Center diff. Upgrade

[DG-FNR]

AWD *Avenger*?

Without knowing any of the details on the suspension (motion ratios in particular) there's no way to tell,

A spool is instant morbid understeer.

DG

 

[jasonlee]

i actualy used the entire gsx drivetrain and suspension. not quite a bolt on, but it all worked.

and when u said spool was instand understeer, wouldnt u meen oversteer? the rear wheels would have a greater tendency to slide then wouldnt they?

 

[jtmcinder]

Under the limit, spools cause understeer because they force the average front wheel speed to be the same as the average rear wheel speed. Since the only time that front and rear averages are the same (under normal circumstances) is when the car is going straight ahead, spools cause the car to want to go straight ahead, which is understeer.

At and above the limit, spools cause you to crash. Whether you hit the wall tail first or head first isn't really that important, is it?

- Jtoby

 

[DSM PERFORMANCE]

About the Quaife, I've talked with RRE about that, and they don't know nothing about any modification or else (?)

 

[DG-FNR]

That's right, RRE doesn't know about it.

Hang tight. I'll relay info as it becomes availible.

DG

 

[DSM PERFORMANCE]

Any update ? still hanging tight here...

 

[MrAWD]

I have no info on any updates from the RRE (sorry), but if you are still fishing for a center diff option for your car, instead of going with a Quaife, I would recommend couple of options. One is to stay with the Speed Design 4 spider diff and a loose VC as JT mentioned earlier (for different diff, but he did mentioned it. With that setup your rear end would have to have higher spring rates and that would give you nice overall balance on the car.
The other option would be with Cusco 35:65 diff, but without any kind of VC unit on it. This setup would require a bit softer rear end, and some other mods that would tame the rear end, but overall that would give you nice balance as well. The point here is that even with 35:65 split, car is better then any other RWD car that is out there and handles 0:100 split all the time! With a bit more touch with the gas pedal, I believe that car can be made to turn very nicely! Still just an old me…


Fedja

 

[jtmcinder]

My understanding of ACM's trick of using a half-dead VC in conjunction with a 35:65 center is that the VC acts in a way that is similar to traction control. If the split ends up causing the rears to break loose completely, then the VC can step in and try to reroute some power back to the front. The rest of the time, the 35:65 is what controls the power and, as Fedja said, 35:65 is better than 0:100 without having some of the problems of 50:50.

One other issue to consider is the following. Quaifes attempt to send power to the wheels with grip. So if you are running a Quaife center, it makes sense to run a Quaife front. But a 35:65 center sends more power to the rear regardless of the grip available at the front. Under these conditions, it might actually be better to leave the front open (i.e., not add a Quaife there), because a Quaife is a two-way and adds some off-power turn-in understeer.

At least, if someone were seriously thinking about following ACM and getting a 35:65 center, it would be very nice if they first tried it with the stock open front, maybe even taking the car up to New England so that it could be tried back-to-back with ACM's. I'd really like to know if my intuitions are correct.

Alternatively, we could just take a collection and put a 35:65 in my car. In return (tee hee), I'd be happy to bop up to Massachusetts a few times and report on the results.

- Jtoby

 

[logic]

The other option would be with Cusco 35:65 diff, but without any kind of VC unit on it.

Fedja,

So to clarify, you're running the Cusco with a VC eliminator? Wouldn't that just lock up a very expensive diff? Or are you just using some kind of sleeve retainer for the output shaft, but otherwise leaving the system open?

Edit: Nevermind, looks like ACM is the one running this, using a half-dead VC; jtmcinder posted while I was composing, it looks like. *grin*

The other curious thing is that aside from Alamo and a couple of websites that look like they haven't been updated in a year, I couldn't find anyone claiming to still sell the Cusco center for our cars. Did they discontinue them? Looking at Cusco's website, it appears they aren't even stocking them in Japan...

 

[MrAWD]

My understanding of ACM's trick of using a half-dead VC in conjunction with a 35:65 center is that the VC acts in a way that is similar to traction control. If the split ends up causing the rears to break loose completely, then the VC can step in and try to reroute some power back to the front. The rest of the time, the 35:65 is what controls the power and, as Fedja said, 35:65 is better than 0:100 without having some of the problems of 50:50.

The problem with running the worn VC with Cusco is in overheating of the coupler. As Charles mentioned elsewhere, there is that hump mode that happens when you overdo your VC. When that happens it kind of become a welded for a little while. Since Cusco has a tendency to send more to the rear every time something changes in there (read it as floor/lift the throttle). That seem to make that already worn VC work extra hard all the time, and as a result you get more locking that you need in there. Of course, there need to be much more testing to really figure out what is going on, but from the car behavior, that seem to be the case.

The same setup with 50:50 diff in there would allow VC to rest a bit more and due to its nature, you could still have it loose overall. That might require gentle aggression from the driver, but I believe it can be done!

One more thing that is related to the behavior of the Charles’s car and the Cusco unit is his rear end. Since it was not working well since he started using it and creating a bit more of the inside tire spinning that we expected, once this gets fixed, I believe that behavior of the Cusco unit is going to change as well. We will know more about that in a few months I guess!!


Fedja

 

[MrAWD]

So to clarify, you're running the Cusco with a VC eliminator? Wouldn't that just lock up a very expensive diff? Or are you just using some kind of sleeve retainer for the output shaft, but otherwise leaving the system open?

Cusco supplies a small piece that you attach where the VC should go, so option is yours. There is no locking with that piece of any kind and only purpose is to hold the output shaft from falling out of the trainy.

And as JT said, all of my Cusco experience comes from what Charles told me about the car behavior with it.


Fedja

 

[ACM]

I run a VC with the Cusco for the same reason one runs a VC with the stock diff - that is, to turn an open diff into an LSD. The 35:65 Cusco diff is not an LSD, despite what Cusco et al put on their sites, it's an open diff. Without the VC fitted all power goes to the wheel that spins, so the VC is needed to maintain some front-to-rear power distribution.

It should be possible to drive a Cusco-equipped car at the limit without a VC fitted, just as it should be possible to race a stock diff-equiped car without a VC fitted. It will, however, take a particularly good driver to do this, a much better driver than me undoubtably.

Without the VC the car exists on a knife-edge, the transition from grip through wheelspin to car-spin is almost instantaneous - certainly much faster than I can react controllably. I have absolutely no doubt there are drivers out there who could manage this, however I am not one of them. Having the VC in place - even a well-worn one - expands that knife-edge to a usably wide plateau. Otherwise it's like driving a 911R and I'm living in fear for my life...I need that Viscous Coupling security blanket to give me the confidence to push that bit closer to the edge. Much more experimentation is obviously required.

The problem with a conventional helical diff - Quaife or Torsen - is that it pulls power from an axle as the grip goes - that's not what's wanted (imho). The power needs to keep feeding to the rear to keep the rear loose - any additional power fed to the front compromises front grip and makes the car push. The power needs to keep going to the rear. With an open diff and a (worn-out) VC this is exactly what happens.

What's really needed is a 35:65 LSD with an asymmetrical torque shift. I'm working on it but I can't put everything together yet to make it happen.

A Helical front diff is less than perfect - as JT mentions - but the under-power benefits of it outweigh the off-power negatives. If a 1-way helical could be sourced, that would be perfect, I don't know that such a device is possible. Which leaves the option of an open front diff. An open front diff reverses the above, so it comes down to what you feel more at home with. I had to work at embracing the Quaife front, but I do believe it is a better compromise than the open front diff. I think...

 

[jtmcinder]

Because of how they work, T2R helical diffs (of which the Quaife is one) are always two-way. They cannot differentiate between torque being transmitted from the input to the wheels or the wheels to the input because all they "know" or "care" about is the relative amounts of torque coming or going from the two wheels. Sad, but true.

Another serious bummer is that there are no T2Rs (that I can find) for the rear of a DSM. If you start looking at RWD cars for inspiration - because you now have a 35:65 center split - then you see that everyone is running these. Rather few people are running clutch-type (Kaaz) rears in RWD autocrossers. But that's what we - or, rather, ACM - are stuck with.

- Jtoby

 

[MrAWD]

Without the VC the car exists on a knife-edge, the transition from grip through wheelspin to car-spin is almost instantaneous - certainly much faster than I can react controllably. I have absolutely no doubt there are drivers out there who could manage this, however I am not one of them. Having the VC in place - even a well-worn one - expands that knife-edge to a usably wide plateau. Otherwise it's like driving a 911R and I'm living in fear for my life...I need that Viscous Coupling security blanket to give me the confidence to push that bit closer to the edge. Much more experimentation is obviously required.

Well, Charles and I already had discussion about this a few hours earlier, so this will be for the rest of people tuning in here. The way his car was/still is set was inherited from the car with all differentials stock (well Speed Design center, but still stock for our discussion). Such a setup requires as loose rear end as much as it was possible. We where there at the time and things that where done where 650 lb rear springs, 1/2” (or maybe even more) higher rear end then front, lots of offset bushings to decrease the rear camber below -2.0 degrees, around 50 psi of the tire pressure, straight toe (or even on the border of being positive) and RM sway bar which is the same size as the front one. Plus, shocks where valved with a tons of rebound in them and we had to use all that was there.

On such a car, switching from 50:50 to 35:65 was a huge change and it made a relatively loose car to start with to an oversteer monkey. Amount of grip was very low, but previous setup required such a setting in order to maintain the balance of the car. Now with Cusco, things where pretty bad. So, tire pressure went down as well as the rear car height, and there might have been a bit of negative toe but I am not sure about that. Still there where so many things left to be done to bring back all of the grip that was removed, but at the time nothing more was done. That still resulted in low amount of grip in the rear and Cusco was doing its work of sending extra power to that side of the car. To make everything even worse, rear Kaaz was misbehaving as well and Nationals was too close for more changes. So, stock VC was fitted and things become better.

Still some of the drawbacks where that hump mode that would make car to be very tight, and to make it even worse it would happen intermittently so there was no much of an adjustment left for the driver. Once you get there that you don’t know what the car will do if you do something, you start to loose your trust and you ended up slowing down as a result. Not a nice spot to be!!

So, the knife edge that was mentioned at the beginning could have been decreased or even completely removed if more of the changes where made to the rear end of the car (possible front as well). I believe that with more rear grip this knife edge would be under control and as a result you would have a lot more grip in the rear, which would still be loose enough. That would create overall higher grip level for the whole car and as result there would be shorter time slips. I guess only way to know this for sure would be after lot more testing, so we will see what will be the final conclusion on this!


Fedja