final drives.
David Billington

Some call it worry. I call it sensible caution (after shelling out 700 quid), to be certain the oil I use won't do damage to my diffs.

Yep, you're right David. The motul is a GL5 oil. Of course it also meets/exceeds gl4 standards. I just saw the gl4.

And yep, Gl5 is no longer corrosive to yellow metals, because as you say it's phosphor is buffered these days. But corrosion is only one aspect.

EP additives work by creating a sacrificial layer on the gears. On steel this is good, because the steel is stronger than the bond, so when the ep comes off, it doesn't take steel with it. But yellow metals are weaker than the bond. So as the sacrificial layer comrs off, it rips off particles of yellow metal.

With Gl4, the bond isn't strong enough to do damage. But Gl5 has 50% more EP additive than Gl4, and the bond is much stronger.

In transmissions, with brass synchros, Gl5 is no go. With no brass, you can use gl5.

The question of diff's with yellow metal is less clear cut. Are the yellow metals subject to pressure loads, sufficient to cause a problem. How much scuffing and pressure, are the cup and thrust washers behind the bevel gears subject to?

I think caution is reasonable. 🙂.

The question is though, why do I need 140? I'm still not sure why. But I can see the logic of a multigrade.

Never 'thicker' than an 85 when cold, and as it heats up, behaves more like a 140. Whereas a 90 or a 140 when hot, gets thinner, less protection.




anamnesis

You need 140 because the dudes that charged you 700quid to give you back a diff that won't do a full turn without you grinding bits off it are scared it might make a noise when it's hot because they set one of them up with the pinion a bit high and couldn't be stuffed pulling it out again to fix it----
From what I've seen here, I'm NOT a fan of their work

Can you get Motorol oils there, used a fair bit in motorbikes---theyve got a GL4 90/140 oil

https://www.motorollubes.com/wp-content/uploads/2019/05/83.MOTOROL-AUTO-GEAR-EP-90140-GL4.pdf
William Revit

Thanks Willy. No comment at the moment. 😐

I've now had both diffs in. Diff 2 is still in the axle.

Questions.

This is Diff '1'. If I read it right, that's a 7 thou backlash marking on the cw. According to the bmc manual, that's a precise figure. But does that only apply when new?

When checking backlash, will the remaining oil on the teeth, alter the reading much?

This is after running to test it. So there's still oil coating the teeth. Is that oil enough to cause a low reading?

Can't find that motorol oil so far, but I'll keep looking.













anamnesis

Additional question.

I've been getting my head round hypoid gears. Specifically the importance of the backlash measurement. In hypoid gears like ours, two teeth are always in contact. As one frees up, another follows into mesh. If the gap, the backlash, is too great, it slams in. That damages and makes a noise.

If that's the case ---

On *used* cw/p pairs, is the tooth pattern vs the backlash, a compromise?

Is it that, to get the best tooth pattern, backlash may have to be under or over ideal?

Or does backlash have more importance than tooth pattern?

In simple terms, I'm wondering, is what I've got, the best that could have been achieved with my two used diffs?

anamnesis

the .007" is the spec the gear manufacturer would have written on it at manufacture
What happens is the the cw/p are set up in a lapping machine where they can move the pinion in-out and the crownwheel sideways to get the correct tooth pattern during and at the end of lapping--When they've finished lapping the 2 together and get an acceptable final finish and tooth pattern it's then that they etch the final settings onto the set so that they can be set up to them exact measurements in the diff housing later on

Yes the backlash will be less wet but probably only a thou or two-if you can secure the pinion so it can't move and put a bit of load on it might be a bit more

The backlash spec on a 2nd hand set isn't that important as long as it's got some and not too much, -the tooth pattern is the target, If it's got a good pattern it's good to go, don't overthink it, if the pattern looks good and it's quiet, then it's good, forget about it, they might have had to set it up tight to get a good mark
diff#1 looked good
diff#2 just looked a bit off to me as though it needed the pinion dropping in a couple or three thou but it might just be that second pic of it looking like it's got runout with the carrier----What did they do to fix that loose bearing issue

Motorol is made in India so it mightn't have got to UK There's not a lot of it here eiter but the bike shops have it so it's got half a chance of being good oil
William Revit

Hi Guys,
Just trying to help break a record .... 122 comments so far.
Firstly I have to admit that I cannot believe how much I DON'T know about this stuff. I must read up on those oil prefixes!
Secondly, I do like Willy's suggestion of 85/140 oil in a road car because that means oil stays relatively consistent in viscosity.
Thirdly, Yes, they can be silent .... I have a 4.55 (Morris 1000) diff in my road TC and it makes NO noise. The 4.2 and 3.7 diffs I use in my race TC, I am pretty sure are silent, but can't be 100% sure because of the exhaust noise.
Fourthly, due to my lack of tooling and knowledge, I just set up diffs so the BLUE markings are right for "drive" and "coast". I believe that backlash and other measurements are only a way of getting to the perfect "blue" marking anyway.
Fifthly, I have never read anything about this, but I know aluminium expands more than steel. Therefore "preload" on a pinion should be minimal because it will increase with rising temperature due to the way the taper bearings work. Whereas "preload" on a crow-wheel will diminish with temperature. There might be different pre-load data for the cast iron housings, but they are few and far between. Or should we be warming our diffs when we set them up? (I never have.)
Lastly, Willy, if you are coming to the Historic Phillip Island event in a few weeks, Do drop in to Peter Cundy's Pit (MG NA 1934) to have a chat.
Cheers,
Bob Schapel
Bob Schapel

Thanks Willy. I'll clean it's teeth, and load the pinion, and measure the lash again.

Recap.
Diff 1, the 7 thou spec 3.9:1, was pretty quiet, but a bit rough, with rrrr rrrr rrrr increasing with speed. Maybe knackered bearings all round, but cage bearings loose on shaft one side. Signs of the inner race spinning, on crown wheel side. Knackered cage.

To fix that, the good cage was taken from my spare 4.2:1 diff, and my 3.9 cp bolted to it. New pinion and cage bearings fitted. All put back into the original 'pumpkin' with it's matching pi nion.

On testing. It is now smooth, but gets noisy at over 40mph through to circa 65, where it quietens down. It still has a bit of rrrr rrrr rrrr. Less, but still there, and the noise is not good.

I'm going to check it again, but this is what I get for run out.

Question.
Can run out be accurately checked with the pinion in place? by turning the cp, by the pinnion flange, as I am doing it.

Book says max 2 thou. I'm getting cirva 3 thou +. Will check again.

And I've got some red poweder dye. I'll mix it with some oil, and see if I can get a tooth pattern again.

What a way to get an education. 😵‍💫






anamnesis

Oil cleaned off. Nice solid contact. No oil cushioning.

Gone from circa 3 thou to almost 5 thou.

Checked it in a few places, different teeth, a gnat's under 5 thou backlash.









anamnesis

Diff 1 runout checked again.

Zero to 3 thou+, up and down around circumfrence. Visible bit of back of C/W looks clean and flawless.





anamnesis

Still Diff 1. Moved in a bit, just in case the edge is raised. Same result essentially, at the same spot. 3 thou plus runout.

Could this be the rrr rrr rrr?



anamnesis

Diff1 Tooth patterns. An abundance.


Hopefully someone can tell me good bad or indifferent.





anamnesis

More Diff 1 teeth.







anamnesis

And Diff 1 again.





anamnesis

And Diff 1 again for good measure, different teeth.

Verdicts?


anamnesis

You've got way too much goop on it to get a decent reading
Wipe it all off
Sparingly paint both sides of one tooth on the crownwheel ,just enough to cover it
Turn the crownwheel so that that tooth is in the pinion gear, then holding the pinion flange for a bit of load get a spanner onto a crownwheel bolt and while loading the pinion up work the painted tooth backwards and forwards through the pinion half a dozen times then wind it back up the top for a look

3 thou runout is more or less max. but if you've got even backlash readings it should be ok
William Revit

Bob
Thanks for the invite-would be great but I'm still not travelling yet, but yeah Phillip Island, should be a good weekend
Also
Sorry to hear about IanCW, a legend, a lot of history gone there--rip Curly

willy
William Revit

In case it is of any help, I have attached a scan of the diagram I work from. It comes from the Jaguar 3.8 Mk 2 WSM. Wouldn't "Bearing Blue" be clearer to see the mesh pattern? (Available at many motor accessories stores.)
Bob Schapel


Bob Schapel

I haven't got any blue, but the yellow looks clear enough I think, --- albeit back to school then, coz I used too much. Lol.

Ok, thanks Willy, will do just the one tooth and report back.

But now that I understand the mechanics of pinion depth, **if** those marks are accurate, the pinion is too high. If it were lowered, it would also increase the backlash.

But there isn't an easy way (is there?), to drop the pinion hight.

https://www.drivetrainamerica.com/blog/pinion-depth-where-do-i-start/

Acording to the bmc manual, max allowable runout is 2 thou. Mine's at 3 thou, and I'm getting rrr rrr rrr increasing with speed.

Isn't that the effect of the runout?

I figured, give the diffs to someone else to do, so I don't screw it up, and waste money and time, and I'd get back two smooth and quiet diffs.

What I've got, is I'm spending/wasting time, trying to understand why both diffs are still noisy, and one still has the rrr rrr rrr it had to begin with( albeit a bit less). And the other diff, as well as noisy, has too much backlash, and when I'm coasting, I can feel it constantly 'hunting' / 'chattering' between coast and drive.








anamnesis

i use ochre, others swear by blue--each to their own but i find a bright ochre is easier to read---used to use redlead/oil but redlead seems to have disapeared from my local boat shop--probably doesn't mix with fish so now use any light/bright coloured ochre/oil mix.

Did you get a warranty with your rebuilds--

I'm a bit interested what method they used to correct the loose side bearing---is that the carrier with the runout
William Revit

I'm only using a yellow paint pigment with a bit of gear oil. Nowhere near me sells the right stuff, not even the local car spares shop.

I'll have to order it online.

This is the best I get drive side, coast I can't seem to get any clearer tham earlier pictures.

Yep, Diff 1, the one I've checked for runout, is the one that had the loose bearing. It had a very loud rrrr rrrr rrrr. It's cage was fooked.

To correct it, they used my spare 4.2 diff, took it's cage, and bolted my 3.9 cw to it. But it is still making an rrrr rrrr sound. Not as bad, but it is still there. And it begins whining quite loudly at circa 40mph.

I didn't ask about warranty. Until now I've just assumed it won't be an issue, because they are so well known.

I haven't rang them back yet, because I wanted to know more, understand more, what I'm talking about when I do ring them.

But I'm sure that's not an issue, it says on their website all work is guaranteed.









anamnesis

Anam
I'm a bit confused. You spent £700 on two diffs, so £350 per diff say, and one is still moaning and whining plus bolts needed grinding? I find that extraordinary - did they have the apprentice on it?

I'd be quite angry and taking it/ them back and asking what they'd done exactly.
Bill Bretherton

Anam

Bill's got a point. I have used Hardy's for many many years (when Dave Hardy started the business) and have never had a problem with their diffs. I am sure Bill would be concerned that you were not happy with the work that has been done. Why don't you give him a call??
Bob Beaumont

Taken Diff 2 back out.

Backlash if anything is increased, is 11 thou now. CW mark says 8 thou.

Surely this is why I can hear, and feel it 'hunting' when coasting. And maybe why it's still whiney.

It's better than before it was rebuilt with new bearing. Before it was rebuilt it was pretty rumble rough, and howled. The bearings were knackered. It's smooth now, by the whine is too much.

Both diffs are whining Bill.

Diff 1, was actually quieter before, but the rrr rrr rrr is quiter, and doesn't rock the back end as it did before rebuild (loose bearing).

I will call Bill Bob. I just wanted more knowledge and information before I do.












anamnesis

Backlash on a different tooth. Same 11 thou.


anamnesis

Taking diffs back to Hardy's tomorrow.
anamnesis

Yep--best idea is to take them back and explain what is happening with each one-
Don't go pulling them apart if you're taking them back, if that was happening they'd probably tell you to go away
Ochre is easy to find-most hardware stores have it for colouring up bricky's mortar----or your local art store

Looking at your nice yellow pics, yeah the pinion on that pattern shown there is sitting way too high and it needs more backlash-----If i was doing that particular diff i'd try about 3thou accross the side shims first which would move the mark back out towards the heel of the tooth and probably that would be enough to lift the mark up out onto the centre of the face as well, this is where outside shims are good ,you can try moving things about without having to press things on and off (or adjuster wheels on your english capri diff)

It is a bit of a combined up down in out adjustment to find the right spot , each adjustment also effects the other
The 9" Ford axles are the perfect setup, they have the pinion in a seperate carrier so you can set up your pinion preload in that carrier the there are shims between the carrier and the main housing and spinny wheels for the crownwheel---magic to set up, just leave the pinion shims out and then it's easy to just move in and out on the bolts and screw the crownwheel around until you get the perfect mark then feeler gauge the gap between the housings and select a shim -finished-- but that doesn't help much here.--sidetracked again.

If you look at the tooth next to your sample you can see where it's been covering most of the tooth in the past----with a used set like this the pattern needs to be very close to where it's been running in the past or between that worn in mark and optimum otherwise it'll be noisy

The correct patern is as Bob's chart shows
Dead centre of the tooth (face-flank) for pinion height
Centre of the tooth (toe-heel) but biased towards the toe for crownwheel(backlash)
A perfect mark would be say about 55% towards the toe and this will move towards the centre of the tooth under full load
The mark should never be off one end or side of the tooth leaving the other end or side untouched

Yep-----I'd be taking them back
William Revit

Thanks Willy.

I don't mind the sidetrack one bit, it's all interesting and informative stuff.

No, definitely not altering them myself. As you say, that's not fair on Hardy's, and anyway would just confuse getting the correct positioning.

But if I was offered a return of money, I would alter them both myself, now that I have a much greater understanding.

Let's see what this morning brings. Setting off soon to return them.
anamnesis

Update.

After a diversion trialling a replacement diff, and testing ep140 oil, my two original diffs are back for backlash adjustment, and a bit of planet gear swapping. Then hopefully Monday, or soon after, I'll have them back with a smile on their crowns, and on mine. And my world will be a complacent passivity of contentment again. 😉



anamnesis

While waiting to get my two rebuilt diffs back again, I've been considering crown wheel/ring gear (whichever term you prefer) runout.

Recap.
I have two diffs. Both have been rebuilt.

In testing my DIFF 1 after rebuild, as well as whining, it was still making an rrr rrr rrr noise (albeit far less than before rebuild), that increases with road speed.

I measured the CW runout at 3 thou plus. The BMC book says MAX 2 thou. If above 2 thou, take apart and clean.

I've been told this amount of runout is not significant.

BUT, the backlash on this diff was set to just under 5 thou (dry measurement). When coated in oil, it was just under 3 thou. The BMC book says the etched 7 thou on the CW, must be strictly adhered to.

My logic says, gear oil is not compressable (much if any), so even though some is squeezed out between the CW an pinion teeth, the resultant backlash in use, must be significantly less than the required backlash.

My thinking is, the excess runout, combined with insufficient backlash, is the cause of the remaining rrr rrr rrr.

The rrr rrr rrr occuring, at the point of peak CW runout, in the teeth of the pinion.

That is, if there is more backlash, it will 'absorb' the excess runout, because more backlash, is greater teeth separation.

Some supporting evidence.

The other of my Diffs, DIFF 2, does not rrr rrr rrr. It also has about 3 thou runout. BUT, the backlash was set at almost 11 thou, compared to the CW etched value of 8 thou. I can feel this extra backlash when driving. It's annoying, and I think contributes to/causes the whine noise it makes. I also think the excess backlash is why the excess runout, does not result in rrr rrr rrr in this diff.

The backlash settings in my diffs are being corrected to the etched values.

If the result is, DIFF 1 no longer rrr rrr rrr's, and esp', if DIFF 2 starts to, it will prove I'm right about runout and backlash, being as important as tooth pattern.

I hope to know early next week.




anamnesis

Meanwhile, in the intermission; ---

Here's a bit of informative fun. 😁





anamnesis

🤣🤣🤣.

Forgot the link.

https://youtu.be/K4JhruinbWc







anamnesis

Here's an example of an unusual differential arrangement I was told about a few decades ago but couldn't find details about till now, that of the Super Sentinel steam wagon about a 3rd the way down the page here http://www.msmec.org.uk/index.php?page=brighton-model-world-2016-2 . The sprockets aren't fixed to either end of the crankshaft but they rotate with it as they're geared together along the crankshaft to form a differential.
David Billington

That's really interesting David.

I'd like to see a decent description of how it works. Just spent 30 min's searching, with no joy. Lol
anamnesis

It's a spur gear differential with a couple more intermediate shafts. There are some animations on youtube if you look for 'spur gear differential'.
David Billington

Ok, thanks David.
anamnesis

Interesting Dave-
it looks very much like the same principal that Gold track use in their torque sensing axles, where the drive goes to the wheel with the most grip eliminating/reducing wheel spin
William Revit

I read some interesting stuff the other day, --- I'll try and find it again.

EV's are quiet. So quiet that conventional diffs are becoming too noisey. One solution is a motor on each driven wheel. No need for differential gears.

But I read a lot of money and research is going into making gears quieter.

Anybody read that too?

anamnesis

This is similar to what I read somewhere else.

"In a previous blog, we discussed gear whine and its importance in EVs because it becomes the dominant sound when there is no ICE noise. We explained how gear meshing and gear deformation are important for the Noise, Vibration, and Harshness (NVH) "

https://blogs.sw.siemens.com/simcenter/bevel-gears-whine-no-more-about-unseen-pressure/





anamnesis

Asked my Tesla mechanic mate here about noises and she said the big issue is rocks and stones that get in on the stone trays, She said you can hear them rolling around on the tray when you go round a corner, so it's up on the hoist ,remove the tray and empty the rocks out----never had an issue with any other noises except had to replace one(1) engine with bearing noise--
William Revit

There used to be a company called Torotrack who had developed a system of integrated torque sensing hub mounted gearboxes. Integrated in that they 'talked' to each other to share the available engine torque. I looked into it in some detail and was sufficiently impressed to take a punt with £1500. They did really well for a couple of years but then rapidly crashed to zero when I think development had become too dependant on a single company and a contract, ISTR with Chrysler, failed to get signed up. I think they were just a few too many years ahead of the times!
GuyW

Did you get any return on the 1500 quid Guy?

I'll bet the last thing anyone ever thought about was noise from the stones rattling around Willy.

Still, you know what Mick jagger says. --

It's only rock 'n' roll, but I like it. 🤣

https://youtu.be/JGaBlygm0UY



Amazing. 1970's Meccano spur gear differential.


https://youtu.be/imwSdJBFOBQ


anamnesis

Willy,

I don't think so regarding the similarity to a torque biasing diff but I can see where you could see a visual similarity in the gear axis alignment. The Sentinel is just an unusual arrangement of a spur gear diff and so should transfer torque efficiently from one side to the other as in a common bevel gear diff. The torque biasing diffs of the type you mention rely AFAIK on the inherent inefficiency of the gear types used to transfer torque, an extreme example would be a worm gear drive where above a certain ratio it's considered non reversible, ie the output can't be turned to drive the input. It's that inefficiency that prevents the diff allowing too much transfer of power to the unloaded wheel.
David Billington

David
Yep, mine was just a general observation of the look of it, mistake on my part relooking at it. As you say the goldtrack system relies on gearing to control drive, in a crude sort of way--The best and most usefull part of their system is that a wheel that looses grip can't spin up faster than the crownwheel speed, as happens with a normal unrestricted spiral bevel sun/planetry gear set----It amuses me a bit(amuses is probably the wrong word) but i've run schools on drivelines and some people can click onto the principal of various differentials straight up and others really struggle with it until they have that lightbulb moment----strangest thing is they usually think they have discovered something new and become the instant expert--can be very frustrating at times

willy
William Revit

Both my diffs were delivered back to me on Monday, with a voluntary refund of my own travel costs, -- which was a nice and welcome gesture. Unfortunately, 6 days in, the 'kin covid still stops me feeling like being in the garage to do anything at all with them. As per my request, on both diffs the backlash measurements were redone. So I'm hoping both will be smoother and quieter, than they both were before being reset.

On the ep 140 oil question, I was told it wasn't only my diff, they recommend it for all the diffs they rebuild. -- Old classics, obviously not moderns.

As I mentioned back on the 24th, at Hardy's suggestion I tried the ep140 on a replacement/swap 3.9 diff. Not only did it make no difference at all to any noise level (I tried ep90 in the same diff), the ep140 added a noticeable drag when cold.

I tried it when it was very cold, frosty/freezing outside. Before it warmed up after a good few miles, the thicker 140 felt like I'd partially left my handbrake on. I could feel the drag/deceleration when coasting slowly, below circa 20mph. Not imagination. When you've been driving the same Sprite for over 45 years, you know when something is different.

I'll stick with ep90 in my diff.



anamnesis

Hi Anam,

Your latest post is interesting. I keep the Sprite in my workshop and if I am doing any dirty work I push it out to prevent any damage to the paint work. I rebuilt and fitted a 3.9 diff last year but don't recall which oil I used. I have noticed that it has become difficult to push the car and it doesn't roll - just like the handbrake is on a notch or two. I put this down to my advancing years but perhaps drag in the diff is the problem. I must try pushing the car after a run whilst the oil is hot.

Cheers,

Jan
J Targosz

Jan I don't know how much effect it would have a such slow speed. Whereas at circa 20mph, I assumed it was the oil being churned. But I guess perhaps even sitting cold in the bearings, a thicker(too thick) oil might add drag from stationary too.
anamnesis

Anam,

Did Hardy give you an explanation for why the settings were incorrect?
David Billington

Hi David.

The backlash originally set in the rebuilt diff's, was a consequence of setting the diff for the best tooth pattern of crownwheel and pinion. So arguably, the backlash settings were 'correct'.

On Diff1 that resulted in very 'tight' backlash (5 thou max dry). Crownwheel mark 7.

On Diff2, that resulted in a very 'loose' backlash (11 thou min dry). Crownwheel mark 8.

Also, after rebuild, both diff's have a runout of 3thou. This means the backlash varies. I didn't measure the runouts before rebuild, but the book says only a max of 2thou runout is permissible. On the loan/exchange diff I was given, the crownwheel had never been taken off the cage, and the runout was barely 1thou max, which meant a much more consistent backlash.

Now, I freely admit to my new found understanding of the complexities of setting up rebuilt diff's. But still, I think my logic is right, that the tooth pattern shouldn't ignore/override the backlash, because excess and insufficient backlash, can both cause noise and other problems. Also, according to the book, a 2thou shim varies backlash by 2thou. I'm equating runout, to adding a shim of the same value at peak runout. I'll know if my logic is, or isn't right, when I refit and test, both reset diff's again.

After rebuild both diff's had a very noticeable quite loud whine (different tones).

DIFF1. In addition to the continuous whine, there was an rrr rrr rrr noise. I reasoned this to be a consequence of the tight backlash and excess runout, causing the teeth to mesh too closely at peak runout.

DIFF2. In addition to the continuous whine, I could feel and hear the excess backlash. But I reason, there was no rrr rrr rrr sound due to runout, because the teeth could never be in too close a mesh, to cause it.

My argument is, setting the backlash close to the marked values, whilst admitedly altering the tooth paterns away from 'perfect', will cure the rrr rrr rrr on Diff1, not create it on Diff2, and reduce(eliminate hopefully) the excess loudness of the whine, on both diffs.

I'll know if I'm completely illogical when I bolt them in.

It may be, that due to cw/p wear in both my diffs, they can never be quiet. If so, I wish I'd known that before paying to have them rebuilt. But I'm hopeful.


anamnesis

I had taken it that the figure for backlash engraved on the carrier in the factory was as measured on the completion of assembly and would therefore take account of any crown wheel runout. I doubt that this measure was taken at the single "worst" tooth (if there was variance) as they wouldn't want to take time to measure every one.
More likely the teeth were ' blued' for contact and lash measured at just a few positions to check that the whole setup was within the design tolerance and it could be installed. I doubt an assembly line workshop would have done any more than that with thousands being dealt with.

They may have had a quality control check of a percentage of assemblies checked to see that the figure claimed by the assembler (i.e.the engraved figure) was correct and not just a quick guess so he could go for a fag break!
GuyW

Guy
I had always assumed the engraved figure was the final backlash after the crown wheel and pinion came off the lapping machine at manufacture such that the assembler of the diff knew what was correct for that unit.
Don’t know which is correct ?

R.
richard b

Richard,

I believe you're correct that they're put there by the manufacturer of the CW&P set as I helped a neighbour install and set-up a new CW&P in his Morgan and they came with the figures already engraved and that's what we worked with.
David Billington

I agree too.

But for a moment/while, you had me wondering Guy.

The cw/p are made and finished as a pair. There is an optimum distance the crown wheel must be from the pinion, to ensure smooth quiet tooth meshing and long service life. That distance is achieved in the complete diff' assembly, by measuring backlash.

As explained in section H of the bmc manual (See picture), describing the refitting of an existing cw/p after replacing the bearings, -- the optimum backlash figure for a matched cw/p pair, is etched on the cw.

The question (my question) is, notwithstanding cw/p teeth wear, how much should tooth pattern override the backlash figure etched on the cw? I argue, not that much at all.

Also, in instances where runout can't be measured accurately, -- uneven rear face of cw for example, backlash should be measured at a number of places around the cw, by turning the cw through the pinion. That gives an indirect measure of runout. If backlash varies significantly, there's a problem with runout.



anamnesis

Anam,

You've forgotten that the pinion height needs to be set first according to the value marked on it, once that is set the CW backlash can be measured and set accordingly.

It should be interesting to hear how the reset final drive units run when you get them back in for a test.

Hardy setting them up according to the contact pattern reminded me of the 'specialist' Morgan supplier that supplied my neighbour's new CW&P, the supplier admitted that he didn't bother with the numbers marked on the CW&P rather relying solely on the contact pattern to set them up. He said the numbers only applied to a perfectly machined housing which you wouldn't come across so he didn't bother with them. That statement really screamed at me that he didn't understand the correct set-up procedure and why the numbers were valid in any housing they were built into.
David Billington

😁😁 David.

Not forgotten the pinion height. I'm ignoring it, on the basis that hopefully it was either correct in both diffs, or it too has been reset to correct the backlash. Arguably, the new pinion bearings are the same 'thickness', and the same pinion height as new will have been restored.

Yes, I agree with you about it being wrong not to consider the etched values.

But it seems new cw/p's for our diffs, now have no such markings, hence tooth pattern alone is used as a guide.

However, I understand there is still a ball park 7-9 thou backlash requirement, but I can't find where I read that now.
anamnesis

Anam,

It may have been mentioned earlier in the thread but while bearings are made to very tight tolerances certain types such as angular contact and taper roller can't be considered interchangeable and should be set-up again in this case that would mean checking the pinion height to confirm it was correct after changing the bearings.
David Billington

Yes, that makes sense David. But I wonder just how much in terms of thou inches, the tapers actually vary from nominal.

Something else I remember reading somewhere.

The in/out movement of the cw, relative to the centre line of the pinion, has a much greater effect on backlash, than the up/down movement of the pinion relative to the centre line of the crownwheel.

Hence a 'small' variation in pinion height won't matter too much.

Wishful remembering?
anamnesis

Ah, OK. I had misinterpreted that same paragraph that Anam shows from the manual. Thinking that was being addressed to someone repairing or refurbishing a diff, and instructing them to adhere to the engraved figure. But I happily accept the figure was supplied by the machinist for original assembly, rather than measured after assembly. Though thinking about it, the result should be the same!

There is still a range of tolerance quoted, so the etched figure must be a target for the new diff as assembled. Although some wear tolerance might be acceptable in use, I would expect a good specialist firm to also be able to set one up to match the exact etched figure.
GuyW

Tooth patern is everything ,if that's not right it'll be noisy
As far as your diffs went An.--yes the backlash was out of spec but as far as them saying they set it up to the pattern is bull sh#t, they were way out, one more than the other and you proved that with your yellow dye
The backlash should never end up less than the marked spec to get it right but depending on wear it 'could' end up more to get a good tooth pat.

As far as markings on the new sets, they are marked but differently to old.
Instead of the pinion having a 0 or +2 or similar it will have a metric measurement from the centreline of the carrier bearings down to the rear pinion bearing and shim stack surface that the head of the pinion presses against, so easy to set up just poke the bearing in there, measure down to it from the centreline, subtract the quoted size off your measure ment and the result is the required shim thickness
Same goes for the crownwheel measurement it's a measurement to one side of the shim pack and then fine tuning during assembly of the side shims to get preload and tooth pattern ---it's a simple system

On your statement on pinion height not being as important, it's just as important as crownwheel position -or moreso, if the pinion is out by more than a couple of though it'll never be right---and pinion height also effects backlash so any adjustment to pinion height needs backlash correction as well----all adjustments go hand in hand
Crownwheel adjustment shims run about equal to size so moving the crownheel say 2though will result in roughly 2though different backlash but pinion adjustments are roughly 2-1 so 2though drop of the pinion will only yield about 1thou of backlash change but a major change in tooth pattern, so there's a little bit of playing one off against the other to get a perfect pattern
There was a troublesome gearset here in Fords, it was only in one ratio 3.08 and it had a period whine right on highway speed 100-110Kph---Ford just kept sending out replacement gearsets but i'd say around half of them were still noisy----having nothing much to do one day we decided to play about with one that was particularly noisy dead on cruising speed 105 klm/hr the spec was4-7 thou backlash so washed this axle out and dried it out, marked it up and found it had 6 and the tooth pattern was dead central on the tooth and very slightly towards the toe--perfect---bugger
Anyway decided to move the pattern towards the toe just for a try-subtracted 1though off one side and added to the other and rechecked 5though backlash and the mark towards the toe more but still back from it
Out the road and the noise had moved from 105 down to 95-100-----back apart and another 2though this time which put the mark right on the toe but not past it and 5 thou backlash---out the road it was as quiet as at all speeds
I think the point i'm trying to make is that yes the backlash spec is important but the tooth pattern moreso, a good combination of heights ,tooth patern and backlash all within spec is the target

A good tooth patern is the main target if that's out it'll be noisy------as you have found
William Revit

Thanks for all that explanation Willy.

I didn't say pinion height wasn't important, I said a 'small' variation in pinion height won't matter too much, because I read somewhere what you confirmed; cw adjustment affects backlash more than p height adjustment. But I didn't know p height had more effect on tooth pattern.

Anyway, on my diffs, I don't know if the pinion heights *are* wrong, or, if they've since been altered as part of resetting the backlash figures.

I was told the tooth pattern was more important than backlash, and it was a good as could be set on my diffs.

But I know what I got back. Backlash was wrong on both, both whined, one clonked, and one had a cyclic rrrr rrrr rrrr.

Having a near perfect tooth pattern is one thing, but even I ignored the whine, if that results in a clonk on one, and an rrr rrr rrr on the other, then that tells me both pattern and backlash are important.

I'm sure I've read that on worn diffs, tooth pattern has to be moved a bit to compensate. Is it drive side tooth mark more to toe?

When I can, I'll see what the new paint marks show on both diffs. The altered markings are quite visible I think. If not I'll see if I can get some of that ochre stuff and paint the teeth before I refit and test diffs. I'm also going to remeasure the backlash and runouts on both.

I've no reason to think both diffs won't be good now though.

anamnesis

Took a few picrures and made a few measurements of the reset Diffs.

Both Diffs, still have 3 thou runout.

Both Diffs now have backlash that varies due to runout between circa 6 to 7.5/8 thou as best as I can measure.

These pictures are Diff 1





anamnesis

This is Diff 2.





anamnesis

looking ok-

diff #1--maybe could have the pinion a little lower by only 1-2 thou.if going for perfect -just to straighten out that D shape and get full coverage - but it's pretty close and should be ok -the pattern tends to move out towards the flank and heel of the tooth under pressure anyway----If anything you 'might' just hear a tiny whine from it on very light throttle maybe, but good under load.
It'd be interesting to see if the pattern is consistent at a few spots around the crownwheel--if it's got some runout they might have averaged the mark
It should be ok though all being fair

diff#2--should be ok

There's an old saying among diff builders that runs true---it's

Toe and heel move the wheel (crownwheel sideways)
Face and flank move the crank (pinion in-out)

To move the mark towards the toe----move the crownwheel towards the pinion
To move the mark towards the heel ----move the crownwheel away from the pinion
To move the mark out to the flank of the tooth--move the pinion further out from the crownwheel
To move the mark down into the face of the tooth--move the pinion in towards the crownwheel

It's all a big juggling act--
William Revit

Thanks Willy.

More Diff 1 from other places round the wheel.

I see what you mean about P height now, with your description of the 'D' shape.

BOTH diff's have pretty much the same 3+thou runout. I'm willing to bet, but can't prove, they had less, within spec, before both were removed from their cages and bolted back on. Diff 1 had to be for a cage swap. Diff 2 was for cleaning I was told.

I'm wondering if it's worth moving them round the cage to look for less runout. It can't be good to have more than bmc says is the max.







anamnesis

I'd run it for a try as is I think , in that second pic ,the teeth round closest to the camera are showing a good full pattern, I'd give it a go as is--

Turning the crownwheel on the carrier usually does nothing, it's the face of the carrier that has the runout, not the crownwheel-------unless there's a foreigner in between the two, but surely a pro diff rebuilding joint would have cleaned/checked for that
Do you know what they did to fix the loose bearing issue-
William Revit

Yep, I'm goint to start with Diff 1, to see if the rrr rrr rrr has now gone. If it's only been reduced, which I expect because the teeth are now not as deeply meshed, but I can still hear it, that can only be down to runout varying the depth of meshing. So it seems to me, I'd at least have to look at the mating surfaces of the cage and wheel, for as you say, foreign bodies.

The loose bearing was the reason for the cage swap; to fix the loose bearing fit.

The 'new' cage was taken from my spare 4.2 diff, that had seen no use at all in over 35 years, and not much before. The original bearings were on it, and were a good and tight fit.
anamnesis

Got back in the garage today and put one of my Diff's back in.

Diff 1 back in.
Backlash reduced, the whine hasn't gone unfortunately, but is definitely less noisy. It comes in at around the same speed, 45 ish, but it's quieter. Quiet enough to live with. I guess it'll have to do is the best I can say.

But it still has a cyclic rrr rrr rrr. That too is a bit quieter, but I'm not happy to live with. It must be caused by the 3+ thou runout. And apart from being annoying, it must eventually cause uneven tooth and bearing wear. The book gives a maximum runout figure of 2 thou, and says any more must be corrected. To add an additional 1+ thou, to an already quoted maximum figure, and then think it doesn't matter and not try and correct it, doesn't make any kind of sense to me.

Since I have to pull Diff1 again anyway to test the other diff, I'm going to take the crownwheel off Diff1, look at the mating surfaces and have a go at improving the runout myself. I've asked Hardy's to send me a set of the ORIGINAL crownwheel bolts with the lower profile heads, so I can replace the ones they used, that I had to grind the heads down on to clear the case.

Having a close look at the edge of the cage where the crownwheel mates to the cage, I'm wondering if this (see pictures) is the cause of the runout. This is exactly where I measure peak runout at over 3 thou. It looks like there's something between the crownwheel and cage surfaces.








anamnesis

Anam,

I checked the 4.22 on the bench and it has apparently ungraded bolts but they're longer than the W grade from a new 3.9 I posted earlier and go all the way through. Maybe weaker so greater thread engagement. Do you have a torque setting for them as my Haynes manual doesn't mention them.

What did Hardy use for the cross pin locking pin?





David Billington

The bruise on the cage looks like the carrier has been dropped or similar without the crownwheel on it, this may have bulged the surface where the crownwheel meets it--but then,them grooves between the two look like someone has hammered something in there to seperate the pair at sometime
Can you get a feeler gauge in there at all either side of the damaged area
When you tidy it up you can fit the carrier without the c/w and check the runout of the surface, if you get that to 0 you're set to go
IF you're going to dismantle that diff i'd be tempted to put the pinion shim on a diet while you're at it, you won't need any parts to do that--just remove the flange and push the pinion out, leave the front bearing and seal in there---if you're reducing the shim the same collapsible spacer can be used as it will be collapsed further when tightening it all up-
I'd be going .002"- thinner shim to get that mark up into the centre of the tooth
Were they new bearings on the carrier or them that were on it
Not being able to hear the noise is hard but a rrr rrr rr sound is typical of faulty carrier bearings at low crawl along speeds in town but then I guess if it has reduced with the latest adjustment it can't really be that--maybe

I'm a bit surprised but not really of the sh@t work these guys are doing, specially when they've been caught out and had to redo their botch job- you'd think they'd be all out to save face.

willy
William Revit

David.
Which cross pin do you mean? The one that the the small bevel gears spin on? I assume/hope, it still has the good one that was in there before it was stripped for rebuild.

I've only got a 2 thou feeler Willy. I can just jam it in a very very tiny bit either side of those marks. But that may only be due to the chamferred edge of the carrier.

Yes now you say that, they do look like 'holes' where something has been hammered in. Surely that's not how you take these cw's off the cage is it? Bugger, it's a pity I didn't take pictures before. But I really can't believe Hardy's would do that, or even need to do that, with all the correct tools they have to rebuild diffs.

Yes, the slower you go the more noticable the rrr rrr rrr is. Just as you say, slow crawl, and esp' when putting on power from slow. But it can't be the carrier bearings, as they have all been relaced with new bearings. All good branded, either ntn, koyo, or timken.

Pinion shim/height. I did wonder about that. But I don't have a puller to take off the bearing to get to the shim. And doesn't that destroy the bearing again anyway? But if it's not causing the rrr rrr rrr, or any other real issue, is it important?

Willy can you confirm?

Just so I really understand it, because I'm still a bit confused by the action/effect of the pinion height.

ROOT=FLANK=BOTTOM.
FACE=TOP=LAND=CREST.
TOE=INNER of ring/wheel
HEEL=OUTER of ring/wheel.

Lowering the pinion/decreasing the height, increases backlash a bit, and moves the contact paint mark OUT towards the FACE, and a bit towards the HEEL.

Is that right?

Full bloody circle. It's precisely because I wasn't, and I'm still not confident doing the pinion bearings, that I decided to pay for it to be done professionally.

If the pinion height really does need changing, if it's not easy to get at the shim, I don't think I have any choice, it'll have to be done by Hardy's.

anamnesis

All good with the bearings, if they're new you'd hope they're not the cause, it must be the runout-

" ROOT=FLANK=BOTTOM.
FACE=TOP=LAND=CREST.

Lowering the pinion/decreasing the height, increases backlash a bit, and moves the contact paint mark OUT towards the FACE, and a bit towards the HEEL.

Is that right? "

Correct--and my apologies i had face/flank mixed up earlier there--sorry 'bout that. didn't mean to confuse things but there you go, did it without trying-

So basically it works different to what you'd think-
By pulling the pinion out a bit you'd think the mark would follow it further towards the end of the crownwheel tooth --but no---it moves accross the tooth towards the face
Same with the crownwheel, moving it sideways you'd think the mark would follow it accross the tooth but no it moves along the tooth

So looking at the average of your marks they look to be D shaped down towards the root/bottom/guts/flank of the tooth, so pulling the pinion out a tiddle will lift/centre that up out towards the face, -it's touchy so only a couple of thou---so doing this lifts the mark towards the centre but because pulling the pinion out that couple of thou will also increase backlash (but not by as much) probably only one thou, it will also mean the mark will move towards the outer end of the tooth(heel) a tiddle at the same time
BUT--because of your runout issue, i'd go one thing at a time, sort out the runout problem first then mark it up again and see what the result is--then progress on if needed-----some of your tooth markings were looking half ok so you never know ,just trueing up the crownwheel might be the magic ingredient
William Revit

Thanks Willy. I'll have a go at the runout today.

Meanwhile, I took another drive to warm it up to drain the oil, to take out the diff.

Yesterday I was concentrating on the whine and rrr rrr rrr on drive. Today I noticed that when the whine is loudest, circa 45-50 ish, it also 'howls'(for want of a better word) a little on coast and decelaration. But again at faster speeds it's quieter.

Did I read somewhere that the pinion moving, is responsible for noise on decelaration?
anamnesis

This is what I read.

"There are various situations which can create rear differential noise. For example, howling of gears is a sure sign of wearing.

If you notice that the howling noise only occurs during deceleration, then it is a good indicator that the pinion-bearing preload has loosened."

https://fuelandfriction.com/weekend-warrior/understanding-rear-differential-noise/

But since I get also whine on drive, is that more simply that my cw/p is worn?


anamnesis

Anam,

Yes that's the one I was referring to, the pin that retains it, IIRC they're NLA and can fail leading to the cross pin coming loose DAMHIK. I've seen it mentioned using tension pins one inside the other or maybe a spirol pin. Just wondering what Hardy used if they had the diff apart to replace the thrusts.
David Billington

Ah right. Yes they did have the diff apart David. Because the entire cage was swapped fot my spare. In fact it was apart twice.

First time he used the old set of worn gears that were in my spare 4.2 cage. The 2nd time, was because I noticed that and asked him to put back my good gears. I'll look and see what pin he used. From what I can tell, there's no need to replace that small pin, as long as it is taken out correctly. But I have read how norm replaced his with a roll pin, and smaller solid pin through the middle of the roll pin, to expand and strengthen it.

Well, DIFF 1 out AGAIN. Now I'll take it apart and examine it. I'm going to count how many times I've done this. 😶
anamnesis

Anam,

In looking for the torque setting for the CW bolts I came across this thread 'Diff torque figures' in 'MG Midget and Sprite Technical' from 28/7/2016. It had a section of workshop manual as an image which I've attached as it's about CW runout. The torque mentioned is 60lbft which is higher than recommended for a 3/8" UNF grade 8 fastener but W is higher strength again. What grade did Hardy fit with their new bolts?


David Billington

Hi David. Thanks for that. I found the same thing.

Here's what I did today.

I undid the cage cap bolts, which weren't that tight. Book says 65lbft, but they didn't feel anything like that. They undid without clicking my torque wrench at 40lbft, so they must have been less than that.

Having got the cage out, I can see the bearings haven't got any brand markings on them at all. I can't see where it says
they are either ntn, koyo, or timken.

I put the cage into my 4.2 casing (no pinion in it) with the crownwheel still bolted on to check it spins freely. It does. Measured the runout again, same 3+thou.

The book says the crownwheel bolts should be 60lbft. 1 of them was a lot less than that though, less than 30lbft, even with the thread lock. The others were very tight.

The mating surface of the wheel is pretty well spotless. But the cage mating surface has a few raised spots I can feel with my fingers, mostly where those marks in my earlier pictures are. And there was thread lock trapped under wheel in a couple of places. Those things including the looser bolt may be the cause of the runout.

I checked the cage on it's own and it runs pretty true except for those few local high spots. Tommorow (thursday) I'll clean up the threads and mating faces, and bolt the wheel back on and see if the runout's gone. I'll try it in a few places.

Edit.
Also, the plain shoulders of the bolts are long enough to be visible on the other side of the cage. I reckon they may well be bottoming into the threads of the crownwheel. If there'd been a washer they'd be better. I'm wondering how well clamped the crownwheel was.






anamnesis

Some of the 'dig' marks. Hard to show, but I can feel raised sections with my fingers. Looks like someones had a go with a file. No telling how recent they are though.

Couldn't get my dial gauge into place, so I check true this way with a feeler gauge. Apart from the dig mark high spots, it runs even all the way round.

Now I've got the cage out, I might as well check the pinion pre-load measurement. Then when I bolt the cage back in with the wheel on it, I'll be able see how much pre-load the cage bearings add.






anamnesis

I spent the day, SUCCESSFULLY getting rid of the crownwheel runout on DIFF 1.

Here's what I did.
1. Used a stone to flatten of the high spots on the cage mating surface.

2. Cleaned off the hardened theadlock from cage and crownwheel surfaces, where it had spread in a couple of places.

3. Ran a tap through the c/wheel threads to clean them out. And cleaned the same junk out of the bolt holes in the cage.

4.The plain shanks on the bolts used by Hardys are too long, and almost certainly hitting the threads in the c/wheel when it was bolted to torque tightness. If they used washers/lock tabs under the bolt heads, they'd be fine. But also, without any washers or tabs, the bolt heads are cutting into the cage, making the long plain shank issue even worse.

4. I ran a die down the bolts to put extra threads on the shanks. I can't use washers because these bolt heads are too deep, and would hit the inside of the diff casing (See earlier posts and this picture).

5. Starting at the position where Hardy's bolted the cw to the cage, which gave over 3thou runout, I turned the cw 180 degrees, and bolted it up again. Put it in my empty 4.2 diff casing and runout was 2thou. I then carried on turning it clockwise 60 degrees (1 bolt position) at a time. At 240 degrees I got 2thou runout again. At 300 degrees it was back up to 3thou. I skipped the original position and checked it at 60 degrees past original, and got UNDER 2THOU. One more turn to 120 degrees, and BINGO. JUST A BIT UNDER 1.5 THOU RUNOUT.

Which proves it can be done, and should have been. I'll have to do DIFF 2 as well.

Phone call to Hardys tomorrow.

EDIT.
Can someone tell me what size the ORIGINAL crownwheel bolt heads are? Are they 9/16"af or 5/8"af?







anamnesis

Anam
Well done but I can't help thinking there's a moral to all this - if a diff is only *slightly* noisy, leave it alone as disturbing it might make it worse!
Bill Bretherton

Agreed, that would apply in some cases.

But is that the lesson here Bill? Or is there another lesson(s)?

Re-read my first posts. It's why/how I started this thread.

"I've got two 3.9 diffs, ----"

"One 3.9 is pretty good if you're deaf, reasonably smooth, but howls a little like a banshee having a bad temper after a night on the beer. Esp' on the over-run."

"The other 3.9 is pretty quiet, but rumbles, and makes the rear end feel as if it has a bent halfshaft. Rrrrr - rrrrr- rrrr -, which increases frequency as speed increases."

Something had to be done.




anamnesis

Anam,

The bolts in my 4.22 are 5/8" AF, the ones in the NOS 3.9 are 9/16" AF. Running a die down the bolt is not a great idea as the new cut thread will be weaker than the original which will have been rolled. Did Hardy not provide you with a set of old bolts as you asked for? Sounds like a better option to me unless specifically advised not to reuse them. You're encountering too many issues with the new bolts which IMO Hardy should have been aware of.

Regarding the bearing brand marking not being present I'd be suspicious of the ones supplied. I have a SKF 7207 and it's marked SKF and images of NTN and Koyo 7207 show NTN or KOYO. Re Timken that'll be the taper rollers for the pinion. All other bearings I have from major makers in that sort of size are marked.
David Billington

Thanks David. I thought the originals might have been 9/16, as that clears the diff casing.

No not got the originals back from Hardys yet. I've asked twice, and they've been promised. I only ran the die down to clear the thread lock, and these bolts won't be used. I just needed them for adjusting the runout.

That's exactly my thinking. Why wouldn't a brand like koyo stamp/engrave their product? I've got a set of koyo rear wheel hub bearings, and they're stamped.

If the cage bearings aren't branded, I'm wondering if the pinion bearings are unbranded too. The point being, I was told branded bearings would be fitted.

anamnesis

Anam,

I've measured the CW bolts on the 2 units and got the following.

4.22 has a drain plug in the final drive casing, what difference did that make if any apart from being an earlier unit.

4.22 5/8"AF, head height 6.9mm, tab washer 1mm, bolt head top to side ~2mm.

3.9 9/16"AF, head height 5.9mm, tab washer 1mm, bolt head top to side ~3mm.

From looking at bolt dimensions the 9/16"AF match standard 3/8" hex bolt dimensions, the 5/8"AF match 7/16" hex bolt dimensions but with a 3/8" thread.

Where were the bolt head fouling your casing as in both of mine the bolt head top is parallel to the side and just shaving a corner wouldn't help unless the heads were fouling the casing nearer the pinion where I can't see unless I dig out the USB inspection scope, maybe a job for tomorrow.

What the head height on your Hardy CW bolts?
David Billington

Anam

I should have been clearer - yes,I do see your diffs needed attention but I was thinking more generally about a diff that's only slightly noisy where it might be tempting to "improve " it but things such as runout could be made worse. Even so, yours should have been "better" the first time round, you shouldn't have to go through all this trauma after spending money!
Bill Bretherton

Anam how do you know the bearings are unbranded as once fitted you can only see one side and the markings may be on the other face?

Trev

T Mason

just reading back through a bit of this and need to make a point-
The tooth marking is the most important point---------
If it's set up correctly the patern will be correct and the backlash will be within spec or very close.
If the backlash is correct and the tooth patern is out it will be noisy.
If the tooth patern is correct it will be quiet regardless of the backlash.
If the pinion height is out throwing the tooth patern off, even though the backlash is correct, it will be noisy
just sayin'
willy
William Revit

Trev, being angular contacts, the only place to brand bearings on the outer race is the visible side. Why would they hide the brand on the inner face of the inner race? I've got koyo deep groove and they are marjed on the face of the outer race. But, yes, strictly, I can't prove they aren't ntn koyo or timken without removing them. Or, I'll email ntn koyo and timken and ask them how they mark their bearings.

Yep see your point Bill, but with runout of over 50% above the bmc quoted MAXIMUM, runout, hard to see how it could be worse. And 50% above maximum can't be ignored. Just as with the backlash figures being so far out.

Thanks David. Scroll down to my posts and pictures on the 17th feb for the full explanation, and see the picture I also posted yesterday, showing where the bolt heads hit. Here's another picture. Enlarge them. You may not think that looks much. But when turning the cw directly, instead of by the gearing of the pinion, the bolt head jammed tight at that point. It was suggested it won't matter, because in use it will wear a bit of the casing away. I'm not joking or exaggerating.

Also I now understand the email Hardys sent me about the bolts. He said he uses 5/8ths bolts because they are "beefier". I thought at the time that only related to head height. Now I realise it's the head width too.

I can also see now the damage caused to the relatively soft cage surfaces, by the much harder bolt heads, by not using washers or tabs to spread the load and facilitate tightening.







anamnesis

can i ask who told you it won't matter and be alright after after it wears the housing away

There's more than half a chance your bearings are marked/i.d ed on the inside of the inner like this in the pic


William Revit

Thanks Willy. So could be ntn or Timken then. I just spoke to a koyo distributor who told me all koyo are marked on both races.

The chap who did the diff rebuild said, not quoting exactly, it would be ok as it would just wear the casing down a bit. He explained that he didn't feel it when doing the tooth pattern, so it can't be that bad.

I said no way, apart from the noise, that doesn't sound good. As they have just rebuilt them, I aksed if he'd mind, if rather than send/take the diffs back, I grind the edges off the bolt heads. He agreed. I was happy to do this because I could get my Sprite back on the road. This was before I tested either of them and found both still whined, one clonked, and one rrr rrr rrr'd.
anamnesis

Seems they could well be nsk, and they are marked on the inside of the inner race, so not visible. I'm happy with that.

Cage bolts are/should be in the post to me.

Meanwhile, as they say, alls's well that ends well. A very ammicable agreement reached. I've accepted half the money back.

I'll sort out the remaining problems myself.

Sorry, just noticed you posted twice Willy.

Point taken about the pinion height. Also, same diff, Diff 1, there is almost no side bearing preload. The cage drops into, and lifts out of the bearing holders, with pretty well no effort at all from just my single forefingers. Same very easy fitting with the caps; doesn't pinch the bearings. Supposed to be a 2 thou pinch. Whereas on my other diff, DIFF2, it does take some noticable significant effort, meaning that diff does have cage bearing preload.


I'll start with Diff 1.
When the bolts arrive, I'll make some temporary lock tabs (I'll order the proper ones), put it back together, chuck it in my sprite, see how it runs/sounds, and take it from there.

Tea time.







anamnesis

Just for clarification and to improve my diff knowledge: I presume pinion depth is the same as pinion height? As I see it the pinion can only move backwards and forwards in the horizontal plane so I can see it would be called depth but why is it also called "height"?
Bill Bretherton

Hi Bill.

I have to say, it's confusing to me too, because the terminology is seemingly confused; -- but may not be.

I thought height and depth were the same thing. But maybe not.

Watch this video.

https://youtu.be/ms4S4q77CkA

Maybe Willy can confirm if everybody uses the same terms, height and depth, and if everybody means the same thing.


One other thing to note, is the pinion can also be deflected 'sideways', if there is insufficient preload.




anamnesis

I presume height is used as it's common to have the pinion vertical when taking the measurement.

Anam,

Are you beginning to wish you had done the job yourself? I wonder why the guy at Hardy didn't notice the lack of preload, I think I'd want to get that right before putting it to use, hopefully Willy will comment later. I noticed the diff cage has 2 notches either side which look to be provided to allow a 2 leg puller to remove the bearings by their inner race.
David Billington

The answer is probably yes David.

I think they have to be pretty 'thin' legs to get in there though. I think a splitter may be better. I'm wondering if external shims may be the thing to do. But at only 1 thou needed each side, they aren't easy to fit, and I can't get them in the uk. Back to the early part of this thread.

I guess really I do need to get the bearings off without damaging them, so the same ones can go back on again, and I don't need to work out new shim sizes.

anamnesis

Pinion height and depth are the same thing--
It almost always was refered to as height because of the method of setting it up-
It was fit the pinion and measure up to the height required and fit the suitable shim to get that height-
Nowadays the measurement on the pinion is taken from the centreline of the carrier bearings down to the top of the shim stack or pinion head face ,depending on the spec of the manufacturer, so it gets called pinion depth -
But really it's all the same.
Shims
an.--This is where the out side shims are the go-
Your shims won't be 1 thou
Your bearings have already got shims in there. probably 5-10-15 or whatever size---it's just a matter of removing the bearings ( with your 2 legged puller if that's what you've got) They're not over tight normally, then measure up your shims and add the thou or whatever to the measurement and get/cut yourself some outside shims to suit, then there's no need to keep pulling the bearings off to change shims
Same with the pinion if you end up getting that out which i imagine you will eventually--press the bearing on the pinion without shims(make sure it seats properly on the pinion, check the radius is compatible between the pinion head and the bearing) and then fit shims under the bearing shell in the housing instead making it easier for shim changes if you need to without having to remove the bearing off the pinion each time-
Your diff joint mate would have a collet puller for the pinion bearing which makes life easy, but after all the dodgey stuff they've done here maybe they haven't---they should have---but who knows
William Revit

Thanks Willy.

You know what's amazing? I understand every word of your last post. Such has been your education of me. 😅😅.

And I was actually thinking along 'similar' lines, regarding the pinion depth. Leave the shim on the pinion, under the inner race. Take out the outer race, and linish a thou or so off it. But as I might not be able to make it dead flat, your idea sounds way better.

As does your way to use thicker outside shims on the cage bearings. Cheers.

anamnesis

Here is 'cool' way to fit the pinion head bearing. Well actually, hot and cool. 😁

Watch from 23 minutes in.

https://youtu.be/H9yKWKV8poc


And an appropriate lyric might be changed to,

I heat up, and then cool down.
You got me spinning
'Round and 'round

https://youtu.be/Ntiet4Rn3lQ





anamnesis

Getting there--
There are rules to stick to ---
modifying bearings is a no no----it's a trap--you'll go to replace it one day and wonder why it's all up the creek---use shims, that's what they're there for-
Not a big fan of heat guns either-too easy to go overhot and some races look as though they have bronze carriers that are coloured plastic just to trap you---I'm a fan of a pot of boiling water, just drop the bearing in for a good heatup/soak for a couple of mins and plonk him on---and plastic cages don't get melted-
Heat guns are handy for some jobs but you do have to know what the cages are made of though-
William Revit

Ok, boiling water it is then. And shims not grind 😄.

Thanks Willy
anamnesis

Anam,

Regarding thin legs, I expect just the tip would need thinning. I've quite a few puller legs that have had to be customised to fit.

Regarding the use of dummy bearings with a slightly opened up bore to allow easy fitting and removal to determine the shim pack required for correct preload, I think use of a DTI as a comparator to check for width differences between the test bearing and the one to be fitted would be an advisable part of the process due to angular contact bearings not being directly interchangeable, unless face adjusted. The difference between the bearings can then be measured and noted and taken into account in the shim pack.

I don't like the sound of putting a bearing directly in boiling water, I'd bag it. If directly in the water I'd use de-ionised as the tap water in my area is quite hard and would leave deposits.

David Billington

I don't think I'll try the dummy bearings route David. I like Willy's suggestion for external shims. I'll see when I get there.

Boil in a bag. Yep, sounds sensible. I must have one somewhere in the kitchen. But would tap water really be so bad?
anamnesis

Meanwhile, this could be useful.

If I decide to lower the height of my pinion, AND I end up needing a new crush spacer, it might be useful to have a torque wrench capable of reading circa 140lbft.

Mine only goes up to 100lbft.

The answer is an extension.

Here's the formula and a link to the calculator for the lazy.

https://belknaptools.com/extensions-calculator/

Just need to rig up a suitable extension to connect the torque wrench.


anamnesis

As Willy confirms, pinion height and depth are the same, as I thought they must be.

I've set pre-load which I checked with a spring balance at a certain radius - AIUI the pre-load figure is more important than the nominal 140 ft lbs. There's a thread about it.
Bill Bretherton

https://mg-cars.org.uk/cgi-bin/or17?runprog=mgbbs&access=&mode=archiveth&subject=97&subjectar=97&thread=2020012714452916483

Here you go
Bill Bretherton

Yep read that Bill. But it's still nice to know what the torque is on the nut, since the torque on the nut against the spacer, is what stops the nut coming loose. And if using a new spacer, nice to know if the crush point is aporoaching. -- Unless using an old spacer, and threadlock perhaps, or maybe deforming the thread end with a punch, once you've got the desired preload.
anamnesis

Just for fun, I'm going to experiment with a used spacer on an old pinion out of the diff casing. I'll restore the length by hammering it on a pipe or bar, and see how much torque it takes to deform it again. I assume it will be much less than the original 140.
anamnesis

"Don't" go deforming the thread of the pinion, you're moaning about the diff guys being rough and talking like this-same with hammering the sleeve, a bit butchery don't you think.
"If" you are that desperate to reuse the sleeve just pack it with a shim and go again

The whole process of setting the pinion up is done without a sleeve or a seal so that you can easily remove it ,change the shim and poke it back in--several times if needed. Then, when the magic spot is found reassemble with all the hardware in place
NOTE---without the sleeve only do the nut up just tight enough to get the correct drag on the pinion-NOT 140ft/lb (nowhere near it) it won't be much more than finger tight

Also--even with a new spacer the torque on the nut isn't set in concrete, some spacers collapse easier than others, you 'have to' go slowly slowly while checking drag on the pinion while tensioning the nut up--correct drag/preload is the target

If you're worried about the nut coming loose use some wick-in Loctite on it after it's all together.
William Revit

Lol. Don't worry Willy, I did say an old pinion, out of the diff, to experiment on a used spacer, **just for fun**. An experiment, to see the effects. They deform at circa 140. I think it will be interesting to know what torque it would deform at, if it was restored to length.

And yep I said perhaps threadlock if re-using spacer. But also if re-using a spacer, and as a result, not having much torque on the nut, punching the very end of the thread in one location to keep thr nut in place, once the preload is set, seems reasonable to me.

But I agree. I think I prefer to start again with a new Spacer as they are very cheap. Hence my investigating extending the range of my 100lbft torque wrench, so I'd have some idea when approaching 140. Because otherwise I have no idea what 140 feels like, other than tight, and as has been said it's easy to overshoot.

anamnesis

Lol all you like -- after 200+ posts this is going nowhere --If you can't see the point of not damaging the thread on the pinion by bashing it there's a problem -- I don't think i'll live long enough to get to the end -- got other stuff to do, I'm out of here
William Revit

Come on Willy. Dot punching the end of a thread, to lock a nut in place, when the nut will then remain in place for yonks after, is hardly damaging the thread, and nothing really much to do with the rest of this thread; which has been very good imo, and is definitely going somewhere.

Thanks for your answers. They've been very helpful.
anamnesis

Can someone talk to me about pinion preload measurement? Paul or Willy or anybody?

Before doing anything else to my 'rebuilt' diffs, I'm trying to establish exactly how much preload has been set on the NEW pinion bearings. It seems to me, but I need confirmation, the preload has been set far too low.

I need help to confirm I'm not doing something wrong, because I can't believe my findings.

On BOTH diffs, I'm measuring preload torque with ONLY THE PINION in the cases (diff cages removed). I THINK, all I'm getting, is a bit less than 2 lb-ins on both rebuilt diffs.

I understand it should be 11-13 lb-ins.

Questions.

1. If when new, the pinion preload (pinion and seal only) should be 11-13 lb-ins, how much would you expect it to be after about 20 miles use? Would the measured preload figure have reduced by much?


See my pictures for next questions.

2. Where I'm measuring it, the pinion flange has a radius of 1 inch. See picture. If I wrap string around here, and hang a 1.5lb weight off it, am I right that the turning torque is 1 x 1.5 =1.5 lb-ins ?

3. I'm using a 9.5 inch T bar and socket on the flange nut, set in the middle so it's balanced. Each arm is 4.75 inches. I've got a bottle of water that weighs exactly 0.5lb. If I hang it exactly 4 inches along, is the torque 4 x 0.5=2 lb-ins ?

Am I making mistakes?






anamnesis

If the spec for new bearings is 11-13 the normal spec for used is about half so around 6 but your bearings are still new-ish so i'd be going around 8

This is a moving torque not from standstill

Using the technical equipment you have at hand and allowing for a bit of weight for the bar itself i'd poke that bar right out one side and hang 1lb on it at 6"-7" from the centre and it should just sit there but if you help it move it should then keep moving round at that------or if you want to get all techy about it you could weigh one end of the bar and subtract that amount from the pound weight and then do it at 8" but remember it's a moving torque not startoff from standstill---
William Revit

Thanks Willy.

Yep, understood about the measurement being moving torque.

Ok, thanks 8lbin for almost new besrings.

Then I am doing it right, and my conclusions are right. I have LESS than 2 lbin turning torque on both diffs.

Diff1 only takes 1.5lbin to start the flange turning.

Diff2 is a bit tighter, it takes 1.8lb to start the flange turning.

On Diff2, doing as you say with the T bar extended, without ANY extra weight, or ANY prodding from me, EASILY spins the flange. It drops immediately, and very quickly, when I take my hand away.

See the still pictures I captured from my video. Pity I can't post the video so you can see how quick it turns.

Also the pinion nuts are 'loose'. I noted the exact positions, and tested the pinion nut torques on both diffs.

Diff 1, the nut began to to turn/tighten at 80lbft. I tested the pinion turning torque again, and it was only barely above 1.5 lbin.

Diff2, the nut DIDN'T turn/tighten at 100lbft. That's the maximum on my torque wrench, so I don't know how tight it actually is. But since there is less than 2lbin preload, I reckon the nut is not near 140lbft.

It seems to me, that on BOTH diffs, the crush sleeves, have NOT been crushed properly, -- that is assuming, they are new sleeves, and that there actually are new pinion bearings in them. Or the old bearings are still in there, with the old crush sleeves, and those are the existing worn bearing preload figures.

The only way to be sure, is I undo the pinion nuts, and strip them down to have a look.

If they ARE new crush sleeves, AND new bearings, I have nothing to lose, because I can just put it back together, and crush the sleves properly, to get the correct 8-lbin preload torque, for almost new bearings.

What do you think Willy?






anamnesis

Yep--i'd be pulling it apart --just out of interest--as you say nothing to loose. AND while you're in there give the bearings a tap to make sure they've been fully seated in the housing
William Revit

Ok, thanks Willy. I'll do it this morning.

Yep good idea about checking if the outer races are fully seated. Might as well see if the head inner bearing race is too.

Given how much I've found wrong, I won't be surprised what I find now, only very disappointed, and somewhat pissed off.

It occurs to me, that if the sleeveisn't crushed enough, esp' on Diff 1, that could account for the pinion gear being too high. And setting the preload properly, may put that right too, without the need to alter/remove the shim under the pinion, and add shims under the outer race.
anamnesis

Does seem a bit slack.

I have a similar socket & short breaker bar and worked out the weight of the bar on it's own was enough to give the 12 in-lb required. It would hold itself parallel to the floor, but move when gently prodded.
Malcolm

Thanks Malcolm.

Bit slack's an understatement.

But I was doubting myself, that I even understood T=fxd. I just couldn't believe after everything else, the pinions wouldn't be right. But they aren't right either.

NOTHING has been done right. And frankly, I'm staggered.

anamnesis

All the above is an example of why I would rather have a go at doing things myself than pay someone to do it :-)

But you will get there!

Pinion out, check bearings are seated right, check what crush tube they have or haven not used.

Re-install pinion and do up the big nut slowly bit by bit (ignore the torque required) until you achieve the required pre load.
Malcolm

Oh yes. After a lifetime of not finding someone to trust, and hence a very big reason for doing it myself, and the confidence to diy, I bottled it.

I bottled primarily over the pinion bearings, and getting them right 1st time.

And now look, the very thing I tried to avoid, is what I'm having to do myself.

I'm angry, including with myself.

But on the upside, I will now be able to know and say, there isn't a single thing I haven't done to my Sprite myself.

anamnesis

Is this video any help in checking pinion preload?

https://youtu.be/idngRF9F7Tk?t=231

It's from a series that a friend has made about the work he's doing on an Austin A40.

Colin
C Mee

Likewise Anam... well... almost...

I couldn't re-bore the block, grind the crank, balance the bottom end or do the head work like cutting valve seats etc.

Although I have done more basic rebuilds of all the above. Honed bores, polished crank journals after throwing big end bearings, rebuilt heads and lapped valves etc.

And most of it has worked most of the time!

Malc.
Malcolm

That looks very good Colin. Thanks. I guess I could buy a digi soring scale. 🙂.

Yep, echo all of that Malcolm. Don't you wish you had a bloody great workshop and the money to buy all the tools, and to buy anything you need? And of course the time to do it all, and the time to 'skill' up.

But, that's why you sometimes have to trust someone else.

So how about this? Yes, it gets bloody well worse.

A USED, AND HAMMERED CRUSH SLEEVE. I just took this out from DIFF1. Does it look new to you, with all that rust pitting, and those even spaced hammering marks around the outside?

I was talking about experimenting with one. And I might have donr FOR MY OWN AMMUSEMENT FOR MYSELF. BBut would I do it for a paying customer- and not tell them?

No wonder the nut was less than 80lbft and tge preload so low.

No wonder the nut was held in place by thread lock.

At least the bearings look to be new NTN.











anamnesis

Not easy to show, but you can see the hammer marks if you look carefully.

New crush sleeve?

My ar*e.





anamnesis

i'd be putting it together without a spacer first and see what you've got, rather than wasting spacers
On the drag measuring subject ,my preference is a long lever small pull compared to short lever big pull, i think the fine adjustment is easier to feel, but then i just do them by feel(hand) anyway so knock yourself out and do what you want, but for you i'd go the long lever approach.
William Revit

Yep, that's what I'll do Willy. And since I want to check the seating of both outer races too, I'll do it without the oil seals too. They're only a couple of quid.

And the long lever method too; as you say more control, seems more accurate, though I guess with a tolerance of a pound or so I guess it isn't so critical, -- is it?

anamnesis

Don't forget to only lightly do the nut up checking for drag without a spacer ---it'll only be not much more than finger tight, but you'll need to measure what drag you have so you can repeat it when you fit the spacer
the torque/drag isn't over critical but it does need to be close, so if you're actually measuring it you may as well do it as close as you can
have you thought of going solid spacer and shims for the pinion-instead of the collapsible sleeve---
William Revit

Yep yep.

I have thought about the solid spacer, but haven't read much about the procedure.

I like the sound of it because there's no worry then of overshooting on the preload.

Am I right that it's only disadvantage is extra time to do it?

Who sells solid spacers?
anamnesis

Happy to say I spoke to Bill yesterday, and again just now when he called me. No quibbles, a full refund of all I paid, plus an unexpected extra amount I didn't ask for. I couldn't/wouldn't ask for more than that, A credit to Bill/Hardy's for that, and I doubt what happened will happen again. The matter is closed. 👍
anamnesis

That's good news Anam. It is surprising what happened as I believe Hardy's to have a good reputation.
Bill Bretherton

Anam,

Did you get an explanation as to why the rebuild seems to have been done so poorly.
David Billington

I used Hardy’s more than 20 years ago and was more than happy with their work, never had any sounds or problems with the diff.

This whole episode has shaken my confidence in their services, and if/when I need such work doing again I’m not sure where I’d go.
Philip Sellen

A few comments on diffs.
I do not like using tab washers as to enable deformation without cracking they need to be soft and so may be subject to further deformation in use with resulting loss of bolt load. Replacement with hardened washers and cross drilled heads for lock wiring in addition to use of chemical thread lock.
I also replace the crush spacer with a solid item and shims (made using the material removed from making the external shims for the carrier bearings). This enables the pinion shaft bearings to be set up without the oil seal fitted to ensure freely rotating but with no actual play. Simple torque check is for rotation using fingers only after applying the full nut torque but checking for free roation during process. Once sorted the drive flange can be removed and seal fitted before final assembly.
S G KEIL

Sometimes things go wrong, mistakes are made. Hardy's care about their hard earned reputation. I'm very sure it won't happen again.

I've had an apology and acknowledgement of what wasn't done as it should have been. I've had a full refund, and an extra payment I hadn't asked for that they didn't have to make, which says a lot. That's good enough for me.
anamnesis

This isn't meant to be cynical in any way at all, but I wonder if Bill from Hardys follows the BBS .
GuyW

SG, is the use of lock tabs really an issue? I don't think anyone has ever reported bmc crownwheels being loose with the original bolts and locktabs. They're used on the flywheel too, and I've never found my flywheel bolts loose.

On the solid spacer, I asked earlier where to buy them. I'd like to use one, but if I do use one, I'd have to have one marginally shorter than I need, and adjust the length with shims, because I don't have the means to shorten it accurately.

Meanwhile a question about the pinion bearing under the flange. How tight should it be on the pinion shaft?

I realise it has to be a close fit to prevent movement. But how tight?

This is a new made in Japan NTN bearing. So I assume it's accurately sized. But it's very tight on the shaft. Very hard to move it down anymore than shown in my picture, and very hard to get it off again, esp' once it's in the diff casing.

I've got an old Timken bearing on my spare 4.2 pinion, and whilst that is a good close fit, it's not as tight as this new NTN.

The shaft is smooth, no score marks. Should I relieve either the shaft or the bore of the inner race?

Edit. Maybe Guy. I don't know.







anamnesis

That bearing is usually a tap on fit and to get it out again just gentle taps with the end of a hammer handle on the threaded end of the pinion and it should tap through ok
Just drop it in a bucket of hot water for a while or sit it out in the sun(if there is any there) and it'll drop straight on----If you think it's too tight you can polish the pinion shaft a bit to get the fit you need, it'll only be less than half a thou.-takes no time at all- don't overdo it-
Just going from memory, the shaft dia where the sleeve fits is either 1" or 1.25"--pretty sure it's 1"---I usually spin up a spacer but have used bought ones--If you go to your local bike shop and ask for stem spacers for a 1" fork they'll have truckloads---again from memory you need a pair to get the right size/length so you need a 10mm and a 8mm spacer to get 18mm and that leaves enough length for shims between them'
There are several different stem spacers about, carbon fibre-no good for this job, aluminium-also not so good unless you can crack it lucky and find a pair of nice chunky ones or steel-best choice, these are used mostly on BMX bikes and owners throw them and replace them with bling carbon/alloy ones----Most bike shops would have stacks of left over steel ones, so 1 x 10mm and 1 x 8mm steel headtube spacers and shim inbetween them.
Be aware of the fit of the spacer against the bearing's roller cage and make sure there's clearance there, some roller cages poke out past the end surface of the bearing and if the od of the sleeve/spacer is too large it can foul there-----naughty, don't do that

William Revit

Excellent, thanks Willy. Everything I wanted to know.

By chance I'm having lunch today in spitting distance of a bike shop. So I'll take one of my used crush spacers along for good measure, and all your info.

I figured the bearing should 'slide' more easily than mine does. A quick test I find I need over 10lbft of torque on the nut to press the bearing on. Much mote than light taps, and quite hard thumps to get it off. I'm not sure if it hasn't 'brinelled' the outer race, with me thumping it out of the diff.

Oil seal removed.
I'm going to confirm this again, but having loosened the nut, I found 1 lbins of turning torque, just to turn the flange through the oil seal. Which suggests barely much more than that on the pinion bearings for preload.

And interestingly (different numbers but same principle using a crush sleeve), reading my FORD official workshop manual for English diffs, --

-- it says this about pinion preload and the oil seal -- See picture.



anamnesis

I have used Hardys since the 1980's when it was then owned and run by the late Dave Hardy (Killed in a racing incident with his Healey 100/4) and now with Bill. They have built a number of diffs for me over the years when I was racing and more recently a 3.7 for the Frog. Absolutely no problems at all. The current diff is quiet and I would certainly have no hesitation in using them again.
Bob Beaumont

Thanks for your comment Bob, but then why such a significant difference in service as experienced by Anam?

It is disconcerting.
Philip Sellen

Bob's last post is a reasonable one as far as I'm concerned.

Amends are often made. Reputations redeemed.

Philip perhaps you could ring Hardy's for reassurance.

My involvement in this thread ends here in this post.

I'll start a new one to continue my diff's rebuild.



anamnesis

This thread was discussed between 31/01/2023 and 18/03/2023

MG Midget and Sprite Technical index

This thread is from the archive. The Live MG Midget and Sprite Technical BBS is active now.