Geoffrey

I don't know about the gearbox your friend is using. I was commenting earlier specifically on the G50 based sequential gearbox found in the 996RSR and 997 Cup which was the subject of the discussion.

Cupcar

Thanks, Bill and thanks for putting together the spreadsheet on the ratios as well.

Jeff Lamb

1. Regarding wear -> Thanks for the additional info. I had no idea there were any wear issues with the torque biasing diffs. That is good to know.

2. Regarding adjustability, I am aware of the ability to adjust the clutch type LSD units by changing the ramps, friction discs and shims.

3. Regarding horsepower loss -> I do realize this is somewhat difficult to quantify between the two different types of diffs. However, my race car currently only produces 250hp (and 215 ft-lb torque) to the rear wheels so I need to keep that in mind. It is a 3.4 liter normally aspirated engine with only 10.5:1 compression ratio. I think with some engine upgrades (raise the compression ratio, upgrade the heads and valve train and re-tune) my engine probably can produce something around 300hp to the rear wheels so that will be my goal for 2010 . . . but I have to live with 250hp for the 2009 season.

In summary, I will most likely go with a friction disc LSD unit but I am trying to make sure I at least get a much better understanding of the torque biasing diffs before I make my final decision.

Jeff

Jeff Lamb

Geoffrey, based on your experience, I agree that you know these boxes WAY better than I do. And, I knew when I typed that I could "possibly shift a sequential trans just as fast an ECU controlled trans" that I was probably pushing it. My main goal was just to point out that a sequential trans can be shifted without an ECU control and that the shifts can be very quick. Certainly quicker than shifting a non-sequential trans.

The biggest hurdle I see with the sequential boxes is the cost to buy one. I might (and that is a big "might") be able to get comfortable with the frequent rebuilds . . . but the initial cost of these boxes is incredibly high. The cheapest sequential boxes I have heard of are $20,000 and then there is the Porsche box at $40,000. Ouch.

One other thing -> Thanks for your feedback on the friction disc LSD units. My guess is that the more I learn, the more I will confirm the friction disc LSD as the way to go.

Jeff

Jeff Lamb

Konstantin, pretty much all street car transmissions use helical tooth gears that run quietly and require synchros for shifting from one gear to another. Your friend may have installed straight cut gears into his "H pattern" transmission to enable him to shift without using his clutch. These transmissions that use straight cut gears and dog rings are sometimes called "dog boxes".

Did you hear any gear whine when you were riding in your friends car? If so, that was most likely from the straight cut gears.

Jeff

NineMeister

Torque biasing diffs have a bias ratio (usually around 2.5) which is the multiplying factor that the diff applies to the grip of the slipping tyre to increase the torque on the gripping tyre. For example, if the slipping tyre will break away at 50lbft, the gripping tyre will receive 2.5 x that so it puts 125lbft to the road. That's the good news. When the car hits a kerb and lifts a wheel in the air, the grip on the slipping tyre is zero, whereupon the gripping tyre gets 2.5 x zero which is also zero....

Which is why on the racetrack, if you have a car that lifts wheels or has a large weight transfer on the driven axle you would normally select a plate type differential which is capable of putting 100% of the torque to the gripping tyre when the slipping tyre has zero grip.

Jeff Lamb

Colin, thanks for your feedback. I had read somewhere that the torque biasing diffs basically behave as an open diff if one of the wheels lifts off the ground. And, I agree that is not a good thing. However, if there is any car that would be able to keep both rear wheels on the ground, I would think it would be a 911 due to the rear engine layout. This is why I was thinking a torque biasing diff *might* just work for racing. But, I am anticipating that the more I learn about these things, the more I will find that a torque biasing diff is not the preferred approach.

One other thing that worries me about a torque biasing diff is that I have the impression that the diff might behave unpredictably when under full power through a corner, especially one with any bumps or undulations. With the friction disc LSD, at least I would know that it wouldn't be biasing the torque from one wheel to the other in strangely unpredictable ways!!!

Jeff

964sepp

Jeff,
as I have done some work on my box (6 speed out of a 993), my experience may be of interest for you:
1. It makes a difference if you talk about 964, 993, 996/997. The 964 has the strongest tendency for understeer, 996/997 the lowest - the stronger the understeer, the greater the benefit of a Quaife.
2. As I have a 964 I made my decision for a Quiafe. I also changed ring/pinion to 32/8, so I did not need to change single gears (exception 6th gear, I took a 30/40 from a Turbo, because on the road I like the lower revs, and on track I do not need the 6th gear...)
3. Theresult is convincing: understeer is significantly reduced, turn in is easy and out of the corner you can accelerate earlier
4. Not to forget a disadvantage of the 8/32 ring and pinion: it is loud - I have checked twice if I forgot to fill in oil ...

Not a mainstream opinion, I know, but probably the best for a 964

Jeff Lamb

964sepp, thanks for the feedback. What series do you race in? How long have you been racing? (In other words -> How hard do you push your car?) And, have you raced with a traditional friction disc LSD before switching to the Quaife torque biasing unit?

The reason I ask the above questions is as follows:

1. If you have some good experience racing with a standard LSD and now a torque biasing unit, you certainly would have a very good A to B comparison.

2. I am wondering if you have experienced the lifting of one rear wheel off the ground causing the diff to go "open" when that happens?

Regarding my race car, the chassis is from a 1977 911S and it has the old 930 suspension setup. I am not sure if this setup is more or less susceptible to understeer than a 964 based car?

Jeff

chris walrod

Dog boxes can be shifted w/o SWOL features (ECU Ignition cut / load cell or switch and GPS pot), Atlantic cars only started SWOL in 2003. Before then, they were shifted manually.

From looking through data, we did find some drivers in practice and qualifying were able to shift these things quite quickly, very near a SWOL shift time. However, it was during races where the driver got tired towards the end of the race where damage was being done, shifts missed etc. So in the end, the driver was the biggest variable in the system.

Same thought process goes for when teams (SWOL enabled) pinch the shift-cut time down to minimum times possible in practice, only to find troubles during the race. It was simply attributed to drivers getting tired, and not pulling hard enough on the lever (when race cars used to have levers). We used shift cut switches instead of load cells as they are generally more reliable, however they have a 80ft/lb spring that the driver must overcome before the switch 'makes' so to speak.

Anyhow, sorry for the rambling, but listen to Geoff and Colin, they know what they are talking about!

chris walrod

Oh, and I used to get a kick out of the Barber-Dodge drivers -- they would do what you were mentioning -- pull on the lever, store some energy in that lever and wait for the rev limiter to activate - this was enough to 'release' the dogs and allow a shift event. Keep doing this and you will wear out shift forks and dog rings (maybe even hubs) in a hurry. The Barber-Dodge folks sure did

Ritter v4.0

Where is the trans code located so I can determine which version I have?

Also does anyone have a part number for the 996 Cup LSD- I may have found a lightly used one (less than 500 miles)?

Cupcar

Cup LSD part# 996.332.083.9C

The gearbox code and serial # is stamped on the bottom of the gearbox (as one would see if the car were on a lift)

kos11-12

hi Jef ,
I don't think my friend have any special gears , but here is the answer that I got from Karl Paton from Emeralmd3 that design ECUs and maped his car , here are his words :

I think you're talking about the flat-shift function ? there's a switch on the clutch that tells the ECU that the clutch is down . When the clutch is down and the throtlle is still open (e.g over 75%) the ECU controls the engine rpm with ignition retard and limiters . the engine rpm won't drop below a certain threshold but also can't increase to far either . The result is high air fow through the engine which is also igniting in the exhaust .. this keeps the turbo spinning and producing boost whilst the driver is changing gear . As soon as the next gear is selected and the clutch is realised normal engine power is
restored .
this function works best on turbo cars as it maintains boots during a gear change . It is hard on the gear box / drivetrain . The latest version of the softwer has a new setting that controls how quickly the normal engine power is restored after a flat shift . This damps down the shock of the engine going straigh back to full power when the clutch is realised .
It's good fun though and makes a quite a difference to accelaration through the gears .

sorry if this is out of subject , but I though it is quite interesting ( to share )

Konstantin C4 92

964sepp

hi Jef,
I do club sport races since 2000. How hard? Allways trying not to loose too much ground on GT3s...
Had the standard 5-speed G50 with LSD before. Losing power because of having one leg in the air isn`t a real issue (at least for me). The advantages I get with a Quaife in a 964 clearly outweight that.
I have never driven a F or G model on track, so I cannot tell if and how strong they tend to understeer.
One last thought: I know someone at Audi, what I hear is they have a clear tendency to torque biasing diffs, as long as the price segment, a certain model is in, does allowe that.
good luck
964sepp