I bought a set of PMO rain hats and filters from another Pelican, and am fabbing up adapters so they will mount cleanly to the Triumph ITBs. More on this soon!

Nice! Looking forward to your comments on installation and setup.

I'm excited about it, hoping to really dig in very soon!

The following post is for entertainment purposes only :D

What does an incurable tinkerer do while waiting for actual pinion depth tools to arrive in the mail? Build a tool just for fun, of course!

I'm not going to trust this thing, but it will be fun to compare it to the real tool when I get it... I took a measurement and it says the pinion is too far in (i.e. needs shimming.) We'll see. Was fun to build.

Basically, I had my machine shop lathe a pipe down to 50mm and I welded a threaded piece in the bottom. I pressed a bearing on one end and the thread at the bottom allows me to preload the bearing and hold the pipe more or less centered in the speedo side race. I ran a piece of allthread straight through which contacts the pinion head, which can then be measured. Don't try this at home! SmileWavy

http://forums.pelicanparts.com/uploa...1517261986.JPG
http://forums.pelicanparts.com/uploa...1517261986.JPG
http://forums.pelicanparts.com/uploa...1517261986.JPG
http://forums.pelicanparts.com/uploa...1517261986.JPG

If you machine the pipe to accept bearings on both ends, and preload the bearings with shims, it could be reasonably accurate.

Volkswagen made a similar tool, VW289b, with a spring-loaded probe. The probe could be released and locked with a knob on one end.
It was used in early VW transmissions with ball bearings, so preloading the bearings wasn't necessary.

https://www.thesamba.com/vw/archives...289b_front.jpg

Actually, the pipe is machined on the other end for a bearing too, so I could definitely set it up properly. The hard part comes in not having access to the threaded rod with a wrench as it is a bit hard to turn. Maybe I could get to it through the mainshaft hole in the bellhousing somehow.

That's one of the downsides with this type of tool. Another is the difficulty in measuring your results precisely.

If you can't access the threaded rod from the outside, you can pre-set the tool to the desired pinion dimension and see if it makes contact.
If there's no contact, then the pinion is too far away. Hard contact means it's too close. Remove the tool, adjust the threaded rod 0.05mm accordingly, and try again...

Or you can cut out an area in the center of the pipe and mount a dial indicator that can be viewed from the outside through the pipe.....

Brilliant!

OK! Finally an update.

My VW 289d tool arrived, and after a few rounds with the machine shop I got it properly modified. Worked like a charm! Photos below, but here are the details:

-Both shafts completely locked down, no movement
-Washer "tool" in place to lock the flange to the diff

-With the VW tool zeroed at design dimension "R" of 66.30mm from the face of the pinion head to the center of the diff/ring gear, the tool measures pinion depth deviation ("r" value.) The deviation measured at +0.09mm with the original shims in place. This value according to the factory should be +0.27mm for my R&P (N-value inscribed on the pinion head.) So the pinion was in too far by 0.18mm, maybe not a surprise considering the machining work done on the diff housing.

-The original pinion depth shims were a 0.15 and a 0.20 together for 0.35mm total. I swapped these out for 3 new ones, 0.15, and two x 0.20 for a new total thickness of 0.55mm (actually measured to 0.53mm.) The factory manual says max 0.50mm, but oh well.

-Got everything back together and torqued down, and now the "r" value for deviation was +0.24mm. Factory manual allows for +/- 0.03mm. So I'm just barely in spec.

-Time to check backlash: came in at 0.15mm. My ring gear has 0.16 inscribed on it, and I'm right in the middle of the range of 0.12 and 0.18mm. Too tight for worn parts?

-Then I pulled the shafts out again so I could verify preload. Its on the low end of the range for FAG carrier bearings, but the initial breakaway measures around 40 in-lb, with a steady torque under rotation of 32 in-lb. Considering I added 0.1mm of total shim to the diff (3.05mm on the speedo side and 3.3mm on the ring gear side) I'm not sure how much more I want to monkey with it. Of course I could add more shim on the speedo side to increase preload and backlash. Should I adjust?

-Finally, I used my $35 tube of unobtainium Prussian Blue from Amazon to print the ring and pinion gears. It is really difficult to get a good photo of the results, but hopefully one of the experts can give some feedback on the results.



Photos:
Modified VW 289d tool, machined to accept carrier bearings and shims so it can be set in place just like the diff carrier. Dial indicator gets zeroed out so it is possible to measure deviation. Spacers in between the setting block and the tool are to compensate for the difference between the distance R for VW gearboxes and R for the 915 box. (This is explained a page back in this thread.) The dial indicator can be viewed from the top when the tool is installed.
http://forums.pelicanparts.com/uploa...1518909538.JPG
http://forums.pelicanparts.com/uploa...1518909538.JPG
http://forums.pelicanparts.com/uploa...1518909833.JPG

Just a photo to show how the indicator arm touchs the face of the pinion head to take a measurement, obviously the side cover needs to be installed and torqued down for real measurements:
http://forums.pelicanparts.com/uploa...1518909538.JPG

Contact patterns in prussian blue:
http://forums.pelicanparts.com/uploa...1518910874.JPG
http://forums.pelicanparts.com/uploa...1518910874.JPG
http://forums.pelicanparts.com/uploa...1518910874.JPG
http://forums.pelicanparts.com/uploa...1518910874.JPG
http://forums.pelicanparts.com/uploa...1518910874.JPG
http://forums.pelicanparts.com/uploa...1518910874.JPG

Evan, I think it's very commendable that you're making this effort to be accurate.
Here are a few thoughts and observations...

I would not use so much blue. I would only put a thin, even film on the ring gear, and leave the pinion dry. However, maybe other people do things differently.
Of course, you should not have any blue on the gear teeth when measuring the ring gear backlash, they should be clean and dry.
You have a Gleason ring & pinion, by the way.

Before your case was repaired, the pinion shim surface was level with the gasket surface. Was the gasket surface also machined during the repair?
You can determine any change in the shim surface if the gasket surface was not machined. The machinist should not have altered either of these surface dimensions without informing you.

If your pinion shaft is torqued, and you have sufficient bearing preload on your tool (it doesn't appear to be shimmed), then I'll assume that your measurements are accurate.
The Porsche tool requires shims for it to have sufficient bearing preload. The shims can be rearranged, if necessary, to place the indicator tip at an ideal location on the pinion.
The spacers between your tool and the setting block should be 7.61 mm for a tool dimension of 66.30 mm.
I see that your indicator has turned over 7 times (7+ mm) to measure in this range. It's possible that the indicator is sitting to deep in the tool, or that the indicator tip is too long. I'd try to reduce that travel by 5 to 6 mm if possible.

Given your results, I would add 0.05 mm to the speedo ring shim on the differential, from 3.05 to 3.10 mm, and see how the backlash and bearing drag are affected.
I would measure the backlash at several different ring gear positions.

Thanks for the nice words Jon, and the input. Much appreciated.


Rookie mistake :) First time I've ever used the stuff. I'll try again.


Cool to know. The backlash measurements were done with a clean and dry R&P.


They machined the entire face of the diff housing flat after the WEVO insert was installed and they informed me of this. So I had expected that some shimming might be needed. The shop couldn't say exactly how much material had been removed, they estimated 0.10mm. They are so used to doing pinion measurements that they probably just flatten the housings and sort it later. I've obviously gone to lengths to figure this out in my home garage...


-Pinion shaft torqued to 181 lb-ft
-Shims in place on the tool. A little less shimming then I have on my diff, but I had total 5.9mm of shims, enough to give some resistance when turning the tool, in other words preloaded. Not tons, but seems like plenty to ensure that the tool was centered.
-My dial indicator is actually too short, so I have it as deep in the tool as it wil go. This means that I'm only able to get about 0,50mm of preload on the indicator, but at least I can preload it on the setting block and zero it out. On multiple tests in and out of the housing to change shims, it always landed preceisely back on zero when placed on the block. So I'm confident I'm getting good measurements.
-The spacers were supposed to be 7.61 based on the dimension given on the tool (as you described a page back.) As it happens they are 7.56, and I have accounted for this difference in all my measuremets above. I didn't mention it, as it did my head in to make sure I had it all correct, but the numbers are adjusted accordingly.


I measured backlash at 4 places on the ring gear, at 90 degress from each other.


Here is where my day got interesting.
I changed out the speedo side shims (were 2.8 + 0.25 = 3.05) for a singe 3.1mm shim and reassembled. Preload: breakaway at 50 in-lb, with 40 in-lb constant. Backlash: 0,13mm. CRAP!

So I took the diff out again, and just for fun I changed the ring gear side from a single 3.3mm shim to a 3.0 and a 0.25 = 3.25mm.

Here is where my day went to crap.
I reassembled and checked backlash again - back at 0.15mm. But this is where it got weird, and not good. Earlier, while I had the carrier bearings off on each side, I blasted them with brake cleaner and compressed air. You see, when I rebuilt the LSD, I loaded the internals with gear oil, which has slowly been seeping out and filling my carrier bearings while I've been waiting on tools and doing other things. Obviously, a bearing loaded with gear oil will probably give different preload numbers then a dry one, so I gave them a good degreasing before reassembly. During my last test, the diff started binding up to the point where I couldn't turn it. WTF. So I pulled out the shafts. No change. Pulled out the diff. Bearings spin freely with no preload, but just a slight downward pressure on either of them and they bind up.

Help! Now what? :confused:

Oil the bearings, and try again ;-)

Your backlash result is inconsistent with the shim change on the speedo ring side.
Are you certain that the measurements of all the shims involved on this side (2.8, 0.25, 3.1) are accurate?
Are these all new shims? If the previous 2.8 shim was not new, then check it for damage, distortion or burrs.

Hi Evan, your carrier bearings should be lubricated when you are taking measurements. Gear oil is fine for this.

Getting "logical" changes to measurements can be very frustrating, as for example you make a .05 mm change to preload and expect a particular change to turning torque. This happens even when you are careful in the work you do. I attribute it to variance in the true thickness of the shims: so a 3.00 mm shim might actually measure 2.98 mm with your micrometer. Especially when you are using more than one shim (eg, two 1.50 mm shims to get 3.00 mm), the errors can accumulate. Patience. Good luck!

I have long said that this process is like baking. People think,”baking? You just follow the directions.” It’s not that simple. Doing it right requires trial and error, as well as good “feel” for the process. There is a reason a shop charges you 6-8 hours to set up a ring and pinion. Even when you do it by the book, it is often several tries with a variety of shims before you get it right. I’m going to start showing people who think they can install an lsd without the right tools and just a flat plate this thread for reference.

A little gear oil and everything is buttery again. :D But now I'm confused... Peter Z goes on and on about things being clean and dry on assembly. I get radically different preload numbers lubed vs. dry. With 3.1/3.3 shims installed and dry bearings I got 50 in-lb breakaway and 40 constant. With the bearings lubed this same setup now yields 30 breakaway and 10 constant. I've got a nice dial torque wrench from CDI/Snap-On, and I measure over and over and over until I'm getting consistent readings. I'm really having trouble making sense of this bit. :confused:

Is is breakaway or constant rotation we are after here? I already have 6.4mm of shims installed now, vs. 6.25 from the factory. When pressing on bearings I give them 2T of pressure after the races bottom out on the shims/speedo ring to be sure they are seated.


Most of my shims are new, all with original numbers on them. I've checked and they are correct. Some of them that have been on and off bear some marks on the underside from my puller, but they aren't distorted.

I agree, the results don't make sense. I'm going through the whole thing again.


I should have a medal in patience by now. I think the diff has been in and out 20 times. See above on the oiled bearing/preload measurements, not sure what to think there. Thanks for the tips. :)


I'll take that as a huge compliment Matt, thank you. This is a fascinating process, but man does it take a lot of time and repeat work, and not least a pile of tools. I'm going to get this diff right if it kills me!


.

Evan, the manuals do not say to measure the differential drag with bearings free of oil or lubricant.
The bearings are usually packaged with a thin film of oil for corrosion protection. That's usually sufficient for this test, but gear oil is fine too.

You are measuring the torque required to maintain rotation.

The nominal preload for 915 differential bearings is 0.30 to 0.40 mm.
If you have 0.40 mm, I would say that it's sufficient, regardless of whether the bearing drag is within spec. I would not exceed 0.40 mm by too much.

http://forums.pelicanparts.com/uploa...1519113926.jpg

Evan, you want to measure the "constant" not the "break-away" torque value. It's the constant, at-speed torque that matters.

Jon, my preference is for the measured turning torque versus a static assessment of the preload, for example as shown in the photo you attached. My experience has been that static measurement is a good starting point but not reliable: too hard to be accurate and repeatable.

It's interesting that the photo shows the cover plate secured with only two nuts: they flex so much when tightened that I use six nuts at a minimum when measuring the turning torque. I cannot imagine that it's easy to get a consistent value around the circumference with the two nuts.

Evan, my "clean & dry" comments were intended for shims, including the speedo sensor ring, and the bearing surfaces (usually the inner race) that contact those shims, or needle bearing races, etc. Because I always use new carrier bearings, it didn't occur to me that my comment might be interpreted wrong, and a technician might wash and dry those new bearings instead of simply giving the contact surfaces a quick wipe with lacquer thinner! It's always good to put a couple drops of oil on bearing balls/rollers, and spin them by hand, prior to assembly or making measurements, just be sure to keep shim contact surfaces dry.

Which is exactly why a pro-shop would be doing mine.

GOD Bless, this is quite the undertaking..:eek: