Center diff. Upgrade

[DSM PERFORMANCE]

Hi, I'm wondering about the best option for upgrading center diff. for RoadRacing. I know Quaife do wonders, but not really recommended because can take out intermediate shaft.

So I'm wondering about Kaaz center diff., I know it's stronger then a stock diff. but does it improve handling over stock ? Is it comparable to a Quaife ?

Quaife Front diff. is the way, but looking the best for center.

Thank's

 

[MNGSX]

Either the kaaz or a 4 spider..

In a perfect world the quaife.

Talk to TRE on the Quaife.. I dont know if they have their ring gears figured out. They may have started treating them differnetly which means they are ok to run..

 

[DSM PERFORMANCE]

Last time I've talked with Ming (at TRE) it was in September, and he recommend me not going with Quaife center, same thing for Sheperd. 4 spyder gears is a good option, but it not improve handling, same as stock but stronger. Does Kaaz improve handling, this is the question, I've ask it to John and can't answer me...

 

[MNGSX]

The kaaz will be an improvement.

It is'nt an open differential.

 

[Poloturbo]

How about the Cusco center diff? Anybody runs them.

When you say: "I dont know if they have their ring gears figured out." Is that the cause of the intermidiate shaft to brake or the casing of the quaife beeing the problem.

 

[jdmspilner]

I have a cusco but I have not used my car yet...

 

[DSM PERFORMANCE]

Poloturbo where do you come from ? Did I know you ?

Here is the answer of Mike at RRE :

People making big power pull the teeth off between the 3rd gear on the intermediate shaft and the diff housing on stock diffs and on Kaaz and Cusco diffs too. All the power goes through that gear connection all the time, especially when you are in third you put a lot of stress on that. Quaife got tired of giving away free differentials for a problem that was not theirs. At least they have parts available. Kaaz and Cusco do not have new housings available. If you kill one of theirs you have to buy a whole new diff. When Mitsubishi upped the power in the EVO in 97 they flipped the motor around to the other side so that they could redesign the whole tranny to make it all stronger and they were able to make more room by talking out a complete shaft.

Dead Cusco diff:
http://www.roadraceengineering.com/eclipsetech/deadstuff/scot-deadtrans.htm

 

[DG-FNR]

Quaife is currently redesigning the case of their diff to eliminate the problems.

DG

 

[Poloturbo]

Quaife is currently redesigning the case of their diff to eliminate the problems.

DG

Very good news indeed... Any date schedule?

 

[DSM PERFORMANCE]

Quaife is currently redesigning the case of their diff to eliminate the problems.

DG

Oh yeaaahhh!!!!!!

 

[Revolution]

Is this new cassing in development or the experimental stages?

Why not make the center lsd into a spool?

Elaborate.

 

[jtmcinder]

The idea of running a spool in the center of a road-racing car is just whacked. Let's just drop that right now.

Your options in the center are a Kaaz (clutch-type), a Quaife (Type-2 Torsen), or the new Cusco 35/65 (planetary). The Cusco is open and really meant for drifting (given the huge bias), but it can be tamed using a half-dead VC, as shown by Charles Moss.

My suggestion is conditional. If you want something that slap in and go, with no testing and tweaking (which involves opening the tranny every time), then get the Quaife. If you are ready to put in the time and effort to tweak, then I'd try the Cusco and half-dead VC.

Why?

I'm against the Kaaz since it locks as a function of input torque. If you're trying to use your right foot to balance the car, I wouldn't want to also be trying to use the same foot to control the diff. Maybe you're a better driver than I am, but that seems to ask too much.

The Quaife is great, but in order to get it to send power to the rear for a good corner exit, you must stiffen the front end of the car a lot. Look at the spring rates that Dennis Grant runs: 900/400. This might give good turn in for autocrossing and might allow you to feed in lots of power for the exit, but mid-corner (steady-state) it would cause understeer. In autocrossing, we don't have much mid-corner to worry about, but in road-racing, you have lots.

The Cusco - assuming that you can tame it - allows you to send power to the rear for a good exit without stiffening the front. So you can have a balanced car at mid-corner and still power out. The problem is the extremity of the bias: 35/65. That's why you need the half-dead VC on top of it.

If someone made a 45/55 (like what came stock in a 3000GT) then we'd be golden. If any of those bastard copy-cats from OBX are reading this, then that's what you ought to make.

- Jtoby

 

[DG-FNR]

Look at the spring rates that Dennis Grant runs: 900/400. This might give good turn in for autocrossing and might allow you to feed in lots of power for the exit, but mid-corner (steady-state) it would cause understeer.

No, it wouldn't.

And in fact, it doesn't.

The steady-state handling bias, as determined mathematically and verified on the skidpad, is slight oversteer with those spring rates and the bars that I'm using.

DG

 

[jtmcinder]

Interesting. Please walk us through the math that predicts oversteer for 900/400 with RM bars. Or do you have a new front and/or rear bar?

- Jtoby

 

[Poloturbo]

Interesting. Please walk us through the math that predicts oversteer for 900/400 with RM bars. Or do you have a new front and/or rear bar?

- Jtoby

Correct me if I'm wrong but isn't Dennis running the stock rear bar and the RM in front?
Rear is 18mm I think. Front would be RM spec 20.6 and with 900 spring rates and Quaife LSD.

 

[DG-FNR]

Correct me if I'm wrong but isn't Dennis running the stock rear bar and the RM in front?

That is correct.

If you'd like to know why this works, start here:

http://www.amazon.com/exec/obidos/ASIN/185960644X/thedsmautocropag

and then move to here:

http://www.amazon.com/exec/obidos/ASIN/1560915269/thedsmautocropag

You will need to know (at least)

CG height
track width
wheelbase
shock/spring motion ratios front and rear
roll bar motion ratios front and rear
roll bar stiffness
roll centre heights front and rear
sprung weight
unsprung weight
corner weights
spring rate front and rear
tire spring rate front and rear
natural frequencies front and rear

DG

 

[Revolution]

"Quote"If someone made a 45/55 (like what came stock in a 3000GT) then we'd be golden. If any of those bastard copy-cats from OBX are reading this, then that's what you ought to make."Quote"


Would it be possible to run that center diff in our cars and hipothetically speaking what would be involved in doing so?

I am not as suspension suave as Dennis , but it kind of does make sense as to what he's doing.

Plant the front more so the rear is forced to travel , but I thought that by having both the front and the back rm sway bars they would promot a better all around balance. Would Dennis' setup work on a less modified car?

P.s im still wondering if that quafie diff is in the experimental stages or would it be seeing daylight anytime soon?

 

[jtmcinder]

Would it be possible to run that center diff in our cars and hipothetically speaking what would be involved in doing so?

Nope. While it is a planetary with the input being the outer ring and the two outputs being the middle carrier and the sun (similar to the Cusco for our cars), the VC on a 3000GT is right there on the side of the diff, while ours is out on the end of the output shafts. For all I know, it's also a different size. You see, Mitsu seemed to be aware of how much power the turbo V6 would put out ... they showed little awareness of how much we'd be able to get out of the 4G63.

- Jtoby

 

[DG-FNR]

Plant the front more so the rear is forced to travel , but I thought that by having both the front and the back rm sway bars they would promot a better all around balance.

The issue with an AWD is that a 50/50 split diff is always going to induce power-on understeer. Always. A-L-W-A-Y-S. Using more sophisticated diffs can *reduce* the understeer, but never eliminate it.

So what a lot of people wind up doing is severely biasing roll resistance rearward, in an attempt to induce power-on oversteer. What they get is a shitload of corner-entry oversteer - which can indeed result in great gobs of corner-exit oversteer if the rears actually break loose.

It takes practice and sensitivity to be able to tell the difference between "real" power-on oversteer and "pitching the car resulting in corner-entry overseer that carries through into corner exit" - sometimes summerized as "Whoops!"

My setup is still rear-biased - corner entry understeer can be useful in mild amounts, and is nice in slaloms. Too much though makes for an unforgiving car that likes to snap-spin without warning.

I got a bit of a leg up on this when I had a centre diff go out in such a way as the car became RWD. It was the twitchiest, most tail-happy car I have ever driven. DSM's don't drift, but this one would get 20 degrees of tail-out by breathing on the throttle - and that was with 750/400 and RM bars on both ends.

Well, that balance was always there; the diffs were just masking it.

The neat thing though is that diffs crosstalk. Grip freed up on one end can somtimes result in grip freeing up on the OTHER end of the car too. I softened the rear bar to tame a little bit of the corner-entry oversteer out of the car. It worked - and I got FRONT grip as well. Shocked the hell out of me.

Once you do the math, the reason why starts to become clear.

DG

 

[jtmcinder]

Would Dennis' setup work on a less modified car?

I have never heard of anyone running anything close to 900/400 with a RM front bar and stock rear bar with stock differentials.

- Jtoby

ps. Dennis - don't forget to put the weight-grip curves for the tires on your list

pps. what the heck is a motion ratio for a swaybar, anyway? all I've ever worked with is the stiffness in pounds per degree of roll

 

[Revolution]

Ok I somewhat understand where you guys are going with this.

First off jtmcinder you'r stating that because the center lsd from the 3000gt is placed differently that it wont work with our cars, correct? In the mood for speculation couldnt their be a way of fabbing it to work ,ie a custom center shaft? Maybe it might be just as much as buying a brand spanking new lsd. Just a thought.

DG-FNR you stated that the reason you changed the rear bar back was in an attempt to further reduce the overwhelming over steer you where expeariencing with you'r previous setup.

My question was to understand whether this was done for you'r own personal preferences or because the car would ultimately handel better in entry and exit.

I my self will be limiting the moddifications to wide rims and tires ,the rm sway bars and of course you'r 2g spec DG suspension setup. Other moddifications i.e a roll cage may be in the works for future endevours for when I decide to begin participating in solo1 .

The reason I ask is if it does make the car a much more well rounded vehicle then their wouldnt be any point in me keeping the rear bar . I ultimately want the car to nuetrally slide into corners rather then have an excess in either direction.

On the topic of lsd's both of you members agree that the quafie would be the best all around choice for a street driven / track car , over say a cusco unit.

Hmm please feel free to correct me if im wrong.

 

[jtmcinder]

DG-FNR you stated that the reason you changed the rear bar back was in an attempt to further reduce the overwhelming over steer you where expeariencing with you'r previous setup.

You may be onto to something. Something along the lines of entering too hot, using trial-braking too much, and softening the tail instead fixing the nut behind the wheel.

Sure am glad that I wasn't the one to bring this up, given how Dennis usually responds to suggestions of this sort.

Or did the unfreezing process cause me to lose my inner monologue?

- Jtoby

 

[DG-FNR]

DG-FNR you stated that the reason you changed the rear bar back was in an attempt to further reduce the overwhelming over steer you where expeariencing with you'r previous setup.

My question was to understand whether this was done for you'r own personal preferences or because the car would ultimately handel better in entry and exit.

A bit of both.

I like a car to be a little bit loose on entry, but when the car is going around once an event, there's a problem.

That's the nice thing about doing the math - it QUANTIFIES exactly how loose the setup actually is. It lets the engineer predict not only the direction of the balance (push vs loose) but also HOW MUCH.

With the previous setup, the answer to "how much?" was "a ton". Grossly excessive corner-entry oversteer. With the bar change, it moved to just "excessive".

And I should point out that I did it under duress. Like everybody else, I'm convinced that I can drive anything, and so backing off on the oversteer seemed like admitting defeat. Imagine my suprise when not only did I NOT lose any turn-in (the car IS still loose, it's just less prone to suddenly swapping ends) but I also gained EXIT speed too.

THAT little suprise lead to the question "WHY?" and answering it opened up new insights, and the car got even faster.

Incidently, taming the back end down a little bit helped ENTRY speed as well, as it increased the amount of driver confidence that the rear wasn't going to snap around and so let me crank up entry speed a bit. It was an enormous overall win, and I wish I had tried it earlier. In retrospect, it's an obvious step in the right direction. The math just reaches out and grabs you.

Part of doing racing engineering is to realize that you cannot force your will onto the car. The car wants a certain setup to work optimally. A lot of rear roll resistance costs overall grip AND offers no benefit in return. Syoez flexible! Learn from what the car is telling you!

DG

 

[Revolution]

We really have to hook up for a coffee next time you hit up magnus...

 

[jasonlee]

this is a very intersting thread. i am gonna lay down another senario and see what you (dennis) or anybody else may think...

what if the car was a tad bit longer? and a tad bit heavier? i am running tein flex coilovers, and rm bars front and rear... on my awd dodge avenger... any thoughts?

i just read an article in "dsport" magazine about handling in turns in regards to suspension setup...

they seem to say that if you have a tight front end, you will have understeer, and a tight rear will have oversteer...

im looking into big power for 2005 and now i am worried about trannys/diffs,
now and the like..i was thinking of going with a spool, but now i am unsure....... i am gonna watch this thread like a hawk!

J