Why you need a LSD if you like cornering!
 

http://i3.photobucket.com/albums/y75...ps8031d56d.png

So a brief explanation of what is going on here

Top graph = throttle position
2nd graph = Front left wheel speed
3rd graph = Front right wheel speed
4th graph = Rear left wheel speed
5th graph = Rear right wheel speed
6th graph = steering angle (negative = left, positive = right)

Blue/purple thin boxes - compare the left and right wheels, and you can see momentary lockups under braking. Decreasing wheel speeds mean I'm braking the car, and the little sudden dips are the wheels on the verge of locking up, but recovering before they actually do. The wheel speed would go to 0 if it actually locked up. This is shown to demonstrate the accuracy of the equipment we are using, and validity of this test.

Red box #1 - under acceleration and turning, notice the wheel speed differences between the rear left and rear right wheels. The rear left wheel is the "inner" wheel in this case, and you can see spikes where it speeds up, and then slows back down as I modulate the throttle. With a mechanical LSD, I would be able to give the car more throttle without the rear wheels going at different speeds or spinning altogether. Spikes on the left rear wheel here indicate that the inside wheel is spinning, rather than gripping. The inside wheel travels a lesser distance than the outer wheel, and should be travelling at a lower speed, not a higher speed.

Red box #2 - Notice the dips and peaks of the rear right wheel, compared to the rear left wheel. The spikes up and down are a result of throttle modulation, but I am unable to put down any power, because when I do get on the throttle, I'm spinning the inside (rear right) wheel instead of accelerating.

Red box #4 - Note the spikes in the rear left wheel. Same thing; I'm modulating the throttle (top graph) to accommodate the car's inability to put all the power down. If the inside rear wheel were not slipping, I'd be able to corner faster, and put down more power.

Red box # - I manage to spin BOTH wheels for once, and need to countersteer.... :)


TL;DR: mechanical LSDs help you power out of corners, and corner faster, regardless of power output levels (note that all the times where I cannot power out are at partial throttle) and driving experience.




*edit*

Some highlights from others.

Thanks Mike, nice writeup.

hey mike, as an aside, did you pull the wheel position sensor using the Solo DL or a more advanced unit?


did AiM release a new FRS/BRZ file

For my own clarification...

I'm reading this as: you can't do this because the car doesn't have a mechanical LSD.

I thought the FR-S & BRZ came standard with a Torsen LSD. Isn't that a mechanical diff? Please clarify if I've missed something obvious.

Src:
[ame="http://www.youtube.com/watch?v=u0EOrzn7-Qc"]http://www.youtube.com/watch?v=u0EOrzn7-Qc[/ame]

This is from the latest AIM Solo DL firmware.

Given that the AIM Solo DL now reads Oil Temp and Wheel Speed sensors, we're thinking about doing a group buy if there is enough interest.

The car comes with a Torsen, but if you lose traction with one of the wheels, then the differential effectively becomes open. Loss of traction can be lifting a wheel, or spinning a wheel. In this case, the inner wheel is spinning.

I just updated! (didn't realize I was on xx.53 and not xx.54)


now I have all these fun things too! :D

@CSG Mike, you're doing this only to torture me until my OSGiken gets installed...jk ;)

@RehabJeff86 is also in the same boat. LOOOL

BUt I thought that if we lifted a wheel the ECU would use the VSC to apply brakes to the spinning wheel.

In theory, but that's really, really intrusive, and you don't want that to happen on track. I've seen guys get sideways at 120+ because of that.

Also, that functions by slowing you down; the goal is to go faster :)

So did you pull the fuse so that the VSC would not respond?

Ummm... SO slowing the spinning wheel would slow you down??
I thought that was the whole concept of a LSD.:iono:

its applying brakes do regain traction is why it slows you down.

it's slowing down the spinning wheel so the LSD kicks in and gives power to the one that's not spinning


but because this is achieved through electronics and well.. brake actuation, it's really just a hack

and what kind of lsd has been proven to be good for this platform?

Doesn't that defeat the purpose of having a limited slip differential? I recognize that it's not a locker, but suddenly having an open diff when one wheel loses traction makes it seem like the limited slip is REALLY limited.

please note this is for when one wheel loses complete traction (or on very different traction surfaces)

A locker is something used on a truck or tow vehicle to force it into locked or unlocked state. A LSD is reactive, and dynamic.

Yes, that's the downfall of a Torsen, but the Torsen type LSD has the least maintenance (change fluid) and makes zero noise no matter the situation.

Any clutch type. OS Giken and Cusco are the two we've tested, and both work well, and can be set up to be very daily driver friendly. The Cusco is cheaper up front, but the OSG has a longer service interval for major services.

The purpose is to prevent the slip to begin with. We use the Pedal Dance.

I've never gotten inside wheelspin in either the S2k (T2) or the FD (T2R). Any time I've gotten wheelspin at the track, it's been both rears.

Both wheels on the ground, you shouldn't be getting inside wheelspin with a torsen diff. If the inside rear is lifting while on the gas, setup may be off. I haven't tracked an FR-S, so maybe its 55/45 weight distribution is more of a problem here on grippy tires...

In my experience, the Torsens in my cars act similarly to the lightly-preloaded clutch/ramp-type in my 240Z. Actually, the FD's T2R acts more like the Z's clutch-type all around than the T2 in the S2k.

Yeah, you can crank in a ton of preload and make a clutch-type act practically locked (very high effective bias ratio), but that sucks for handling (unless you just like understeer). With a torsen you're limited to about 2.5-4 bias ratio, but usually that's enough. At least with a well-balanced car...
Then again they didn't put a Torsen in the Boss 302 LS (which has similar weight distribution issues as the FR-S/BRZ and a ton more hp) because it was *worse* than the clutch-type that went in lesser Boss 302s...

Personally, I'm happy either way. Never felt the need for a Torsen or Quaife gear-type LSD in the Z while others were happily spending $$$$ for them, and don't feel a need for a clutch-type in the S2k or FD. If it works, it works. And if it doesn't work for one driver/car at the track, it doesn't necessarily mean it won't work for another car/driver at the same level.

My opinions based on my experiences with other cars, so take it fer what it's worth :)

Mechanical diffs are quite brutal in a daily.

I have one. I run it because it's so much better for drifting. But it does take away alot of street comfort.

Datalogging would say otherwise on my S2k... I have the benefit of having DBW/VSA, and individual wheel speed sensors.

If you're at the limit of adhesion of the rear tires under cornering and accelerating, you'll get single tire wheelspin before both tires spin, because the torque going to the inner and outer wheels are not the same. If you launch the car, then yes, you'll spin both wheels because they're getting relatively even loading, but under cornering, that just isn't the case.

Do you have someone locally that can reconfigure it for you? I may be able to give you some guidance.

What brand do you have, and what is your current configuration?

Don't get me wrong, I'm happy with it.

I'm just stating that the axle binding and tyre chirping that comes with these diffs is something most aren't aware about.

I have a TRD 2way.

Those two things can be completely eliminated for street driving with proper configuration, which is the downside most people are afraid of.

The problem is, most retailers just sell LSDs, but don't have a clue about how to actually configure/customize them.

Some of your posts can come across a bit arrogant FYI. (I'm one to talk)

It depends how tight you set it. I wanted mine super tight, because its for drifting.

There is always going to be binding and tyre squeal in a daily with a mechanical. It just depends on how tight the diff is set, as to how often it happens.

Having it so lose that it NEVER happens in a street car, would mean its too loose IMO.

Your S2k is different from mine.

With an open diff, torque going to the inner and outer wheels is *always* the same. The reason you get inside wheelspin is not because more torque is going to the inside wheel, it's because the inside wheel has less grip because it is unloaded.

With either a clutch-type or a Torsen, frictional forces transmit torque that's trying to spin the inside wheel up over to the loaded wheel side.

If setup is such that the rear becomes totally or nearly totally unloaded, then a highly-preloaded clutch type will have the advantage of continuing to drive the outside rear wheel when the torsen will spin the inside.

It just *is* the case for my S2k and FD. At the track (not a drag strip), I always spin both rears at the same time, never get inside wheelspin.

PLENTY of Torsen equipped cars have no problem putting pretty big power down at the track on DOT R tires.

If your setup is such that you are lifting the inside rear a lot, yeah, you'll benefit from a highly-preloaded clutch-type diff.

Lots of Z guys ran rear roll stiffness biased setups that required a superstiff diff, but I went with softer rear roll stiffness (no rear bar) and a very lightly loaded diff. Both methods work, but I liked the way mine handled better...

Would you be able to achieve exactly what you want with a lower initial torque setting (dd friendly) with a higher rate of lockup (aggressive engagement)? What is the benefit of making your 2 way so tight that it is basically always locked? You might as well use a welded diff if you've eliminated the "differential" part of "LSD" by effectively permanently locking the diff.

IMO, the purpose of a LSD is to be able to have the wheels locked only when you need/want it to be. My interpretation of what you wanted would be that you want them locked permanently.

Quite simple really, you need it that tight for drifting in the wet, when you're giving it very little beans before wanting both wheels to engage. Having it that tight on decel means it feint drifts like a boss.

If I'm super smooth and slow during a slow maneuver than I can stop it engaging, but its so slow often I don't bother and just let if bounce/bind.

Also remember that no matter how tight the diff is, you can always use it like an open diff by pressing the clutch in. I'll often do that say if I need to make a u-turn, I just get enough momentum to do it without power.

Every drift car and most of my rwd street cars ive run Nisbro diffs. (NZ Slang for a welded diff) They're good because they always work, unlike a torsen. (My torsen was already getting lazy after 2 drift days) But the 2 way gives me the compromise I wanted, basically all the good things of a locker, but with some small areas where it performs like an open.

I don't disagree that I run it tighter than what would be ideal for circuit. I wouldn't even run a 2 way for a circuit car. I'd be much more inclined to run such a loose decel setting that I might as well run a 1 way. But my point from my previous post still stands, that if its so loose that it NEVER binds/engages on the street, then its probably too loose.

Welded diffs are great. It's actually one of the best modifications I did to a 240SX at one point. It's funny driving around grocery store parking lots though because one tire is constantly creating noise. :bonk:

Two things:

Some people are working to a budget. I certainly was when I had my first couple of silvias. Spending $1500 on a mechanical LSD was the last thing I was going to buy, when I could weld it up for effectively nothing.

Having my welded diff chirping around a car park was a small compromise in my eyes, for the realiability and predictabilty it gave while using it for drifting.

In that sense, you're right. It was probably the best modification you can do for virtually nothing. I paid $50 for my S14 diff, welded it, cut some wood bushings, drilled them out and boom..."solid" bushings for the S13. :party0030:

However, comparing that to a mechanical clutch type LSD was a very nice experience as well. While Torsen works great for what it does, I do desire more at times, hence the recommendation of the clutch type LSD. :)

Tight diff in general requires more rear roll stiffness to counter the understeer, which unloads the inside rear more, which requires still more diff tightness! Can be kind of a vicious circle.
I remember pushing a friend's ITS 240Z with a "Kennesaw locker" (welded diff) in the pits, when he cranked in a little steering, the car just stopped! This cannot be good for handling, thought I...
Point: there's more than one way. I don't think a tight clutchtype diff is necessarily necessary for putting the power down at the track. Some approaches and some driver preferences will lead that way though.

It's always in preference. The way you drive a windy course with a welded diff is definitely going to be different compared to an open one and a LSD setup. It's really assimilating to the car's abilities in order to extract the most out of it.

In this case, we have actual data that proves the settings we were looking for worked exactly as intended. The OTS settings in the past locked up a little harder than what we wanted. ;)

I did do some circuit days with mine prior to the 2 way.

I'm no CSG level driver (not a dig, I'm not) but I noticed almost straight away the 'auto-lsd' type aid the car has. Definitely not good for corner exit speed.

A few of us in NZ have 2 ways now, and what becomes apparent when you go from a torsen equipped car to a mechanical equipped car, is how much more the mechanical equipped car seems to spring out of corners better. Where as in realty, its the torsen equipped cars that are being held back by the car.

Which brings up another point: OTS (Off-The-Shelf) solutions are not likely to ideally suit every driver. It's not as simple a problem as "Stock Torsen sucks, I got Brand X clutch type and it's the bomb". For most people who track these cars, I think they're gonna be better off at least trying to optimize setup around the stock diff and spend that $1500 on track time and tires/brake pads instead. With experience they might have a better idea how to spec a diff to their liking if the stock Torsen isn't working for them. For someone like myself, I bet my preferences would lead to a setup that would work with the stock Torsen LSD.

That is a point that is very important to make. @CSG Mike presented data we derived on a track that utilizes mechanical grip more than grip obtained from aero. In stating this, we wanted to see a CSG configured diff put through its paces to find out how well the LSD was performing. Data is important to validate claims, statements, and proving theories. In either case, I believe these conversations with you and @diss7 show that our goal and points are, at the very least, communicated. Every setup is a specialized setup, but it definitely shows that the differential does improve performance when done properly. ;)

Realistically, having a CSG spec'd diff is a golden opportunity for most. Because its going to be based off real world data, using the good drivers and this car.

Its a not brainer to get it if you..
Are a circuit guy
Not really sure how to set a diff up (most poeple don't)
Don't have the time/money/desire to pull the diff out and try different settings (almost NO one does)

If I was going to focus on circuit I would buy a pre specced CSG diff, purely for the R&D put into it's settings.

how much would an autocross spec dif differ from a circuit type, I wonder?


would one aim to have more lockup for easier rotation by throttle?

After reading Mikes first post my only question was, 'why aren't you trying to load up the unloaded tire to prevent the wheel spin instead of ****ing around with the diff? CSG already has the ability to adjust rear roll stiffness.'

Of course, my question has already been answered in that it comes down to personal preference. But along the way I've learned yet more about the artistic side of suspension setup. Thanks guys!

No, more lockup => more understeer off the gas, more understeer on the gas.

Which is one reason I'm kind of philosophically opposed to tightening up the diff. You're adding work to the tires that is trying to make the car go straight. My approach would be to try to make the car work with as little diff preload/tightness/lockup as possible.