Need Bearing PreLoad Specs

[dentist90]

I have just swapped my old diff w/ old ring and pinion set into a newer final drive housing. I checked the preload of the old setup before I swapped over using a beam torque wrench, with oil seals removed

The diff spun very freely in old case... it didn't even take 1 lb-in to get it turning. If I grabbed both axle flanges and gave them a good spin the diff would rotate almost 1 full turn before friction brought it to a stop. There was no measurable play in the bearings either as determined by prying crown wheel with two large screwdrivers.
I have now placed the diff assembly into the new case, using the sideplate and bearing races from the old case that match the existing carrier bearings. It feels tighter than the original. It takes 4~5 lb-in to get it to start turning and about 1 lb-in to keep it slowly turning. I don't want to ruin these good bearings by overtightening on preload. Anyone have experience with this, or factory specs for setting initial preload?

ALSO: is 1 (one) beer enough to do this job properly, if it IS a Canadian beer?

[dentist90]

No one has an opinion as to how tight diff carrier bearings should be?
How about an opinion as to whether Hillary or Obama has a chance in 2008? ( Although the answer to preload question would be of more immediate use to me!)

And could someone get back to me on the beer question... I've got 11 more cold ones waiting

[slodave]

You've only had one beer in 6.5 hours? Shame on you!

Can't help on your problem.

[bazar01]

Don't have any opinion as to wheteher Hillary or Obama has a chance,
but I have some info on a 944 transaxle diff preload setting.
It should be close.

Here is the pic from the WSM.

[Gunter]

A search might reveal some info.
If not, I would contact JW on this importan issue and post his reply here for future use.

Canadian beer in an Alu-can is a recipe for heart-burn.

Become a connoisseur with Pilsner in a glass or Grolsch, Becks, DB, Dortmunder Union served in a glass mug.

[rick-l]

subscribe

250 (N cm) = 22.1268645 inch pounds force Is that what the 944 document says?

What does the adjusting device do?

[bazar01]

Originally posted by rick-l
subscribe

250 (N cm) = 22.1268645 inch pounds force Is that what the 944 document says?

What does the adjusting device do?

The torque meter is not an adjusting device.
It's just a measuring device to determine the correct tapered bearing preload.
The preload is set by the S1 and S2 shims behind the bearing cups on each side of the diff case.
Too much preload (too much shim thickness) will lock up the bearings and cause it to overheat and fail.

He is swapping the diff case so he needs to set the S1 and S2 shims to correctly preload the bearings.

However, I think it is still necessary to check the pinion distance with the VW universal ring and pinion gauge.
This gauge is very expensive.

These info might help though it's for the 944 it should be similar.

[dentist90]

He is swapping the diff case so he needs to set the S1 and S2 shims to correctly preload the bearings.

However, I think it is still necessary to check the pinion distance with the VW universal ring and pinion gauge.
This gauge is very expensive.

Correct. I'm hoping to keep the same S1 and S2 shims. I have come up with a simple and hopefully accurate way to determine if the backlash and pinion depth are correct, using measurements from the old case. However I do not have a value for preload. All I can tell now is that it is a bit tighter than original. 22Lb-in is WAY more than I have now. Wow. That sounds pretty tight.

"Canadian beer in an Alu-can is a recipe for heart-burn."
...and 12 of them in aluminum cans is a recipe for a headache! Lagers are my favorite, Stella Artois in particular right now. What were we talking about again...?
Oh Yeah, did search long ago. In RarlyL8's ring and pinion whine thread Grady Clay and Jim Williamson come very close to mentioning it but never got that far.

[Jeff Alton]

Did you get the bearing sorted? I had no luck and forgot to email you back stating that, my apologies.

Cheers

[dentist90]

Yeah, I think so. You're referring to the smaller mainshaft bearing. Spoke to a Jason @ Pelican and he says although it is listed as a pinion shaft bearing it was also used in later 915's. I haven't got it in my hot little hands yet, I think its at the border right now in a Pelican box (more stickers...Yea!!)

[Jeff Alton]

Cool that is how it appears in my parts catalogue as well, and that is where I got lost on the trail.

Cheers

[Walt Fricke]

Setting preload on the 911 is pretty much the same as that given for the 944. Take out the pinion. Install the differential stack with its bearings and your shim pack.

You start with measuring the gap around the cover plate (without rubber O ring gasket) pressed into place but not fastened down. This should be about 0.15mm with a feeler gauge.

Then when you tighten the cover down and pop a flange in you measure the drag with a torque wrench. SKF bearings should be 22-30 INCH lbs. FAGs are 35-57 in. lbs.

I don't have a tool to do this - you want a quarter inch inch ounce beam torque wrench, I think, plus extenders and size converters to the socket you need for the long flange bolt's head to turn it all. I've faked it with the feeler gauge. But a guy ought to measure.

Measuring also is done with no oil seals in the cover plate to drag on the flange you installed.

Anyway, getting things right at this stage gives you the total shim stack thickness. That may or may not be the same as from your previous tranny, which is why you measure.

Then you can go on to check backlash (or set pinion depth, if you haven't done that yet, and then on to backlash) with the R&P installed. I haven't done enough trannies to know how likely it is that your old shim pack split was just right for the new case. I'd doubt it would be spot on, but that's a guess.

You have probably figured out that backlash setting is largely an iterative process involving moving shims from one side to the other so the total shim pack thickness is the same, but the split varies until backlash is where you want it.

Shops have a box of shims of various thicknesses. They vary some in actual thickness even if their nominal thickness is printed to be the same, so you need to measure each one. And it seems you never have just what you want. They are (or were) available, per the manual, in increments of 0.10mm from 2.4 th 3.7mm. Along with a shim at 0.25mm (I wonder if that is a misprint in the manual - 0.025?). The manual says you can adjust to the nearest 0.05mm. If they are new you should deburr them.

I don't know how a guy by himself deals with this - I have a few spacers from previous tranny jobs, but there are two Porsche tranny shops where I live and they've let me borrow their spacer box (full of old spacers pulled from junked trannies).

If you are replacing the bearings, keep the old ones and hone their ID out. It is very tough to pull the bearings off, and the spacers reside under the bearings, so each of the cuts and tries calls for removing both bearings (on the backlash setting task - one will do for preload calculations). But a honed old bearing you can slip on and off, and its dimensions are not apt to vary enough to cause trouble.

The manual says that spacer S1 should be 0.1mm (0.004 in) thinner than one half of the sum of spacers S1 and S2. S2 should be correspondingly thicker. But this has to be an initial setting, to be adjusted as needed.

Walt Fricke

[bazar01]

To back up Mr. W Fricke's statement, here are a few more info from the 944 WSM.

When I did my S2 transaxle's S1 and S2, I did what Walter said, to grind down (machine shop did it) the tapered roller brg cups OD so they can be mounted and dismounted easily from the case brg housing.
It took a couple of tries to set the pre-load and backlash. Just imagine mounting and dismounting an interference fit bearing cup 4 to 5 times.
You might damage the case bearing housing.

[MBEngineering]

HI why don't you stop pissing about with the 944/924 setting as this gearbox is not a 915 box, and as for the FAG/SKF bearings drag, that has been up dated ages ago, just what are you going to do with the drag No's on a after market bearing or the NON made in Germany bearings, that this setting was for, the manual has the setting you require and the info for the correct preload/diff set.



you will also have to set the pinion depth with the clamp plate shims

regards mike

Ho one more thing ,Don't rely entirely on the kindness of strangers over the internet.
quote from; Walt Fricke

[bazar01]

Will that make you happy now? Sorry to piss you off with the 944 gearbox.
You are correct, the ring and pinion construction for the two are entirely different animals.
I guess I'll have to wait till my G50 transaxle needs rebuilding, then I can say, they share the same rebuilding techniques using different bearings and shim part numbers.
Not sure when that's gonna happen though with my weekender.

[MBEngineering]

HI bazar01 no the G50 is different again ( similar setup to 915 ) If you need any info on the G50 give me a shout, or have you the 3.2 CARRERA shop manual??

regards mike

[Gunter]

dentist90:
Did you end up setting the pre-load for the bearings before checking the backlash on ring/pinion?
Meaning: Checking the pre-load with pinion gear removed per Jim Williams:
............ Checking the back lash isn't possible without checking the pre-load, which isn't recommended without removing the pinion. If you have your gear box apart, and think any of the differential parts might need checking, you can check the pre-load for the differential carrier while the pinion gear is out of the way. The pre-load setup seems to be independent of what is done inside the gearbox proper, and needs to be done before the backlash can be measured. Jim Williams

What did you end up doing?
What pre-load figures, and backlash figures, did you come up with?

The pre-load factor would depend a little on the viscosity, and temperatur, of the gear oil also.

[Walt Fricke]

Gunter is right that gear oil could affect preload measurements. But these are taken without gear oil present beyond what may still be sitting there in old bearings, or what the manufacturer uses to ship new bearings. And everything is done at room temperature in the world of shop manuals. Since the pinion has to be removed to measure preload bearing drag, that is a good assumption, I think. So the factory thought these variables weren't important.

You can do an estimated setting of preload using the feeler gauge method with the pinion installed - just put all or most of the spacers over on the side away from the ring gear so there is no chance that it is touching the pinion.

Preload comes from the cover plate bending. If the cover plate takes a set due to continuing side thrusts over many miles, preload can be lost.

But for measuring preload drag this won't work, so doing it with feeler only would be reserved for a screw up - the whole box was assembled before the happless would be mechanic realized preload had been overlooked. Of course, if you aren't changing parts like differentials or cases or case covers, you can just use the old shims and their distribution and figure (or hope) nothing changed. The bearings are really accurately made, it seems, so new bearings shouldn't affect things much.

But dentist 90 has his pinion out, and (unlike me) he has the right tool to measure preload drag and is headed down the right road. He is concerned by the fact that his old tranny had lost all of its preload (or on some rebuild it wasn't set correctly), and the accidental preload of his new tranny case and cover plate is still way below spec with the old shim pack.

But 22 inch pounds is within the correct range.

I'm not as adept with the scanner as MBE and Bazar are or I'd have just copied the pages from my manual. It would have taken me longer than typing what I did. And talking about things helps me remember/sort out how this all works beyond the rote manual following we all start out with. So I get something out of the deal too.

Walt Fricke

(who didn't know, but isn't surprised, that 944 preload is set the same way)

[dentist90]

Mike, Walt , Gunter and everybody else, your input has been very helpful. Particularly the pages from the shop manual... thanks mike. I had gleaned most of this information from other threads and shop manuals. My tranny is completely apart: one thing I have not done is remove the carrier bearings or shims from the diff. Before I got into this the only problem was worn/broken synchros. Then the cracked pinion bearing bore showed up and I ended up getting a new diff case only. The pinion and ring gear were still in good shape and quiet before the engine drop.
I tried to measure the preload before swapping anything around. I removed the axle flange seals and put the original diff case back together. There is no side-side play in the diff itself, but as mentioned earlier there was almost no measurable drag using the 1/4" beam torque wrench. I also took some non-factory baseline measurements to try to establish location of ring gear in relation to the pinion bore. I then painted the ring gear and did a mesh pattern with the pinion and it's original .40mm shims to see how the contact pattern looked before changing anything else. I'll post these tonight... my 6yr old is yelling for me to come play with him right now.

[dentist90]

Anyway, I wanted to see how close the new case is to the old diff case. I had hoped that the preload would be acceptable and then I could go about checking the backlash measurement.
Now to my non-factory measurements!
I need the ring and pinion to mesh just as they did in the old case. Measuring backlash is quite a cumbersome task on this type of diff housing. Backlash is determined by how much the ring gear protrudes into the housing:


I came up with an alternate way of trying to measure this by using the wall of the pinion bore (same spot on both cases)with the pinion race out so I could see the crown gear. I measured the old case prior to dis-assembly and found the measurement to be 1.25mm (.050")

I am not measuring depth here, just using the tool as a thickness guage. The measurement with the diff installed in the new case is very, very close. As the ring gear can only be moved in .10mm increments I am starting to feel more confident that the relation of the ring gear to the pinion gear is very close to the old case. This is the same measurement on the new case:


The images here are represenative. I was much more careful and used both hands to steady myself when doing actual measuring. And I wasn't holding a camera at the time.