S54 advice and considerations

[gobuffs]

I added a roundy round all aluminum triple pass radiator for $199. Cools better and cheaper than anything you'll find that is specific for BMW and is 1/3 the price.

[VinniE30]

This would be so cool:
All aluminum M42 style radiator with built in reservoir

[e30m3s54turbo]

VinniE30 m42 should work.

[VinniE30]

Would a M42 radiator work OK with the s54? Cooling capacity? Seems like it works just as good as the other radiators with the 24v swaps.

[Thing30]

Anyone have personal experience with SC Typhoon ECU...

[Wanganstyle]

as far as S54 news is concerned; currently working on a parallel DME solution to allow for "obd2" port function with S54 and OBD2 non s54 DME......

Soo, about those S54's :D

s54 port porn.




silly little fingers are awesome.


hiding in the woodwork, machine shop was double thumbs up on all the OEM CNC work.

[giantkeeper]

Soo, about those S54's :D

[Wh33lhop]

Thanks! I was thinking I was going to have to be the ass who points out what Torsen is... ;)

Like Wanganstyle said, changing the number of clutch disks changes the relationship of locking torque to clamp load on the disks.
IE, (pulling numbers out of the air) a 45 degree ramp with a 2 plate clutch pack might deliver 200 ftlbs of locking torque at a given driveshaft torque. The Bellville spring might give it 50 ftlbs of locking torque from preload.
The same ramp with a 4 plate clutch would deliver 400 ftlbs of locking torque at the same driveshaft torque. The same bellville spring in the 4 plate unit would have 100 ftlbs of locking torque from preload, BUT using a softer bellville spring can drop the 4 plate unit back down to 50 ftlbs.

So the number of disks *DOES* affect the slope of the locking torque vs drive torque curve.
LJ851 posted while I was writing, but I already knew what a bellville spring is ;)

Again, I got that in theory, but from how people were talking about building a diff it seems like they were just stacking the clutches one on top of another, which made no sense. A simple misconception is all.

[Wh33lhop]

Extra clutches adds MU by doubling surface area; just like a single/double/triple plate tilton for 1,000hp++ supra.

I get how that is supposed to work but that would require a mating friction surface as well. I was under the impression that you just stack the clutches which would do nothing for surface area and just act as a shim. I guess I was mistaken--which is good, because it would be pretty dumb to do it that way.

[The Dark Side of Will]

Ready for some pedantry? ;) Torsen. TORque SENsing. :p

You know, it occurs to me that adding extra clutches, because of how they are added--since you do not increase frictional area--does not actually increase increase locking rate, just preload. They basically act as shims.

Rather, if you consider a locking force vs. torque curve, adding extra clutches does not increase the slope of the curve like extra ramp angle, it only shifts it up. So, ramp angle modification would possibly be a more streetable solution as it will lock up more at higher torque levels (when you are goosing it), and lock up less at lower torque levels (when you are driving normally).

Thanks! I was thinking I was going to have to be the ass who points out what Torsen is... ;)

Like Wanganstyle said, changing the number of clutch disks changes the relationship of locking torque to clamp load on the disks.
IE, (pulling numbers out of the air) a 45 degree ramp with a 2 plate clutch pack might deliver 200 ftlbs of locking torque at a given driveshaft torque. The Bellville spring might give it 50 ftlbs of locking torque from preload.
The same ramp with a 4 plate clutch would deliver 400 ftlbs of locking torque at the same driveshaft torque. The same bellville spring in the 4 plate unit would have 100 ftlbs of locking torque from preload, BUT using a softer bellville spring can drop the 4 plate unit back down to 50 ftlbs.

So the number of disks *DOES* affect the slope of the locking torque vs drive torque curve.
LJ851 posted while I was writing, but I already knew what a bellville spring is ;)

[LJ851]

the ramp angle def. affects the locking. i.e. 2 way ramp, 1.5 way ram (asym), the lower the angle slope the easier it is to lock up.

ramp angle modification locks clutches more/less when drive shaft/axel imputs change.

ONE can add more preload by adding preload- there is a part in the diff (bellvue spring) for adding preload. Adding preload can be done with THICKER clutches and thicker dogs, but adding more clutches requires removal of spacer - so if done, one does not add any more preload (unless more preload is requested and specifically added.).

no added preload is required when adding clutches, many are adding preload when DIY rebuilding diff's and not completely understanding the end product (more potential understeer)


one could even add additional clutches and DECREASE preload :)


BMW also has OEM 4 clutch LSD's that are UNLOADED -no preload, no bellvue spring. BMW Started doing this in small runs starting circa E3 / E9 era :)

Extra clutches adds MU by doubling surface area; just like a single/double/triple plate tilton for 1,000hp++ supra.

for Aftermarket (KAAZ, OSG, CUSCO) LSD units with many discs one can change the pairing from clutch/dog/clutch/dog to clutch/clutch/dog/dog pairing to DECREASE locking :)

More friction surface pairs = more reaction to locking - one can alter these characteristics by changing ramp angles, altering clutch/dog pairs, altering bellvue spacers (adding, changing spring rate, or removing entirely)

More pedantry? Bellvue is a city in Washington state and although they probably have a spring or two there, it is not in your diff.

A bellville spring (also known as a conical spring) is in your BMW clutch type LSD.

[Wanganstyle]

Ready for some pedantry? ;) Torsen. TORque SENsing. :p



You know, it occurs to me that adding extra clutches, because of how they are added--since you do not increase frictional area--does not actually increase increase locking rate, just preload. They basically act as shims.

Rather, if you consider a locking force vs. torque curve, adding extra clutches does not increase the slope of the curve like extra ramp angle, it only shifts it up. So, ramp angle modification would possibly be a more streetable solution as it will lock up more at higher torque levels (when you are goosing it), and lock up less at lower torque levels (when you are driving normally).

the ramp angle def. affects the locking. i.e. 2 way ramp, 1.5 way ram (asym), the lower the angle slope the easier it is to lock up.

ramp angle modification locks clutches more/less when drive shaft/axel imputs change.

ONE can add more preload by adding preload- there is a part in the diff (bellvue spring) for adding preload. Adding preload can be done with THICKER clutches and thicker dogs, but adding more clutches requires removal of spacer - so if done, one does not add any more preload (unless more preload is requested and specifically added.).

no added preload is required when adding clutches, many are adding preload when DIY rebuilding diff's and not completely understanding the end product (more potential understeer)


one could even add additional clutches and DECREASE preload :)


BMW also has OEM 4 clutch LSD's that are UNLOADED -no preload, no bellvue spring. BMW Started doing this in small runs starting circa E3 / E9 era :)

Extra clutches adds MU by doubling surface area; just like a single/double/triple plate tilton for 1,000hp++ supra.

for Aftermarket (KAAZ, OSG, CUSCO) LSD units with many discs one can change the pairing from clutch/dog/clutch/dog to clutch/clutch/dog/dog pairing to DECREASE locking :)

More friction surface pairs = more reaction to locking - one can alter these characteristics by changing ramp angles, altering clutch/dog pairs, altering bellvue spacers (adding, changing spring rate, or removing entirely)

[Wh33lhop]

Agree. I had a mz3 torsion in my s54 powered car. worthless half the time. one wheel wonder exiting over bumpers. Skip that. Give me an e30 locker.

j

Ready for some pedantry? ;) Torsen. TORque SENsing. :p

Street/milder setup track cars really dont need any more than 3 disc; a local track car (who drives it to the track) is running a 3 disc stock ramping angle 3.46-188mm paired with ls7 7.0L vorshlag setup, he is pretty happy, balanced item is key.

You know, it occurs to me that adding extra clutches, because of how they are added--since you do not increase frictional area--does not actually increase increase locking rate, just preload. They basically act as shims.

Rather, if you consider a locking force vs. torque curve, adding extra clutches does not increase the slope of the curve like extra ramp angle, it only shifts it up. So, ramp angle modification would possibly be a more streetable solution as it will lock up more at higher torque levels (when you are goosing it), and lock up less at lower torque levels (when you are driving normally).

[Wanganstyle]

Street/milder setup track cars really dont need any more than 3 disc; a local track car (who drives it to the track) is running a 3 disc stock ramping angle 3.46-188mm paired with ls7 7.0L vorshlag setup, he is pretty happy, balanced item is key.

Usually many on a quest for higher locking setting items are way overboard for chassis/footwork/driver setting. Jean's 4 disc is paired with full race+R comps; without the R tires and if was a street car would be pretty disagreeable.

David is running secret++ super locking setup in his, but car is only for 1/4 mile (9.05 currently). His diff is set to go STRAIGHT. no turning for that car

Items like 90/10 asym or similar staggering LSD settings are very specific; may cause irritation in driver results as it may really not be your style.

Suggestion? Start easy, increase locking when needed; LSD TUNING is specific to the driver and car :)
Since you can build an engine and a DME from scratch, you can probably tune your own LSD!

[nando]

how about something intermediate like a 3 disk setup?

I saw some pretty sketchy "rebuilt" 40% lockup LSD's on ebay.. made with used parts, for over $600. yeah, right!