Porsche 915 Pinion Depth - No Shims Required??

[mikedsilva]

(Realised I posted this in engine building forum by accident so posting here now)

I'm setting up a transmission for someone... (I don't normally work on transmissions) The box is a bit of an unknown quanity.

915/72
Bearings bores are actually nice and tight.

It needed
- replacement pinion and crownwheel (original was damaged)
- new gearsets for 2nd and 4th
- all new dog teeth, syncro bands, sliders
- all new bearings

This is the first trans I had pulled apart with only 1 shim (0.2mm).
The old pinion didnt appear to have an N number.. but the matching crownwheel had 66.63 etched on the side, next to the backlash number.. this may or may not be relevant.

The differential carrier was also pretty worn and the carrier bearing on the cover side, had been spinning on the shaft, so we sourced a replacement diff carrier.

So I'm at the setup point. I'm armed with the porsche manual, the Wiki from P Zimmerman, and also the excellent videos online by ************ (but his stuff seems to just work out perfectly every time!)

First thing I went to check, was pinion depth.
The pinion shaft (and input) was assembled in the housing, and torqued to spec so the big nut was tightenting everything down. Note, I did this without installing any shims.
For measurement, I am using a VW289 tool that I bought from Evan in Norway who also put up a thread here.
The tool was supplied with a gauge block and the tool is scribed with 58.69mm. I assume this means that when zeroed, the dial gauge is reading 58.69 from the centre of the rotating axis to the tip of the pointer on the gauge surface.

Evan had supplied it with two specially made spacers (7.61mm), so that when the dial gauge is zeroed, it is reading 63.30 mm. (58.69 + 7.61)

The tool is inserted, the side cover fitted and torqued down, the tool has no axial or lateral movement.
When rotating to the face of the pinion, it is giving me a reading of 0.2mm. So I believe that means, that with the 63.3 + 0.20 = 63.5mm is where my pinion setting is without a shim.

Now, the N number on my pinion is 0.22. So that means the desired deviation is 63.3+0.22 = 63.52.

So that means, that without any shims, I am 0.02mm away from ideal spec but within +-0.03 tolerance.

Not sure I am using this correctly...
Has anyone ever come across a gearbox that didn't need any shims under the bearing retaining plates?

[kevingross]

I've not had that happen to me in any 915 transmission, nor in any other Porsche transmission.

The 66.63 on the original r&p was the ideal pinion face offset from the diff's rotational axis, as I think you surmised. Getrag/Porsche shifted to providing that value later in the 915 lifetime. I have a '85 Euro 915 box on the bench with this absolute marking rather than the N deviation.

Are you confident that your VW 289 is reading correctly? I'd offer to compare it to my VW 385 set measurements, but you're a bit distant and shipping may be expensive.

[mikedsilva]

Originally Posted by kevingross

I've not had that happen to me in any 915 transmission, nor in any other Porsche transmission.

The 66.63 on the original r&p was the ideal pinion face offset from the diff's rotational axis, as I think you surmised. Getrag/Porsche shifted to providing that value later in the 915 lifetime. I have a '85 Euro 915 box on the bench with this absolute marking rather than the N deviation.

Are you confident that your VW 289 is reading correctly? I'd offer to compare it to my VW 385 set measurements, but you're a bit distant and shipping may be expensive.

hi Kevin
short answer = NO, I am not confident I am using it or reading it correctly!
Going to take it to someone else in coming days for second opinion.

The tool belonged to Evan, and he used it to set up his box and from what I can tell it worked well.
I plan to use some gear compound too before I pull it apart, just to see...

[mikedsilva]

Originally Posted by mikedsilva

When rotating to the face of the pinion, it is giving me a reading of 0.2mm. So I believe that means, that with the 63.3 + 0.20 = 63.5mm is where my pinion setting is without a shim.

Now, the N number on my pinion is 0.22. So that means the desired deviation is 63.3+0.22 = 63.52.

Not sure I am using this correctly...

ok, i'm a bit of a goose. I was reading the gauge in reverse.
Instead of 63.3 + 0.22, I should have read it as 63.3-0.22 = 63.08.

And because I want it to be 63.52, that means I need a shim of approx 0.44.

At this stage I only have the 0.2mm shim.

I used some gear marking compound at both settings. Not entirely sure how to read them but I think it's moving in the right direction after I added the 0.2 shim.

First pics are with no shims installed.




And these next pics are with the 0.2mm shim installed.

[Matt Monson]

Mike,

.20 is the thickest. You'll need 2 of them. The factory never made a .05. 0.40 may be close enough. You'll have to check it after installation.

I recently started making a kit. They are 0.05, 0.10. 0.15, and 0.20, sold as a package. Email me if you need more than you can find there locally.

[mikedsilva]

Originally Posted by Matt Monson

Mike,

.20 is the thickest. You'll need 2 of them. The factory never made a .05. 0.40 may be close enough. You'll have to check it after installation.

I recently started making a kit. They are 0.05, 0.10. 0.15, and 0.20, sold as a package. Email me if you need more than you can find there locally.

hi Matt, thanks for this.
I actually sent you an email around 27th oct... not sure you got it?

Mike

[Matt Monson]

Dunno. I’ll look.

[mikedsilva]

Originally Posted by Matt Monson

Mike,

.20 is the thickest. You'll need 2 of them. The factory never made a .05. 0.40 may be close enough. You'll have to check it after installation.

I recently started making a kit. They are 0.05, 0.10. 0.15, and 0.20, sold as a package. Email me if you need more than you can find there locally.

Here are pics with the 0.4 shim total.. pinion depth is in spec here.
Preload is approx 40inch lbs (fag bearings) but backlash is too high at this point.. Im doing the marking to try to save my results for future reference.

[mikedsilva]

Here's my backlash check. I did it at 6 points.
The number scribed on the side of the CW is 0.15.
When I lined up the pre-determined mark on the pinion and CW where it was originally measured, I got 0.15. But you'd think because this is a used pinion and CW it might have opened up a bit.

[mikedsilva]

My final setup was to add another thin shim to the s2 side of diff to slightly increase preload and backlash.
Pulled the input and pinion shaft out... checked preload and it is around 35-40 inch lbs which is about right for FAG bearings?

Rechecked my pinion depth and I am within 0.01mm of spec.

Checked backlash and it is slight larger.. took 6 readings and the average is 0.158 (spec is 0.15)

The book says you dont need to check the pattern.. but here is mine. Not sure if I am using too much compound. I engage gears, and then use a thick leather glove to brake the drive flange while rotating the input shaft forwards, and backwards.

My patterns dont look anything like the "footballs" that the websites and youtube vids mention.
But according to the numbers, I think it should be right?

fyi - I only apply the compound to approx 5 teeth on the CW.

[PeteKz]

That's bit too much marking compound for me, but it indicates that the pattern is about the center of the drive face of the ring gear. I don't think you can do much better than that, and if you keep messing with it, you might end up worse.

As for those perfect long-oval shaped contact patches in the books, I've not had much luck achieving them, but all my R&P experience is on American iron and British scrap iron. I haven't done a 915 yet. So if Matt has different guidance, listen to him.

Another thing I've found is that if the R&P run quietly, leave the mesh adjustments alone. I've had the experience of setting up a Ford 9" diff so the pattern and lash were within tolerances, but it then made more whining than before I adjusted those clearances. It quieted back down after a few thousand miles, but if yours was quiet before, be conservative about changing it. Mercedes used to set up their diffs at the factory by running them and measuring for minimum noise, then marking the ring and pinion adjustments to .01mm. To set up those diffs required a special jig that allowed the factory measurements to be duplicated, but only a very few shops had access to those. That's sort of the extreme case of German precision, but at least MB thought it was the right way to do things.

[mikedsilva]

hi Pete
yes, a friend here also said the same about too much compound. I might have another go with a little less compound and put up the pics.

I find gearboxes fun, fascinating and frustrating, all at the same time.

As for the pattern I really struggle to understand what is good or not, so get others to help. All part of my learning process.

But as far as the numbers go, I believe it is set up within spec.

Would LOVE to be able to run it in some sort of jig... I guess the car is a type of jig huh?

[mikedsilva]

here's another attempt with a little less compound...
The drive side I think is ok.. the coast side looks a little offside to the heel (is that correct term for outside circumference?) and not sure what this could mean or what change would move it to be more central...

[dannobee]

From the last pic, the pinion is in too deep. Needs a thicker pinion shim.

[mikedsilva]

Originally Posted by dannobee

From the last pic, the pinion is in too deep. Needs a thicker pinion shim.

how much are you thinking?

According to the tool, I am 0.01-0.02mm away from spec.. ie IN by 0.01-0.02... and I don't see how I'm going to make up for that?

It seems that the movement of the contact patch is opposite of pinion movement.. in my first photo, with zero shims, the contact patch on the drive side is right out near the heel.
And as shims are added, pulling the pinion further away from the CW, the contact patch moves toward the toe.

[dannobee]

Start with 0.05. If Matt's making a kit, take him up on the offer. Otherwise you'll need to mix and match. With the earlier cars sometimes you're stuck with the least undesirable option and not the optimal solution.

[burgermeister]

I am setting up an Albins R&P. These differ from the Porsche parts in that they are clearly lapped, and the lapping pattern can be seen on the gears.

Gearset is N20 B0.16

Here is a pattern with N18 B0.16

[burgermeister]

And here is one with N13 B0.16







The N18 pattern seems to overlay the lapping pattern quite nicely. This suggests to me that the lapping dimensions are unaltered - so no effort is made to account for any change that might happen during operation, such as a hot transmission case or deflections caused by operating forces.

The N13 pattern is really not all that different. A slight shift towards the toe for the drive side is about all I can see. Backlash did not change switching to N13, at least not within my measurement error. This suggests that the pattern is not super sensitive to pinion depth. Also not entirely unexpected.

Given the differing coefficients of thermal expansion for aluminum and steel, the pinion distance is likely to increase 0.05 - 0.07mm when the case changes from 65 deg (workshop) to 165+ deg (happy track use).

I did ask Weddle industries if I should bias the pinion depth to account for expected dimensional changes at operating temperature, and they said no, though they are just the distributors in the US from what I can figure out. Can't find a contact at Albins. Given the lack of sensitivity to reasonably small changes in pinion depth, it probably doesn't matter a whole lot. I am still inclined to go with N13 over N18 - curious what other folks have done.

[dannobee]

That's a nice looking pattern, don't change anything. Make sure the bearing preload is good and ship it!