V2 Diff rebuild + Gear Ratio Swap

[maupineda]

I also wanted to refresh the fasteners but they are too many and changing them for new was expensive, so I got myself a small ultrasonic cleaner and run through in a bath of rust remover for the bad ones, and vinagre for the only dirty ones.

one note, the rust remover does take the zinc coating off, so I will put a coat of black oxide which should be good enough for some time as I don’t drive the car in the rain.

they came out pretty clean

[bigjae46]

I think it is over kill the replace those every time. I know BMW spec is to replace but my guess is it goes beyond the fastener wearing out. When BMW repairs a vehicle there is absolutely an expectation of appearance. Rusty or marred bolt heads are a no-go. And I think BMW wants to do everything within reason to avoid a repeat of the symptom. Customer satisfaction. Again, just my opinion.

[maupineda]

I think it is over kill the replace those every time. I know BMW spec is to replace but my guess is it goes beyond the fastener wearing out. When BMW repairs a vehicle there is absolutely an expectation of appearance. Rusty or marred bolt heads are a no-go. And I think BMW wants to do everything within reason to avoid a repeat of the symptom. Customer satisfaction. Again, just my opinion.

100% agree, in my case is just to maintain it rust free so in year to come I don't struggle to remove seized bolts. if left as shown above, they are raw, and would rust by just touching them LOL, I will post a pic of how they look with black oxide.

[maupineda]

Small update, or more like a small review on my work.

I completed the install of the diff (thanks Slideways) last weekend, and since then I have driven the car about 300 miles. and I will start with things that are relevant for forum readers that are thinking about this type of job as a DIY.

Below are the specs I setup the diff to.

2.2NM pinion preload (20in-lb)
2.0NM for carrier preload (18in-lb)
0.06mm of backlash

Driving the car up to 75mph is perfect, however, from that speed on a whine starts to be noticeable and the faster the car goes, the higher the pitch and more noticeable. is not terrible, but for a car with stock diff bushings it should be quite.

After reading online, and revisiting the RD videos, seems I made a bad decision by sticking to the lower end of the spec for backlash.

From E36 TIS data, we have that the preload, or frictional torque varies by bearing brand and for the Koyo items the spec goes from 2.4 to 4.8, which seems a lot. RD in their videos aim for 1NM, and when I was setting my diff I know 1NM felt loose and would allow for the bearing to make some noise. so I sticked to a value that is withing the range of the BMW spec as well as other builders. So I have no concern with the preloads I used. However, the backlash is bit less clear on the BMW files.

Per TIS it is:
Backlash - 0.06 to 0.14 (0.0024 to 0.0055) - assumes this is for EVERTHING but:
Vehicles with M60B40 - 0.04 to 0.1mm (V8 cars would have taller gears and thus less sensitive to a tight diff)

It seems that the higher ratio gearsets need more tolerance as they become more sensitive to thermal expansion due to the larger number of teeth and so the backlash is best to be set on the higher end. RD recommends 0.1 - 0.12, which is still within the BMW spec.

The sad part of this is that it means another round of diff off/on.

So sharing this experience here for future DIY'ers as very few share data on their setups and save others from the iterative process this can be.

Lastly, after reviewing my many photos of the gear pattern, seems it was best with the OE pinion shim, although it looked a bit high on the tooth, it also had the best contact area, so I will remove the 0.1mm shim I added to the pinion and roll with the original shim thickness only and aim for 0.1 - 0.12 backlash.

As for the shorter gears... well yes, it is very much noticeable and the car does feel way peppier, all gears remain very usable including 1st gear. In 2nd gear it can flash the traction control light as it wants to spin the tires a bit. On the road the car does feel stronger and no longer needs a downshift going uphill. The car is a bit busier at 75mph and you have now 3.4k rpms v 3k on the stock gearing. 400rpm is not much, but it is something that can annoy others, particularly with loud exhausts, I will say this, a 3.91 would be a mid point as it is 8% shorter so it is 80% of the 4.1 with a bit more flexibility for highway users. I personally don't bother it, and I love the feel of large engine in a small car it gives. and with the CSL airbox, the experience is just awesome! What it is refraining me from full happiness is the whine, but I think it will be fine once I fixed the setup with the correct backlash

Final setup (with 0.1mm shim under the pinion bearing)





This is the contact patch with just the OE shim

[bigjae46]

IIRC, I targeted .0035” for the backlash. Me no so
peak metric.

So you swapped ratios? It is not uncommon to have some whine with shorter gearing. I’d leave it alone unless you’re not confident that your backlash reading is spot on.

[maupineda]

Here is another data point. This is what I measured in my original diff, this has 90k kms or 55k miles. 100% stock and original to the car.





P.S. If you read my first posts you will see the spare diff measured 0.0025, but this was a used unit which had color marks on the side caps and a bad front threaded hole, so who knows if it had been messed with at some point. Still in spec though, so could be the spare unit was on the tight side.

[bigjae46]

I've never gone below .003". I feel like the shorter the gearing, the closer you need to be at .003 to .0035 to get the gear pattern right which I believe is Diffsonline's spec. A 4.10 has a small pinion gear so I would guess that you need to be more precise to get the proper gear pattern.

I have heard of failures with 4.44 gears related to pinion gear damage. Wonder if they were within the wider BMW spec or the tighter alleged Diffsonline spec.

[maupineda]

I've never gone below .003". I feel like the shorter the gearing, the closer you need to be at .003 to .0035 to get the gear pattern right which I believe is Diffsonline's spec. A 4.10 has a small pinion gear so I would guess that you need to be more precise to get the proper gear pattern.

I have heard of failures with 4.44 gears related to pinion gear damage. Wonder if they were within the wider BMW spec or the tighter alleged Diffsonline spec.

Have been looking at this a bit more, the BMW 4.1 has 41 teeth in the crown, which means pinion is 10 for a 4.1 ratio.
The stock gears have 47 teeth crown, and the pinion has 13 (3.615 which rounds to 3.62)

So there are more teeth in the 3.62 OE setup, but relatively speaking, the 4.1 setup has more teeth on the crown in relation to the pinion and due to the higher velocity difference between the two they would generate more heat and potentially higher thermal expansion. I wish I had a way to measure all this. all I can say at this point, is that my diff is just too tight and need to go to the higher end of the spec. I asked Dan from diffsononline and he said the patter is not bad, and racing diffs said they pinion preload was set too tight, so their feedback is not cristal clear.

[bigjae46]

Yeah, shorter hearing has fewer teeth on the pinion and the head is smaller - very noticeable when comparing a 3.15 to a 3.91. I'm sure there is more heat and wear with a shorter diff - the wheels are spinning faster at any given RPM.

Agreed, want to get closer to the middle of the spec.

Here are the specs I use:

Carrier Preload - 9 to 11 in/lbs
Backlash - .003" to .0035"
Pinion Preload - 14 to 16 in/lbs

Here are the BMW specs I got from Bimmerforums

Carrier Preload - 11 to 23 in/lbs
Backlash - .0024" to .0055"
Pinion Preload - 12 to 23 in/lbs