Differential pinion preload too low / nut too loose?

[cohny]

About 2 months ago I replaced the differential input seal on my E39 M5. After about 500 miles I started to get a bad whine when decelerating / coasting only. I used the mark the nut method to make sure I tightened the pinion nut appropriately but now I'm not sure I did so. From extensively researching this topic it looks like a whine when decelerating is caused by a pinion preload that is too loose. '

I removed the exhaust from the car and disconnected the driveshaft from the differential. There definitely appears to be too much backlash. The input flange can rotate about 5 degrees before the output flanges move at all. This is much more than on my E36 M3 which essentially has the same sort of differential. The problem is I would expect the input flange to be loose too if the nut wasn't tightened enough. Is that assumption correct? The marks are still lined up but I suppose it's possible that the threads are different now, or that the nut went back on differently.

I have everything set up to tighten the nut some more but I am nervous to do so. I realize now there were better ways of ensuring the nut is tightened back correctly but hindsight is 20/20.

Is it possible that the input flange could have no axial or radial play, but that the preload is still too low?

[pbonsalb]

BMW has repair instructions that cover this for later models that should also apply to earlier models. I don't have the full set of instructions, but there is an RA in the instructions attached that you can search for and try to find via google.

Attachment 602795

[shadowpuck]

just thought i'd remind everyone - bmw repair procedures and such are available via bmwtis.com.
yes, it is a pay service - i know, i know - but, you can search for any info bmw has (repair info, torque specs, recalls, TSBs, etc.).

[cohny]

I was able to find TIS here: http://workshop-manuals.com/bmw/

I wasn't able to find anything that helps me in this situation. I realize that I can rebuild the differential if I need to, and I have the specifications, but I am trying to avoid doing so. The question is this: if the differential is howling when decelerating, does that point to a loose pinion? And if so, is it possible that the pinion could be "loose" (preload too low) without the yoke showing play?

[bmwdirtracer]

https://www.ringpinion.com/Technical...n_settings.inc

From the above source:


Technical Help
Four Critical Ring & Pinion Settings Four Critical Ring & Pinion Settings Although proper ring & pinion setup can be accomplished at home, I really recommend having a professional do the job. However, if a person really insists on trying it (and that is really the only way to learn), here is some information that should be helpful.

When setting up a ring & pinion, there are four critical settings that need to be correct in order to get proper setup and good gear life. They are: Pinion depth (how close the pinion is to the ring gear), backlash (how close the ring gear is to the pinion), pinion bearing preload (how much bearing drag is on the pinion bearings), and carrier bearing preload (how much bearing drag is on the carrier bearings).

Given enough time, just about any approach will work when setting up a ring & pinion. However, I suggest adjusting the four settings in the following order so that the job goes faster with less frustration.

First, set the backlash using only light carrier bearing preload. The proper preload can be added closer to final assembly. Next, check the pinion depth by way of the contact pattern. Make sure you use only real gear marking compound (available from GM as part # 1052351), as other marking substances are difficult to read. There are many theories about how to read the contact pattern, but I won't open that can of worms now. After checking the pattern, it is usually necessary to adjust the pinion depth. After changing the pinion depth, the pattern should be rechecked. The backlash may have to be readjusted as the pinion depth is changed. It should be close to specifications in order to get a good pattern reading. Once the correct pinion depth is established, the pinion bearing preload can be adjusted.

When setting the pinion bearing preload on a crush sleeve design differential, it is always necessary to use a new crush sleeve every time the pinion nut is removed, or if the sleeve is over-crushed.

When installing the carrier, the same adjusters or shims that are used to adjust the backlash are also used to set the carrier bearing preload. Rather than trying to change the backlash and preload at the same time, I recommend getting the backlash correct and then adding preload for the final assembly. If the carrier has to be moved to set the backlash, it is easier to remove if the preload is not too tight.

Once the backlash has been set, the carrier bearing preload can be added. Some books give exact specifications of how much preload to add to the carrier bearings. I find that the given specs are not always easy to measure or achieve. If they call for 0.015" preload, a zero point must first be established, and that is not easy either. I recommend setting the carrier bearing preload as tight as you can get it without damaging the shims during installation. For screw adjuster type differentials, I recommend about 100 to 150 ft lbs. of torque on the adjusters. From my experience, I have found it difficult to over preload the carrier bearings. I have only seen it done in a few cases, and those involved a mechanic with Schwarzenegger-like arms. Many people claim that setting the backlash too tight is good for high impact applications so that the backlash will not open up too wide. I have been guilty of this claim in the past; however, experience has shown me that it is lack of carrier bearing preload that causes the backlash to open up and not some magical transformation due to high shock loads.

Now that these four settings are correct, the installation is almost complete. Remember to use good high quality oil and follow the manufacturer's break-in procedure


------

Now, I can promise you that I've had a few pro techs tell me that you can cheat this, one way or another.
But the best pros I really trusted said that, "Dammit, Chris, you just don't undo that nut unless you install a new crush sleeve, because you can't maintain the same preload AND the same pinion depth if you try to cheat."

And I believed them

Oh, the two times I SAW a tech "cheat" the pinion....the diff failed in less than 6 months, ring and pinion trashed.

Just my 2 cents.

I won't answer your identical thread at M5Board. I do not deal well with a certain moron in the E39 M5 forum, so remain attached to the E34 M5 forum (my previous M5).

[shadowpuck]

yes, there are various versions of TIS running around all over the place - when I need to know for sure, i go to the source. the cost is low and the info i've needed has always been there.
anyway - i've seen a few folks try this with the pinion and they often had trouble a few months down the road....

[pbonsalb]

It's not a hack when you have BMW instructing their techs that they can remove and install a diff flange without replacing the crush sleeve.

However, I agree that if you try something and end of with worse results, you should seek professional help.

Could it be a case of mistaken symptoms, like you are hearing noise now because you installed stiffer diff mounts or left the reat seat backs down removed the rear seat bottom?

[cohny]

Now, I can promise you that I've had a few pro techs tell me that you can cheat this, one way or another.
But the best pros I really trusted said that, "Dammit, Chris, you just don't undo that nut unless you install a new crush sleeve, because you can't maintain the same preload AND the same pinion depth if you try to cheat."

And I believed them

Oh, the two times I SAW a tech "cheat" the pinion....the diff failed in less than 6 months, ring and pinion trashed.

Just my 2 cents.

I won't answer your identical thread at M5Board. I do not deal well with a certain moron in the E39 M5 forum, so remain attached to the E34 M5 forum (my previous M5).

I don't mean to spam lots of forums, it's just hard to get help with things like this. Plenty of people will help you with exhausts or making awesome wheels fit, but this stuff is much less sexy.

So!, last night I got under the car and tightened the nut some more. With some force the nut essentially "jumped" another 5 degrees or so and then hit some kind of true stop. I don't think I actually tightened the nut enough to get to the crush washer. The nut is now about 3 degrees past the mark I made before I replaced the seal. I believe this makes sense given the wear the nut and threads experienced. After hitting this "stop" so to speak the nut essentially became unmovable with an 18" breaker.

BUT, now that I have a truer understanding of how this works, I recognize that while the depth is probably fine given the fact that I haven't crushed the sleeve any farther, I basically can't really set the preload properly. There is simply no truly accurate way to ensure that I put the right amount of additional force on the bearings without a rotating torque specification for the entire, fully assembled differential (kind of impossible to define really).

That being said, TIS calls for the mark the nut method, and I think it's pretty obvious that I didn't tighten the nut enough the first time. I am going to bolt everything back together and see what happens. If it's clear this hasn't worked I'll pull the differential and rebuild it. It's just odd to me that your experience with techs who have used this method in the past have seen failures when that is what's prescribed in TIS.

[bmwdirtracer]

It's not spamming, my friend, you're just looking for good information.

Can you show me where TIS says to "mark the nut" ? Because I've never seen such a factory direction, and I'd like to examine the logic.....

Please let us know whether your noise is gone with the new setting.

Chris

[cohny]

http://workshop-manuals.com/bmw/5_se...ive/page_1838/

That is the page in TIS for replacing the input seal. I believe the logic is that if you tighten the nut back to the exact same position as it was tightened to when the differential was originally set up, then the preload and all of the settings should still be intact. This, however, doesn't take into consideration what the immense forces do to the threads. I am hoping that is what's happened to me, and that I needed to go just slightly past the mark to get the right value, but I'm not exactly optimistic. I will definitely post back with the results.

[bmwdirtracer]

Thank you, cohny, for providing that link. I rely on TIS as a most reliable authority, so I'm not prepared to argue with the directions you provided. I understand the logic in the TIS document; however, I do have to wonder if preload is correctly achieved.

That said, you saw the Ring/Pinion link I gave, and all other authorities I've seen agree with that.

Therefore, I'm going to ask you to add your experience to the database, as you achieve a satisfactory conclusion. Hopefully, other experienced people will add their input, too.

Please keep us advised.

Chris

[walczyk]

How did you do extensive research and NOT replace the crush sleeve? I've heard about this quite a few times.

[pbonsalb]

Diffsonline swapped a 6 bolt input flange for a 4 bolt on a 188mm E36M3 LSD without touching the crush sleeve. I ran it with 400+ rwhp without issue before selling it to someone who tracked it without issue. Done right, the BMW service method works. Done wrong or done on a diff that already has issues, it does not.

[cohny]

I did extensive research on how certain failure modes present themselves. There was a lot of information on M5board regarding differential that make noise when you get on the throttle, but mine only makes noise when I come off of it. That seemed significant, and I eventually came to the conclusion that the pinion must be getting pulled back into the ring gear when I let off the throttle which would suggest that the nut was too loose.

[bmwdirtracer]

With pbonsalb supporting the technique, on top of the TIS document, I'm going to have to believe that this is a valid way of keeping the preload. Wish I knew why there's so strong a disagreement, with other eminent authorities on the subject, and why the very high percentage of failures by very careful pros?

I'm going to try to keep an open mind on the subject, and I'll look forward to hearing your report, after you've readjusted and reassembled.

[pbonsalb]

There are reports of DIYers swapping 6 bolt for 4 bolt flanges or vice versa on 188mm E36 LSDs and having issues. It may be that the 4 bolt and 6 bolt flanges differ so it is hard to recreate the original preload by tightening to the same place. The flanges could vary in thickness slightly but enough to matter. An experienced person might be able to do it by feel, though, like Dan at diffsonline who did the flange swap on the 3.38 I ran years ago. More challenging for the DIYer.

If I was doing it, I would go a hair on the tighter side rather than exactly as it was, if reinstalling the original flange.

[cohny]

Back with an update:

Unfortunately I've had no luck. If anything the noise is worse, now presenting itself off throttle but also when slightly on the throttle. I am pretty demoralized. It took me a good amount of hours to get to the nut to try tightening it. I am starting to question what the noise even is. I have only been DIYing for less than a year, and while I'm very proud of what I've managed to accomplish so far, I think I am going to bring the car to a professional to make sure it's actually the differential. All of a sudden the noise seems to be coming from the center of the car and I feel like I'm being gaslighted. It could be the CSB or something.

If it is the diff, I need to decide whether it's even worth having this one rebuilt. I guess in that case the course of action would be to remove the cover and inspect the ring and pinon gears. If they look ok I'll order the bearings and see if I can find someone who can rebuild it for me. Otherwise I'll have to try and find a used diff, which would be great because I'm sure E39 M diffs are readily available on the used market ha ha.

I know some of you were hoping for a verdict on the mark the nut method, but I'm not sure this is a good case study. I did drive the car for about 1000 miles after the noise first presented itself which could have damaged the differential enough that tightening the nut wouldn't have mattered. Or it could be something else, I don't know. Anyways, I will definitely keep this thread updated with what happens.

[shadowpuck]

I don't blame you for wanting to get it looked at. I've been doing DIY for many years now and I still routinely take cars to the pros - I think its a good idea.

is your location NE - Nebraska, or NorthEast?

If anywhere near Lincoln, Nebraska - I can recommend a superb shop.

[bmwdirtracer]

If you decide that you need a used differential, you should send a PM to "clemster" (Adam) at M5 Board E39 forum. His email is adambaj AT aol.com He is well known for selling good E39 M5 parts, and will likely have a couple of diffs in hand. (I've bought several items from him, myself.)

[cohny]

I don't blame you for wanting to get it looked at. I've been doing DIY for many years now and I still routinely take cars to the pros - I think its a good idea.

is your location NE - Nebraska, or NorthEast?

If anywhere near Lincoln, Nebraska - I can recommend a superb shop.

Unfortunately I am located in NE.

I took the car for a drive with a professional and he told me that it definitely sounds like the noise is coming from the differential, but that he didn't think it sounded bad enough that he would expect the R&P to be damaged. He agreed that it seems like somehow the preload is wrong. He told me it sounds more like its too tight than too loose, but either way I am now going to move forward on some kind of rebuild.

First things first, I am going to remove the differential from the car. Next I am going to drain it, pull the cover, and inspect the R&P just to make sure they're ok. If for some reason they aren't hopefully Adam has one available. If they are though then I am going to try and rebuild it.

Here's where the next questions come in:

Is there any reason to believe the shims could not be reused? In other words, do I need to be concerned with only the pinion preload, or the pinion depth, backlash, and carrier preload as well?
Is there actually any reason to believe the carrier bearings need to be replaced?
Do I really just need to remove the carrier, replace the pinion bearings and reset the preload with a new crush sleeve?

[Wojcieech]

For screw adjuster type differentials, I recommend about 100 to 150 ft lbs. of torque on the adjusters.

150 ft lbs = 203Nm. I think you meant 10 to 15 ft lbs so about 2Nm.

[bluptgm3]

Well, will the stack-up in the pinion shaft with new bearings change? There is some variability in the bearing assemblies, not much but the pinion depth could fall out of its documented depth. The pinion should be engraved on the aft end with the depth range. That will have to be checked even re-using the original shims - a measured depth from the center of the carrier bearing journals or 1/2 the journal diameter + the distance from the end of the pinion to the journal bore forward edge. (Either done with no shim or the existing set), adjust shim stack as required. Then with a new crush sleeve, torque the nut (toward some documented value,) checking the torque required to rotate the pinion as the sleeve is crushing. Re-check the pinion depth, adjust the shim stack as required. Repeat pinion preload w/new crush sleeve as required. Then set ring back-lash and then preload. Check torque required to spin the assembly. Adjust preload as required. Hopefully, some BMW centric differential shop will sell the required shim pack, if needed.


Sent from my iPhone using Tapatalk I

[bmwdirtracer]

I have since used the TIS method of marking the nut to change the pinion bearing, and it works just fine.