Need help speccing a differential

[QQ_23]

Had my car about a month now and always noticed a humming/rubbing noise coming out of the back while making tight turns. Usually at low speed while leaving my parking garage. Did a diff service and this did not help so I've been told to just change the diff.

Looking at DiffsOnline I see options to change the ratios as well as WaveTrac and Quaife upgrades. Will going for a different ratio result in a noticeable torque increase? If so, what ratio should I look for and what would be the drawbacks?

The WaveTrac and Quaife options are $1600 and $1800 over the base price. What are the differences between these and the M variable?

If it helps, I intend to track the car 3-4 times per year. The rest of the time will be spent commuting and going for leisure drives. So I'd like the car to be able to take the occasional abuse while still being reasonably comfortable.

Posting the page I'm looking at for clarity:
https://diffsonline.com/bmw-e9x-m3-differentials.html

Thanks

Edit: This is for a manual E92

[Redd]

Yes running a shorter final drive will result in noticeable torque increase. I have DCT and sent from stock 3.15 to 3.45 for a 10% shorter ratio. The additional torque makes the car so much more frantic and exciting. The trade off is higher cruising rpm which can be tiring.

For $2k you can get an OS Giken Superlock LSD which is a clutch type (superior to torsen/helical) which is silent and never needs to be rebuilt. I have one and absolutely love it for both street and track. It really gives you a lot more confidence to push the car compared to the stock LSD.

[F32Fleet]

Originally Posted by QQ_23

Had my car about a month now and always noticed a humming/rubbing noise coming out of the back while making tight turns. Usually at low speed while leaving my parking garage. Did a diff service and this did not help so I've been told to just change the diff.

Looking at DiffsOnline I see options to change the ratios as well as WaveTrac and Quaife upgrades. Will going for a different ratio result in a noticeable torque increase? If so, what ratio should I look for and what would be the drawbacks?

The WaveTrac and Quaife options are $1600 and $1800 over the base price. What are the differences between these and the M variable?

If it helps, I intend to track the car 3-4 times per year. The rest of the time will be spent commuting and going for leisure drives. So I'd like the car to be able to take the occasional abuse while still being reasonably comfortable.

Posting the page I'm looking at for clarity:
https://diffsonline.com/bmw-e9x-m3-differentials.html

Thanks

Edit: This is for a manual E92

Wavetrac - When one wheel gets airborne you do not lose power to the other wheel like you would with the Quaife.

Wavetrac and Quaife are maintenance free vs the M-LSD

[CTMike]

If you want to read and learn more than you thought possible about diffs… see here:

https://www.m3post.com/forums/showthread.php?t=788992

[Scharbag]

Definitely go with a ZF, OS or Drexler clutch type. Love my ZF over the stock M-Variable unit. Swapped mine out of an E39 M5.

I used to have a Quaife in my VW and it was OK. But the issue of loss of drive when one wheel is off the grounds sucked.

Not familiar with the WaveTrack. But from what I have read, the tried, tested and true clutch style LSDs are the way to go for long term reliability and drivability.

As for gear ratio, it does wake up the car. As said, cruising RPM does go up. I sold my 4.10 when I was in the prairies as cruising RPM on long drives was annoying. Regretted it when I moved to the island and all the roads are twisty. I solved that with a SC

Cheers,

[nycplumber]

+1 for multiple clutch style LSD. I have done a fair bit of research on everything else mentioned and it doesn't compare. I'm sure more experienced people can chime in on that. FWIW I think you could get a better diff from that guy Florian who posts on here than diffsonline.

[Scharbag]

Originally Posted by nycplumber

+1 for multiple clutch style LSD. I have done a fair bit of research on everything else mentioned and it doesn't compare. I'm sure more experienced people can chime in on that. FWIW I think you could get a better diff from that guy Florian who posts on here than diffsonline.

I think the Diffsonline guy makes great Diffs. SYT_Shadow can likely provide more feedback. Only issue I have had with Dan is he has a gruff bedside manner. Quality is top notch but he is not cheap and he will not help you out in any way unless you are a paying customer.

Florian (driftflo) seems like a much more personable enthusiast. Also, from what I gather, he makes great diffs and really likes Drexler units. He is in Europe IIRC.

Either one of those two would be my first choice if I were buying a race diff.

[DRLane]

Originally Posted by Scharbag

I think the Diffsonline guy makes great Diffs. SYT_Shadow can likely provide more feedback. Only issue I have had with Dan is he has a gruff bedside manner. Quality is top notch but he is not cheap and he will not help you out in any way unless you are a paying customer.

Florian (driftflo) seems like a much more personable enthusiast. Also, from what I gather, he makes great diffs and really likes Drexler units. He is in Europe IIRC.

Either one of those two would be my first choice if I were buying a race diff.

Dan is simply from North Shore Boston area. It's not gruff, just a bit of a different interaction if you're not from NE.

Florian is incredibly receptive however! Either or won't steer wrong.

[SYT_Shadow]

I wouldn't change the final drive unless it's a dedicated track car. Unlike other cars, Ms are geared very well from the factory.

Dan@diffsonline is a real professional, but can be challenging to get ahold of and do business with.

Until recently we had diffsonline 'racecarrier', their top dog 10 clutch setup and a 3.62 final drive. It was spectacular, you could do a two wheel burnout leaving a turn and the car would track straight and true. It's now sold as the E90 is going back to a street car.

[DrFerry]

Originally Posted by QQ_23

Had my car about a month now and always noticed a humming/rubbing noise coming out of the back while making tight turns. Usually at low speed while leaving my parking garage. Did a diff service and this did not help so I've been told to just change the diff. Looking at DiffsOnline I see options to change the ratios as well as WaveTrac and Quaife upgrades. Will going for a different ratio result in a noticeable torque increase? If so, what ratio should I look for and what would be the drawbacks? The WaveTrac and Quaife options are $1600 and $1800 over the base price. What are the differences between these and the M variable?
If it helps, I intend to track the car 3-4 times per year. The rest of the time will be spent commuting and going for leisure drives. So I'd like the car to be able to take the occasional abuse while still being reasonably comfortable.

Posting the page I'm looking at for clarity:
https://diffsonline.com/bmw-e9x-m3-differentials.html

Thanks
Edit: This is for a manual E92

I recommend you read and study this thread:
https://www.m3post.com/forums/showthread.php?t=788992
driftflo

[Redd]

Based on my research - and correct me if I'm wrong - the main difference between clutch-type vs others (viscous, torsen, helical) is that the clutch-type LSD is positively locked on acceleration. This gives it much faster reaction to a loss of grip or slip situation. The other types operate as open diffs (no lock) until it senses a torque/speed differential between the two wheels, and then only it locks. Problem with this is the response time is much slower and in that split second you're already in a slide. This is what makes the clutch-type the gold standard of LSDs. And also the fact that it works even with one wheel off the ground.

The stock M-Diff in our cars is a viscous LSD. While there are clutches in the diff, the lock is a viscous coupling that is notoriously slow to react. So when you start to slide, you actually gain quite a bit of angle before the LSD 'wakes up' and sets you straight (Hand of God effect).

The problem most people claim clutch-type LSDs are only for track use is that clutch-types are generally noisy and jerky. They grind in slow turns and can rattle at low speeds. However, for our cars, the OS Giken and Drexler clutch-types are known to be smooth and silent and perfectly streetable. I have a Giken in my daily driver and love it!

Main reason I went with the Giken is their long-life. OS Giken claim they have never needed to rebuild one of their diffs and that appealed to me cos my car is mainly a street car and I don't need a ton of tunability or the utmost performance from my diff.

Hope this helps.

[driftflo]

Originally Posted by Redd

Based on my research - and correct me if I'm wrong - the main difference between clutch-type vs others (viscous, torsen, helical) is that the clutch-type LSD is positively locked on acceleration. This gives it much faster reaction to a loss of grip or slip situation. The other types operate as open diffs (no lock) until it senses a torque/speed differential between the two wheels, and then only it locks. Problem with this is the response time is much slower and in that split second you're already in a slide. This is what makes the clutch-type the gold standard of LSDs. And also the fact that it works even with one wheel off the ground.

torsen does not lock the axle at all. it can just help transfering a bit of torque to the stronger wheel while still maintaining different rotation speeds of the wheels.
viscolok oem lsd can lock the axle but is very slow with a huge delay, correct!

Originally Posted by Redd

The stock M-Diff in our cars is a viscous LSD. While there are clutches in the diff, the lock is a viscous coupling that is notoriously slow to react. So when you start to slide, you actually gain quite a bit of angle before the LSD 'wakes up' and sets you straight (Hand of God effect).

a lsd that is locked will not set you straight! it will keep you sideways! if you want to get the car in line as soon as possible hit the clutch or get a 1-way lsd. but why would one want to get the car straight as quick as possible.
you want to pull forward as fast as possible in a racing situation.

a clutch type will make the cars reaction much more direct and predictable.
so for an unexperienced driver the car may be even more difficult to drive with a proper lsd. but for a (at least slightly) experienced driver the car is much easier to handle as it is much more consistant and predictable.

Originally Posted by Redd

The problem most people claim clutch-type LSDs are only for track use is that clutch-types are generally noisy and jerky. They grind in slow turns and can rattle at low speeds. However, for our cars, the OS Giken and Drexler clutch-types are known to be smooth and silent and perfectly streetable. I have a Giken in my daily driver and love it!

no one who ever drove a drexler lsd (with a setup that makes sense) will complain about NVH. this is only a problem of more basic lsd designs with steel/steel fricition materials (like japanese, chinese and some UK lsds).

Originally Posted by Redd

Main reason I went with the Giken is their long-life. OS Giken claim they have never needed to rebuild one of their diffs and that appealed to me cos my car is mainly a street car and I don't need a ton of tunability or the utmost performance from my diff.

OSG has a massive NVH problem once you set it to a high performance spec that is good for racing (it is steel/steel... this is what steel/steel does). if you leave it in standard spec-s it does not make any NVH but does hardly lock at all. performance is really on the very low end of what is possible with a clutch type lsd.
if you want minimum NVH with still pretty good performance, carbon/steel is what OEMs currently use (for their next gen cars). but this is too expensive for small third party suppliers like osg.
and for racing moly/steel is still the way to go for all diff/gearbox builders in professional motorsports and high end sportscars (drexler, gkn, ricardo, xtrac, hewland, zf, graziano, ...)
osg does not supply any professional motorsport company afaik.

concerning rebuild: clutch type lsds (and even torsens!!!) work with friction. friction generates heat and wear. the question is not IF you need to rebuild but WHEN you need to rebuild. and if a manufacturer uses high quality friction materials and surface treatments a lsd can last +200k km.
you can easily drive a drexler +200k km although it will loose a bit of bite and performance (of course). so does the osg... and any other unit.

[Redd]

Thanks for the detailed explanation driftflo. I always learn something from your posts.

Is there a direct link between the number of clutches in an LSD and how long it can last? Ie. Will an LSD with more clutches usually last longer than one with less?

[driftflo]

Originally Posted by Redd

Is there a direct link between the number of clutches in an LSD and how long it can last? Ie. Will an LSD with more clutches usually last longer than one with less?

in short: yes!

in detail:
lockup is achieved via # of clutchpacks (friction surfaces) and the pressure generated on them via the ramps of the pressure rings.

so for a certain amount of lockup "x" you can use many clutches and steep ramps (few pressure on many surfaces) or less clutches and more shallow ramps (more pressure on fewer surfaces).
and of course the version with less pressure on more surfaces is the one that will last longer (plus - even more important - reacts quicker and smoother).

additionally, durability also massively depends on surface coating, surface treatment, steel quality, oil/additive package used, oil change interval, ...

[Redd]

What's the recommended LSD oil change interval for a mainly street driven car?

[driftflo]

for the drexler friction material i would recommend 2 years/20k km.
for other friction materials you should get that info from the manufacturer.

if osg says there is no need for a rebuild (= no wear) i guess there is no need for an oil change as well ;-)

[Redd]

Lol I'm not silly enough to think it doesn't need an oil change.