5 speed shims

[fig]

So I've just finished carefully measuring bearing height from the case edge, and pocket depth for each of the recesses in the cover. I used a measuring plate and gasket for the case. I thought I'd check the end-play figures I have against the original shim stacks I removed during disassemble to see how much slop there was from the factory. Problem is the thickness for each stack is greater than the end-play figures I have for each shaft. I definitely wasn't expecting that and I'm now wondering where I've gone wrong. In hindsight I guess I should have measured this after I first cracked the box open but I didn't so now I'm wondering if I've made a measuring error or maybe I didn't get the shafts fully seated in the case when I re-assembled it. Any ideas?

DF

[Ray from Indiana]

Some ideas from a non-expert:
1. Are the original shims cupped and is it possible you didn't compress them sufficiently when you used a micrometer to measure them? Try measuring eh individual shims, add them up and compare to your overal stack measurement.
2. Be sure you measure the pocket depths from the edge of the cover. I found on mine that the bearing receptacles were proud of the edge and had to adjust for that. You can use the shim plate bolted against the cover to check to see if the pocket edges are in the same plane as the cover edge.

3. Have you zeroed your measuring instruments and have you checked the plate for flatness?

In my opinion, if you are using a caliper for these measurements try to borrow a depth mike to measure instead.

[robert]

Did you include the gasket?

[Ray from Indiana]

used a measuring plate and gasket for the case

I think he addressed gasket thickness

[fig]

Thanks for the replies. I did measure the pocket depths as you described Ray, I had the same issue with receptacles being proud or even below the edge. I didn't measure the shims individually but I did compress them really well so I think those measurements are pretty accurate. Yup I used the gasket with the plate and had it torqued down to spec, no gasket when I did the cover. My measurements were done with a Mitutoyo caliper with the depth base attachment connected. I measured and remeasured a bunch of times from six locations around the bearings and pockets then averaged the numbers, it was all pretty repeatable within a few thousandths. Could the original gasket have been thicker than the one I'm using?

[Ray from Indiana]

To eliminate the possibility of not having the bearings not fully seated on the clutch side couldn't you just re-heat the bearing areas and tap again?

Bolting the plate to the cover is only good if the bearing pocket bosses actually sit at or below the plane of the cover edges. But if that really is your situation and the plate is flat I would think your measurement plan was ok. If the shim plate is stiffer than the cover and the bosses are proud of the edges, this could fool you into thinking there is more of a gap to shim than really needed. If the bosses are below the cover edge plane and you measured from the boss down you would be thinking fewer shims are needed.

Are you sure the bearings were pressed fully home on both sides of the output and layshaft?

Let me qualify my responses by saying I've only done one of these BMW 'boxes but I'm just trying to reason this out with you.

[Ray from Indiana]

Fig

You say you are using a caliper with the depth attachment.

I would put the caliper depth attachment against the flat plate and push the slide down to make sure you still read zero and not a positive number. Push back up and make sure you don't get a negative number.

I've seen those attachments go loose and out of whack when re-tightened. One of the reasons I prefer a depth mike for work like this.

[fig]

I'm certain the caliper is zeroed properly, I did just what you described before I started measuring. I'll double check to see if the bearing pockets are proud, I see how that could mess up the measurements. Bearings are on the shafts all the way but I think I may re-heat the bearing pockets to make sure they're fully seated in the case. The thing is, the input shaft is also effected and I didn't change the input side bearing so it should be sitting in the case exactly like it was when I took it out.

[Ray from Indiana]

Can you tell us how much is the difference between the original stack and what your shimming calcs are indicating?

[fig]

The original shims are about 5 thousandths greater according to my measurements. I haven't had time yet to reheat the case and make sure the shafts are all in all the way. I'd like to rule that out. Could it be that the box was set up wrong to begin with and thats why the bearings were notchy when I got it? I don't know the history or miles, got it on ebay.