OMG Matt! That's alot of dead LSD inners. Well, it all started with you jumping in with my DIY when I got stuck and your excellent customer service paid off :thumbup:

I think Lar's 2010 GT3 LSD will be on it's way to you. I may tackle a sequential rebuild/swap with upgrade LSD if I get a Cup car ... stay tune. Have a happy new year.

Cheers

Mike

 

Finally got Gisele back this morning from a rebuild of the diffy. And I have to say that Matt at Guard Transmission has the BEST customer service of anyone I've ever dealt with. Can't praise him enough! Thank you, Matt!!!

Now I'm looking forward to getting her out on the dance floor and making her sweat! LOL

 

Enough of the dirty talk and time to beat up your 2 wheel drive GT3

 

I heard that's just what he did last Sunday with a run through Marin County up to Point Reyes and back...

 

Nah, can't work up a sweat on public roads. Gotta get her out on the race track to work up a good sweat! LOL

End of this month, then she'll get a good workout.

 

great info. thanks everyone.

 

I've been reading this thread several times now, but still struggles a little bit with getting a firm grasp on preload and what it really does. I found this pice of information somewhere else on the net:

This makes a lot of sense to me, and is also in line with the experience different drivers in here has posted: When the preload is gone, the tail wiggles in the transitions phase in turns.
But the LSD can still be able to lock up fully while under load, both in acceleration and breaking and thus be "working".

But when the tails wiggles and the cars gets unstable while under heavy breaking, this is indicating that the LSD itself doesn't lock up anymore.

So the role of the preload is to make the diff function less like an open diff when the difference in wheelspin is not big enough to make the LSD lock, and a very high preload will make the LSD lock much more of the time. (If I appear slow on this subject, I was thrown off a little bit by an early comment made by Matt of GT Guard stating that their belleville washer actually pressed the rampes together - this would indicate that the effect of higher preload would make the diff act more like an open diff and not the opposite which is clearly wrong). A very high preload will ensure that the diff always is always locked up, is that correct? But different preload will affect how much slip the plates will be allowed, while the rather large force applied under heavy breaking or acceleration will assure that there is no slip at all (when the diff works correctly) as the force on the ramps are so large.

Also, high preload will affect the car in slow, tight turns as it will fight difference in wheelspin, and will (for some people) be a negative experience on normal road use. I guess this is the downside of preload that has to be evaluated for a car that is to be used on road and not only on track.

Anyway, great thread!

 

I think we need to start out with some definitions so that we are all on the same page here, and then we can look at my comment which has been taken somewhat out of context. The terms high (or even very high) and low preload are getting thrown around a bunch here. Those terms don't really mean very much unless you've got some sort of relative definition. Especially if you go back and look at the Giken sales pitches where the spin he's putting on it is that pre-load= push and a locked diff without really putting any sort of relative reference point into the conversation.

Here's my definitions:
Low preload = 35-50ft/lbs. We pretty much never go below that on our LSDs, though some of our customers do. In fact, I've got one customer who has me remove the belleville washers completely and replace them with flat plain plates for a zero preload configuration.

Moderate preload = 50- 80 ft/lbs. This is pretty much how I configure ALL of our LSDs. With only one exception (Tom Trew and his 6 plate GT2 build) every person in this thread running a GT rebuild has an LSD in this range

High preload = 90-150 ft/lbs. This is a pure motorsports configuration. And many (maybe even most) of our motorsports customers do not use this sort of set up. The ones that do tend to go on the low end of 90-100ft/lbs. The Porsche factory motorsports LSD is generally 120ft/lbs. Brian Copans is known to go even higher than that and go as high as 140-150ft/lbs.

Ok, that said, let's look at how preload works in our LSD as I configure it. For starters, I don't disagree at all with the comments about the impacts of high preload on braking and on turning. In fact, those effects are part of why we use a moderate preload on our LSD. The LSD has what I call the "open phase". This is not a technical term. It's my own terminology but it describes well the steady state of the LSD when it's not fighting wheel slip. Preload, in whatever amount, does play against the LSDs efforts to function like an open differential. It's a balancing act when you play with the preload because you want enough to give some stability under braking, but you don't want so much that it makes the car act like the LSD is locked all the time. This is the thing we struggle with and on each car it's slightly different.

Let's look at the context of my comment regarding the preload helping to keep the LSD functioning in the open phase. Where this specifically takes place is under partial throttle midway through the turn.

One of the liabilities we've seen in the zero preload LSDs is under those circumstances the zero preload LSD sometimes has a tendency to snap from open to locked pretty instantaneously. There's no progression to the lock up. It's open and then it's locked. And if the driver doesn't know it's about to happen it can unsettle the car and break traction, causing snap oversteer on corner exit. Once the driver knows it's going to happen, he can prepare for it and even adjust his driving style to deliberately set the LSD and modulate the throttle to help keep the car from over rotating. There are also ways to tune around it, and if you look at someone like Chris Musante, who has been running the Giken LSD for a season, you will see that the LSD can be restacked in such a way that negates a lot of that and lets you dictate where in the turn it locks. The Giken just does that tuning more with alternating friction and plain plates instead of with preload. It's a different approach to the same end.

Then there's the Porsche Motorsport high preload approach. As described above, that LSD tends to act locked pretty much all the time and it's behaviour through the same corner is that it's going to make the car want to push. This is exactly what the Giken salesman was talking about and is a known behaviour of the high preload motorsports units.

Lastly, there's our midline approach. Our moderate preload does give you that bite and stability under braking. But then when you start going through the corner and start to roll onto the throttle the amount of force is not enough to keep the two wheels from working independently. It doesn't make the car push in the same way that the high preload Motorsports approach does. The moderate preload simultaneously helps to keep the cups (aka ramps) together and out of the real locking phase (which occurs when there is significant wheelspin and the pins push the cups apart really locking the LSD). What it does is it makes for a progressive lockup. People driving our LSD report that there's a smooth transition from open to locked and that it's very predictible and something they can feel occurring. In our opinion, no preload makes it happen too abruptly and high preload makes it always somewhat present. Moderate preload is our solution.

I stand by my statement that our moderate preload helps to keep the LSD in the open phase without prematurely locking. We do not use a high preload configuration except by request. High preload does do exactly what you say it does, but when I said that the preload helps keep the LSD from locking, it was specifically under the circumstances above in specific reference to our LSDs and how we configure them. We don't put enough preload on our LSDs to make them act like they are locked at rest. I'm sorry if I wasn't 100% clear about that in the earlier discussion. Does that make more sense?

I want to close by saying that tuning LSDs has a lot of black art involved in it. You can ask 5 top teams how they configure their LSDs and get 5 different answers. We recently spent more than a month of consulting with Rick Deman on the LSD set up for his Grand Am Boxsters. Rick would call me up and bounce ideas off of me and then he would go test them. He'd come back a week later with results and then we'd change something else in response to what he had found. In that process we changed ramp angles, stack height and preload. Sometimes he would send out two cars that were identically prepped except for preload and see the results. Eventually we found a set up that he felt worked, and the results speak for themselves. Deman went out and set a track record an NJMP and put one of the cars on the outside pole at Daytona for the innaugural Continental Tire race of the season.

 

Thank you Matt, great stuff.

 

BTW, rep points awarded for asking well thought out and educated questions.

 

Just read the whole 16 pages and learned a lot of information regarding LSD, thanks Mikymu for starting these threads a learnt a lot from your threads, Matt also from Guard Transmission and everyone who contributed in this informative thread.

After reading the whole thread and from what im seeing from my cars behavior I think my lsd is also gone bust, by moving from a red light my rear right wheel spins while the other doesn't and this isn't while launching the car just by normal driving ( happens 6 out of 10 times and on even normal roads ). Or im i wrong and it might be a sign of something else ?

If its the lsd, what preload should i go with ? Im planning to keep my car as a dedicated track car with little to no street driving.

Thanks.

My car is an 08 997 GT2 with fvd software, catbypass, bmc filter, evoms diverter valves, cup cables, lightweight flywheel and a sachs puck clutch.

 

GT-Silver how do you like the single mass flywheel and did you have to have the computer tuned for it?

 

GT-Silver,
We don't set an LSD up to a certain preload. We set an LSD up to a configuration based on the driver's skills, vehicle, and tracks that they run. In general on the GT2's most people tend to like it with a little less diff. than the GT3's. In particular, a little less lock up on corner exit tends to help keep the car from having traits of snap oversteer. As such, we usually go with a 40/60 ramp set and a moderate preload.

 

For high speed tracks such as Mosport with its long sweepers do you want more or less preload?

 

Please take a look at post #233 for my thoughts on preload. I suspect you guys may be using preload interchangably with the choice of ramp angles. They are two very different things. As a business we don't subscribe to Porsche's high preload model for motorsports LSDs. However, if a customer comes to me and says that he wants a high preload LSD like Porsche does because he thinks it works better, I am not going to argue with him. Ultimately how a team or car owner thinks the LSD works best for their particular configuration is how I'll configure it for them. And then they may still end up playing with it, buying additional thickness internals, and adjusting it several times before they get the LSD doing what they really want it to do.

For you recreational guys I can really only provide you with a baseline set up that it going to offer you good performance, usually for a very wide range of needs. As a gross generalization I would set up an LSD for Mosport with 50/80 ramps. However, if you then take that same set up and try to run Calabogie with it, you're going to be fighting all kinds of push. Some of you will be good enough drivers to drive around it. Many of you won't, but again it's not my place to judge or make the ultimate decision on your behalf. Just like many guys will retune their JRZs from one track to another, pro teams will often do the same thing with our LSDs. But you don't have that luxury and aren't going to pull the LSD apart and switch it over to the 40/60 ramps one weekend and then next month when you're back at Mosport, put it back to 50/80. This is why my bias in this thread for most people is to encourage you to go to 40/60. 40/60 is a better all around set up. It may leave a little bit on the table at Mosport, but will work well everywhere. If you only ever run Mosport or want to be fastest at Mosport then 50/80 is the way to go, or maybe even 80/80. We don't talk much about 80/80 in this thread, but the pro teams running our stuff usually choose between 50/80 or 80/80 and rarely run 40/60 on anything but the midengine cars like Boxsters and Caymans.