Missing shims on a 915 tranny

[jstein]

During my teardown of the 915 I discovered that the PO eliminated (or forgot) the shims between the shaft bearing plates and the differential housing. My question to the community:

1) Is there ever a case where the shims are NOT necessary?

2) How do I go about determining how many shims should be there (or verifying that no shims are appropriate).

Thanks.

[jstein]

I surveyed the shop manual and become a bit disheartened - perhaps the adjustment is too difficult for a normal human to attempt so the PO just ignored it?

[rob justice]

Is there any size engraved on the rear face of the pinnion?

[jstein]

Here's what the end of the pinion looks like :

[rob justice]

Try the matching face on the diff - there should be a pinnion/diff clearance size engraved - the scribing on the right hand side is fir alignment - they're will be 2 scribed marks on the diff.

[Grady Clay]

A few weeks ago I started the explanation ”915 Repair - What tools needed to change carrier bearings?” but the thread petered out.

Your pinion matching number is 359. You should verify that the same number is etched into the ring gear also.

Post some more images of the appearance of the leading and trailing contact patterns on both the pinion and ring gear.

Your “Deviation” value is 3. That means the proper location for the pinion placement is r = 0.03 mm more than the design dimension (R = 66.30 mm). Adding the two together we get the adjusting dimension (E = 66.33 mm) which is where we want the pinion set.

Using P258 set the gauge to zero on gauge block P258c. The gauge now indicates the gauge dimension 0.000 mm but is the sum of the number on P258 plus P258c. In my case that is 54.015 mm for my P258 plus 12.590 mm for my P258c for a total of 66.695 mm. Subtracting the Design Dimension (R) of 66.300 mm from that, my tool reads 0.305 mm when measuring the Design Dimension and zero Deviation (r = 0). With your existing pinion (r = 0.030 mm) the proper setting would indicate 0.335 mm when the pinion is set correctly.

We want the number of pinion shims to locate the end of the pinion where the P258 gauge indicates +0.335 mm (+ is in the direction away from the differential centerline).

I have never seen a 915 with no shims. I suspect they were simply left out. Always use new shims. Particularly the thin one is very susceptible to getting wrinkled. Never use a wrinkled shim, even if new.

I would measure the pinion position, ring gear bearing pre-load and ring-to-pinion backlash with everything set the way you found it. Those numbers may give us a basis for a difficult decision.

The difficult decision is weather to put the gears back the way you found them (probably incorrect setting) or adjust them to the proper dimensions, pre-load and clearance (possibly no longer valid because they have been run incorrectly set for some time.)

Most would recommend you start with a new ring and pinion and set them to the correct position, pre-load and backlash. An alternative may be to measure the current settings, and, using lay-out blueing, visually inspect the tooth contact pattern. Next reset everything to the proper settings and again inspect the contact pattern. Only then MIGHT you make a better informed decision.

A good used ring & pinion set that had never been mal-adjusted might be a better choice.

I think that so long as you have it set at or close to the design specs and have acceptable contact pattern, the worst case would be ring & pinion noise (whine).

Best,
Grady


EDIT to correct my math or typo. This is why this Forum is Great!

[john walker's workshop]

if it's any consolation, i rebuilt a 915 a while back that had no shims. didn't look like it had ever been apart and it was quiet on reassembly.

[jstein]

Grady,

Thanks for the info - your description is a bit more verbose than the 911 manual so is starting to congeal in my head. Of course, I don't own P258 or P258c, so that measurement will be a bit difficult; I'll see if there is anyone local that has these tools.

In your discussion you state that the adjusting dimension will total to 66.32 mm - shouldn't that be 66.33mm? Just trying to clarify the math.

I will take some more detailed photos of the pinion tonight and try and post - here is one that I took when I first took off the shaft :



This doesn't show much - I was highlighting some burring on the edge of the gear here that I was concerned with - this has since been removed.

Along with some other pieces parts I've already ordered for the rebuild of this tranny I ordered an assortment of shims from our host - with luck I'll have the right combination when/if the proper depth is determined.

[Grady Clay]

Thanks; yes, that was an error on my part. That is what is so great about this Forum. It is self-correcting. When I make a mistake, I hope everyone questions it so errors get corrected. I will speculate that a century from now people will be combing these archives for this kind of first-hand information.


John, I had always suspected that but have never seen one without any shims. Can you remember what the “r” value was? Hehe. The differences are much larger with the casting dimensions than with the gears themselves.


Noah, as you found out this process is VERY time consuming if done correctly. The tools exist in every reasonably size city and with many enthusiasts. It probably isn’t economically feasible to pay a normal shop rate (what? $100/hr) to have a talented mechanic do this many times. Good ‘ol DIY. This is one of those situations (similar to MFI) where a Pelican DIY can take his time and do it better with the help from this Forum than almost any commercial situation. The exceptions are JW and others who keep their brains exercised.

When you are looking at the contact pattern, I think the issues are:
Similar patterns in the leading and trailing contact.
A full width pattern.
An even pattern.
The pattern not go to any edge.
That said; I have never seen a “perfect” pattern even with all new parts. That is why Porsche runs-in the gears looking for the quietest spot within some parameters.

The R&P killers are too little (or no) backlash, the pattern going off the outboard edge (axially or radially), line contact, way too much backlash or loss of pre-load on the differential bearings. Of course all the normal things apply; too much torque or horsepower, too high temperature, inadequate lubricant or lubrication and too weak castings for the job.

The 915 is a well designed transmission. It successfully addressed many of the shortcomings and limitations of the 901/911. It also introduced some new challenges. Many (most?) were addressed during its production life. Our goal here should be to make the best with what we have, recognize the limitations and where we can successfully update components.

Anyone contemplating exceeding the 915’s design limits by a lot should be looking at a 930, 950 (G50) or 964 (G64).

Best,
Grady

[jstein]

An update : I traded some emails with John Walker and verified that the tranny he refers to as one he previously worked on is the one I'm currently working on! It seems likely that it's going to be kosher for me to reassemble things without the shims seeing that the diff seemed fine without them.

However, I'm anal. I would like to at least do a "gear paste" test to let me sleep better at night after this thing is bolted back together - I couldn't find any decent description of this procedure. Does anyone feel like describing that to me?

[einreb]

Originally posted by jstein
An update : I traded some emails with John Walker and verified that the tranny he refers to as one he previously worked on is the one I'm currently working on!

You didnt by any chance ask him if he recalled filling the tranny fluid did you?

Thats a riot. Small world, eh?

[john walker's workshop]

the trans was installed by the customer, so hopefully he filled it properly.

[jstein]

No implications on John's skills were intended - right bernie?

I'm accepting all responsibility for driving this thing with no lube - I just want to get back on the track.

Besides, it's thanksgiving - what are all you dorks doing reading this crap?

[edit]: oh wait -what does that imply about me?

Have a great holiday!

[einreb]

Originally posted by jstein
No implications on John's skills were intended - right bernie?

I'm accepting all responsibility for driving this thing with no lube - I just want to get back on the track.

Besides, it's thanksgiving - what are all you dorks doing reading this crap?

We all know JW is a hack. Are you kidding me? I put on my WWJWD bracelet every time I go out to wrench on the damn car.

'taters are boiling now.

[jstein]

No one feels like writing up a short description of a "gear paste" test?

[KTL]

Awhile back I tried to post a picture of a page taken from Carroll Smith's book, courtesy of RoninLB. But it was too large according to Pelican. I'll email it to you when I get home- pic's on the home PC.

[Grady Clay]

jstein

How about a little more of the history of this 915? Was that pictured pinion run without oil? Details. Details.

Let’s have images of both the pinion and ring gears in both directions.

The decision that has to be made is if you need to replace the R&P or not. Not an easy decision in this case. A very critical decision for long-term use.

Some of the things to consider are:
Is there indication of too much heat?
Is there damage to other gears?
Is there damage to any of the bearings?

I’m now more concerned about the lack of any pinion shims. My recollection is there are normally 2-3 shims in the range of 0.15 to 0.25 mm. Let’s assume they normally total about 0.40 mm. If the range of adjustment for a given Factory R&P setting is in the order of ±0.05 mm, then the pinion setting with no shims is 10+X off. Not good.

That means the pinion could have been an interference fit to the ring. Sounds to me like potential destruction.

My recommendation is, with no other information, to replace the R&P with new or a known virgin Factory set one. Yes, you can set yours with pattern. It has defiantly been “around the block.” I’m not sure I would give it long life.

I will be willing to do an article on contact-setting a 915 R&P if I can do it with images of five or more set-ups. There is a lot of subjective skill here. That won’t happen this week.

Best,
Grady
gradyclay@hotmail.com

[jstein]

Grady,

Here's the quick synopsis.

The car was purchased from member of this board (POPS) by the person whom I bought it from (double-second hand if you will). That person is the one who shipped it to Illinois, and realized about 6 months later that it was not quite the car for him. I bought it from him at that point, but still consider POPS the PO. The person who brought the car to Illinois had a PPI done by JW himself.

[sidebar] This transmission was worked on by John Walker and he recalls it not having any shims. I've corresponded with him regarding this issue and he feels confident in leaving it alone ("shimless"). Basically he stated that it ran quiet so I should leave it alone - (John, correct me if I've mispoken please). [/sidebar]

This being my first 911 it was understood (for me at least) to be a learning experience.

I became very fond of DE track days and took the car to 3 or 4 events before losing a half-axle at gingerman (detailed elsewhere on this board for those with a morbid curiosity).

It was after replacing the axle and throttle linkage parts that it became obvious that something was amiss in the transmission - after a few test drives the shifter seemed to fail to disengage the gears (see posting here).

I then procedeed to drain the (very little amount) of transmission fluid which gave me the first sign that something was very VERY wrong. (see posting here).

So this lead to an engine drop (my first) and the dissasembly of the transmission (also my first).

Removing the input and output shafts provided the evidence that it had been running 'almost' dry for a while - there was rust in the needle bearings and the race surfaces. However, there was still a fair amount of fluid in the diff case - so I think the R&P was never run completely dry. Also, the shafts were not BONE dry, probably due to viscous pickup from the little fluid resting in the bottom of the cases.

I have yet to crack the case on the diff - priorities have changed a bit with the holidays approaching, but I will do that this weekend and take photographs of the ring gear.

Let me know if there are any special details you would like captured by the camera, or if you would like any more details.

Thanks in advance for you help,

[jstein]

Grady and other experts of all that is 915:

Here are some pictures - the pinion ones are ones I posted earlier, put here to keep things together. The ring pictures were taken this morning...

pinion end:


pinion side showing edge - I've since removed the burring:


Some images of the mounting bolts holding the ring gear - note how two of these bolts have backed out quite a bit!






And here are the gear surfaces:






more in the next post...

[jstein]

Some more differential pictures:

Ring markings - they seem to agree with the pinion














and an inner shot of the diff:



Let me know what other shots may help in determining any issues.