Big LSD Shootout... OS Giken, Drexler, Titan, GKN, ZF...

[driftflo]

sorry for the delay, i was on holidays for a few days...
for inquiries please contact me via email: service@limitedslip.de

[driftflo]

Originally Posted by DavidZ

I recently sealed up my diff for the first time and even though I laid out a relatively thin bead of gasket maker (Permatex Ultra Black), by the amount of goop that squeezed out on the outside after sealing, I'm sure there's even more on the inside. I assume that even if it breaks off on the inside, it'll be harmless to the gears/LSD?

the seal on the inside will stick together but even if it would break off it will not do any harm to anything inside the pumpkin. at least if you use a reasonable amount of sealant.

[buddalun]

driftflo / Florian makes excellent diffs for our cars based on the type of use/driving style your car will see.

I recently had Florian build me a custom Drexler LSD and it has transformed the car. With DSC off, the car is infinitely more adjustable with throttle input mid-corner. Wish I would have done this sooner as the car is much more engaging to drive in a spirited manner.

[nycplumber]

Originally Posted by buddalun

driftflo / Florian makes excellent diffs for our cars based on the type of use/driving style your car will see.

I recently had Florian build me a custom Drexler LSD and it has transformed the car. With DSC off, the car is infinitely more adjustable with throttle input mid-corner. Wish I would have done this sooner as the car is much more engaging to drive in a spirited manner.

6mt or DCT? Stock ratio?

[Redd]

I have a OSG Spec-S about to go in, together with 3.45 FD. Will rebuild the rear diff at the same time with new bearings and seals.

Anything else I should do while I'm back there? Any adjustments I should make to the OSG?

[buddalun]

Originally Posted by nycplumber

6mt or DCT? Stock ratio?

DCT & Stock 3.15 ratio.

It would have been nice to shorten the ratio, but I did not want to deal with the headache.

[driftflo]

Originally Posted by Redd

I have a OSG Spec-S about to go in, together with 3.45 FD. Will rebuild the rear diff at the same time with new bearings and seals.

Anything else I should do while I'm back there? Any adjustments I should make to the OSG?

please see post 1 and post 30.

[bigjae1976]

Question...

Do I need to mess with backlash and gear mesh patter if I just swap the LSD unit and use the same ring gear?

[driftflo]

yes, you should definitely check backlash and gear mesh pattern!
quite often it is ok, but not every time!
a check is done quickly and if it is not ok it is better to know it before you installed it in the car and find it out the hard way.

[roastbeef]

driftflo do you ever use titanium hardware to reduce weight?

[driftflo]

unfortunately not!
potential customers do mostly complain the already high prices for the drexler "light" lsd cores. using even more expensive materials would not make sense for bmw applications i guess. but i have new lightweight pressure rings that save ~0,5kg... so i am still searching for cost effective performance and weight improvements!

[driftflo]

here are the lighter pressure rings used in an M4 LSD
i do also have them in stock for m3 e92 applications.

[roastbeef]

later on when i have you build mine, we'll try to do some lightweight stuff for the fun of it.

[bigjae1976]

Originally Posted by driftflo

yes! but it is definitely worth the extra effort!

How do I determine how much to machine off the top plate? Or can I send you my top plate to do the machining?

[driftflo]

Originally Posted by bigjae1976

How do I determine how much to machine off the top plate? Or can I send you my top plate to do the machining?

i do measure every rebuild kit before i send it out. so i will provide the exact thickness that has to be machined.

of course i can do the machining for you if you send the topcap!

[kyoo]

does anyone have any links on the oem rear differential, and how it works/is electronically controlled, with the various levels of traction control on/off?

[CSBM5]

Originally Posted by kyoo

does anyone have any links on the oem rear differential, and how it works/is electronically controlled, with the various levels of traction control on/off?

It's not electronic but totally passive. It's a GKN Visco-lok LSD unit. Requires speed differential between rear wheels to (quickly) begin locking action.

[dogbone]

Originally Posted by kyoo

does anyone have any links on the oem rear differential, and how it works/is electronically controlled, with the various levels of traction control on/off?

I did a long three part write-up about the differentials and compared the OEM unit's performance to an OS Giken Superlock. It has tons of AIM Solo DL charts showing how the rear wheels spin in relation to each other at Buttonwillow Raceway. Start at post #173.

https://www.m3post.com/forums/showth...1158445&page=8

It's really really long, but I think it might have something in there for you.

But a couple quick answers to your questions:

-the E9x M3 diff is not electronically controlled. It's a speed sensing unit and requires one wheel to spin up for it to lock.
-traction control does not prevent the rear wheels from spinning at different speed. It still allows that to happen. Otherwise, there's no way for the unit to lock. Most of traction control is done by cutting engine power, and individually engaging the front brakes as needed.

I also did a write-up showing traction control behavior. Many people think traction control uses the rear brakes mostly. This is incorrect. The AIM Solo DL can show how all four calipers activate individually. I posted some fairly quick laps from several tracks showing traction control engaging the front brakes almost 100% of the time. You can check that out here:

https://www.m3post.com/forums/showpo...&postcount=747

[kyoo]

Originally Posted by CSBM5

It's not electronic but totally passive. It's a GKN Visco-lok LSD unit. Requires speed differential between rear wheels to (quickly) begin locking action.

Originally Posted by dogbone

I did a long three part write-up about the differentials and compared the OEM unit's performance to an OS Giken Superlock. It has tons of AIM Solo DL charts showing how the rear wheels spin in relation to each other at Buttonwillow Raceway. Start at post #173.

https://www.m3post.com/forums/showth...1158445&page=8

It's really really long, but I think it might have something in there for you.

But a couple quick answers to your questions:

-the E9x M3 diff is not electronically controlled. It's a speed sensing unit and requires one wheel to spin up for it to lock.
-traction control does not prevent the rear wheels from spinning at different speed. It still allows that to happen. Otherwise, there's no way for the unit to lock. Most of traction control is done by cutting engine power, and individually engaging the front brakes as needed.

I also did a write-up showing traction control behavior. Many people think traction control uses the rear brakes mostly. This is incorrect. The AIM Solo DL can show how all four calipers activate individually. I posted some fairly quick laps from several tracks showing traction control engaging the front brakes almost 100% of the time. You can check that out here:

https://www.m3post.com/forums/showpo...&postcount=747

thanks for the quick and dirty's. when you say requires one wheel to spin to lock, is it by a percentage of speed differential, or literally as soon as the speed is different? the speeds between the two wheels would be different almost all the time, except in a perfectly flat straight line?

[dogbone]

Originally Posted by kyoo

thanks for the quick and dirty's. when you say requires one wheel to spin to lock, is it by a percentage of speed differential, or literally as soon as the speed is different? the speeds between the two wheels would be different almost all the time, except in a perfectly flat straight line?

The OEM unit does not lock as soon as there's any speed difference. You wouldn't want that. You want a diff to be open when you enter a turn. Otherwise you'll understeer into the turn because the rear wheels are spinning the same rate and want to go straight. There has to be a solid spike in spin rate of one wheel versus the other due to gas pedal input in a corner, then the unit senses it, then the unit applies locking. But it is certainly not instant. The charts from the 3 part writeup, show the OEM diffs behavior.

The OEM diff is a "safe" diff. It locks gradually---and only when there is a significant rate of spin difference. A fast locking diff like an OS Giken Superlock can send a car sideways much more easily. This is why they do not put fast locking mechanical diffs on street cars. Some guy with a lead foot on the streets in a corner will send that car sideways right away and traction control will struggle to intervene in time. In more recent cars with electronic diffs, like the F8x M3/4, you can keep the diff open on the street, and then when the car is churning out high performance, the diff can lock more quickly. Pretty cool.

Spend some time reading that writeup.

[M3MPH1S]

Originally Posted by kyoo

does anyone have any links on the oem rear differential, and how it works/is electronically controlled, with the various levels of traction control on/off?

I know this isn't really what you're asking, but I figured I'd include it as a fun info tidbit for everyone

Here's a thread from the coding subforum that shows how traction control is calculated between different models - specifically post #8 (and refer back to post #1)

https://www.m3post.com/forums/showth...light=euro+mdm

I will summarize shortly here:

There is some sort of equation that isn't known to non-BMW engineers, but what we do know is that the car is measuring coefficient of friction in some way and there are values assigned to each chassis as to when to engage traction control.

For M3_UEBERSTEUERN_MUE_1_2_M:
wert_00: !(E90+(S65B40+US,S65B44+!US+MAN_UMF_GTS),(E92,E93) +S65B40)
wert_01: E90 + S65B40 + US
wert_02: E92 + S65B40 + !US + !MAN_UMF_GTS
wert_03: E92 + S65B40 + US
wert_04: E93 + S65B40 + !US
wert_05: E93 + S65B40 + US
wert_06: E92 + S65B40 + !US + MAN_UMF_GTS
wert_07: E90 + S65B44 + !US + MAN_UMF_GTS

Here are the coefficient of friction that are associated with each value:

wert_00: 2.5
wert_01: 4.09375
wert_02: 2.09375
wert_03: 3.296875
wert_04: 3.09375
wert_05: 4.484375
wert_06: 2.09375
wert_07: 2.09375

It was interesting to me just how much the US E93 got shafted on traction control compared to Euro E92

[dogbone]

Originally Posted by M3MPH1S

I know this isn't really what you're asking, but I figured I'd include it as a fun info tidbit for everyone

Here's a thread from the coding subforum that shows how traction control is calculated between different models - specifically post #8 (and refer back to post #1)

https://www.m3post.com/forums/showth...light=euro+mdm

I will summarize shortly here:

There is some sort of equation that isn't known to non-BMW engineers, but what we do know is that the car is measuring coefficient of friction in some way and there are values assigned to each chassis as to when to engage traction control.

For M3_UEBERSTEUERN_MUE_1_2_M:
wert_00: !(E90+(S65B40+US,S65B44+!US+MAN_UMF_GTS),(E92,E93) +S65B40)
wert_01: E90 + S65B40 + US
wert_02: E92 + S65B40 + !US + !MAN_UMF_GTS
wert_03: E92 + S65B40 + US
wert_04: E93 + S65B40 + !US
wert_05: E93 + S65B40 + US
wert_06: E92 + S65B40 + !US + MAN_UMF_GTS
wert_07: E90 + S65B44 + !US + MAN_UMF_GTS

Here are the coefficient of friction that are associated with each value:

wert_00: 2.5
wert_01: 4.09375
wert_02: 2.09375
wert_03: 3.296875
wert_04: 3.09375
wert_05: 4.484375
wert_06: 2.09375
wert_07: 2.09375

It was interesting to me just how much the US E93 got shafted on traction control compared to Euro E92

(You can change the Wert Number in the coding. It's on the same page as the Euro MDM setting. I switched my E90 to Wert_02 and traction control got less intrusive. As far as we know, the least amount of traction control that you can have is, switching to Euro MDM and changing Wert to Wert_02. I ran at the track with traction control on for 4 years, so I experimented quite a bit with it.)