Larry Herman

Darren, I had occasional problems with this "ice mode" in my RSA. It was most prominent when first getting on the brakes from a very high when the track is falling away from the car, like before turns 1 and 8 at the Glen. I found that upping the low speed compression for the front shocks helped because it allowed for a faster initial transfer of weight to the front tires when you get off of the gas. I also had the Cervelli ABS cutoff switch, and had a few occasions to use it.

As far as the Steering Angle sensor, I wonder if you could replace it with a resistor to provide a constant reading of no load so that it never thinks that you are cornering. Just something that I might try if I were experiencing your problems.

bobt993

^^^^^^^^^^^^^^ What's the fun in that Larry? Darren loves new project and cool car stuff.

onefastviking

Yes, I've seen it in a few without PASM.

onefastviking

Best way to describe it that I know of.

924RACR

Yeah, good luck with that!! LOL

It'd never work. It's far more complex a sensor than that. You'd be better off just unplugging it, so that it's always off and you don't have to deal with the system changing performance while on-track.

More to the point, it's not likely the steering angle sensor; rather, it'd be a matter of trying to reduce the gain on the lateral g sensor, by a factor of 2/3rds based on my knowledge (figuring you're likely pushing to the range of 1.6-2.0g peak values, and that threshold would typically be set around 1.6g or so on such a car). However that could still cause problems with plausibility models at lower lats comparing to steering input and yaw. And it would REALLY such driving around with PSM on - you'd really start getting hits ALL over the place. AND most likely in those later cars, it's a CAN-based message that you'd have to intercept with a CAN box and re-scale... and and and...

IOW - so not F'in worth it.

Maybe this explanation about how ABS systems work, in a small area of detail, will help with what you guys are calling "ice" mode...

OK, so simply put, it's not so much that there's a "mode" that you get locked into.

The thing is, every time you get on the brakes, on an ABS system that has to be able to see street use (which implies winter use including on ice) - the system must assume worst case, that you are on ice, for the initial operation, and then, once we can confirm you're on a stickier (higher-mu, mu being the coefficient of surface friction) surface, we can adjust control consequently.

Actually, this is a little more significant for PSM (ESC) applications than even ABS; this is also why a Race ABS like the fine Bosch Motorsport units (shameless plug! ) is preferable - it doesn't have to worry about ice, snow, and gravel, just wet and dry pavement.

But remember - ABS and PSM/ESC are SAFETY SYSTEMS. These are not race performance systems. You want those, you need to buy a Corvette ('nuther shameless plug! ). So the system must first ensure safety and stability; then performance comes along.

So how do we determine what kind of surface you're on?

Well, simply, in the case of ABS, you can look at the wheel dynamics as pressure is applied. Obviously, if the wheels decelerate very quickly (while the driver is ramping in pressure), then we're on a slipperier surface (lower mu). If the wheels develop slip (go to lock) more slowly, then the mu (surface grip level) must be higher.

This surface friction estimation then goes into and affects the rest of the calculations for control during the rest of the stop; initial pressure decrease size, then consequently the holds and pressure ramps and decreases through control till the vehicle stops or the driver gets off the brake.

Now, as the system must continue to adjust, it is possible for it to detect what you might consider an underbraked case (where wheel pressure is low, and the vehicle could stop more quickly). Under normal situations - say, you hit a spot of gravel, ice etc on the street, that causes one or more wheels to slip deeply - this will happen; you'll have a moment of lower decel, while the system recovers the wheels that started to lock, then it detects the higher mu again and ramps pressure back up to stop the car at a faster rate (higher decel).

HOWEVER.

Your cars are so far modified from the systems as delivered, with brake mods, wheel and tire mods etc... that the wheel dynamics and inertias are far from what the ABS system was initially tuned with. This can make both the initial slip/surface detection and follow-up control (including pressure recovery) suspect to errors.

This is why a pure-race ABS system, like the Bosch unit (which allows both initial setup for brake system hydraulic and wheel sizing, and tuneability on-the fly in the car), is so much better for your applications. It will be installed and calibrated specifically to how your car is now, vs. what it was originally built with by Porsche. This makes the algorithms effective once again, as intended.

This is ALSO why you find a notable difference between hammering the pedal hard vs. ramping pressure in smoothly. When you give the brakes what I like to refer as the "high hard one" - fast, hard, spike apply - the wheels go quickly to lock, and the ABS is more prone to think it's on a lower-mu surface. When you squeeze into the brakes, and give the wheel a little more time to develop slip, this is less likely to cause underestimation of the surface grip/mu available.

Unfortunately, this is not really the ideal way to brake an ABS-equipped car on the racetrack; the faster you get to locking pressure, the later you can brake and the shorter the stopping distance. If your cars were stock, you should be able to brake like that.

So clearly - the cheapest solution will be to learn to threshold brake.

Hope that helps at least to explain what's going on...

Darren

Wow, thanks for that! Fascinating info!

jscott82

Yet one more person I add to my list of folks I'd love to have a beer with....

Like they say sometimes, you don’t know what you don’t know... I had no idea ABS did all of that (estimations etc).

924RACR

Just the tip of the iceberg, really! After all, there's a very good reason you can't just slap on some generic electronics and think it'll all work and be safe...

About the only thing uglier than locking the driver out of all four brakes (corners, I mean) is grabbing one rear wheel brake without reason... which is why there's an incredible amount of code, development, and testing behind all that ABS and ESC technology...

Larry Herman

Thanks for the info Vaughn. With a system so complex, I should have known that there was not a simple work-around for it.

kurt M

+1! Thanks and please continue with more on this as you see fit.

trackjunky

This may be displaying my stupidity, but why can't someone just write a code for the "ice mode" that changes the G levels to be more consistent to what you'd see for track use and flash the ABS computer?

Sort of like an ECU flash.

onefastviking

They have, but they have written it for the other two systems previously mentioned.
Also many of the 996 models are 3 channel systems which wouldn't give you the desired results. I am sure 924RACR can elaborate much better than I on this, I know these systems at a much more basic level that his expertise.

Terry L

Darren - when I had this problem in my Lotus Elise, the pedal was high and hard as a brick - it would sometimes drop to normal position at the last second, which kept me out of the kitty litter. It didn't give an initial tug, like it did when it was working right, just at the very end of brake actuation (or at least that is what it felt like to me.) How does your pedal react when the problem occurs? Is it like a rock from initial application or does it take a split second to cut in?

Darren

At first the brakes feel fine and then after maybe a second, or what seems like a second, the pedal comes up a little and then its rock hard but the car won't stop very well.

924Racr's explanation was that the wheels are slowing down faster than the ABS system thinks they should be able to (because of bigger brakes and slick tires) and then it misinterprets the road surface as being extra slippery so it "helps" me out by reducing my braking force.

J richard

How many of you guys think the benefits of ABS outweight the potential of no brakes at all when you need them? I've always found that if I was into the ABS I was way beyond threshold and the only thing the abs saves me is a flatspotted tire (which is a nice feature) but as far as performance gain? Worth much? anything? I had a spin that cause the ABS to let go and found myself rolling back onto the track backwards with no brakes, after that the overrides were put in both cars, I even considered wiring it to a position switch on the clutch and brake, both feet in hard to the stops disables the ABS...but I am still considering yanking the whole thing and saving the weight. I know its not the proper "engineered" solution, but sliding off the track into a wall I'm into practicality...and then theres the practicality of a $5-10K...

Another thought is a tire is at its most efficient with a certain slip angle even moreso with a race tire, and that is true in braking as well, how much are the street abs systems giving up ultimate capacity by releasing a wheel that could be at an optimum slip angle for the tire, but exceeding the parameters of the ABS settings? and thinking in FC terms combinations of lateral and rotational slip angles could make things worse...