MFGJR

I'm looking to understand excessive traction control intervention on corner exit when I was last on track with a new-to-me E46 BMW 330, i.e., what was I doing to trigger the intervention. But, I first want to rule out a malfunction in the ABS/traction/stability control unit.

Looking at wheel speeds from my Solo DL has me confused--rear wheel speed is consistently higher than the front, yet the front and rear wheel speeds come together whenever the throttle position goes to zero. Here's a snippet showing the left side wheel speeds (right side is the same) on a straight with no steering angle. The blue trace is front wheel speed, red is rear wheel speed, and black is throttle position. The first two lifts are shifts, the third is braking.



Track tires are square 245-40-17. The factory setup is staggered 225-40-18 front and 255-35-18 rear, with only a nominal difference in circumference, so I don't think the intervention problem is rooted in the system being calibrated to different front/rear tire sizes. And that wouldn't explain what happens to the wheel speeds when the throttle is lifted. Note this is a low-powered car and it's definitely not wheel spin that we see in the data.

Any rational explanation for this?

Thanks.

boxer-11

I've spent some time in a Spec E46 race car which is basically an E46 330i -- certainly it still has the DSC from the street car in it. First thing you do on the track in that car is turn off the DSC. If you don't it will just rob you of throttle coming out of corners and generally eat your rear brake pads while making the car feel unstable. And you don't really seem to need to be pushing very hard to trigger those effects based on what I've felt when I forget to turn it off (or it comes on again which does happen sometimes)...within a corner or two you'll immediately feel "why am I going so slow out of this corner??" and/or "why is the car lurching through this corner so much??"...oh, yeah, DSC is still on -- DOH!!

What I was told is that it's just not tuned for the track and when it sees more than a little slip angle it assumes you are doing something too dangerous for the street and takes corrective action.

I haven't looked at AIM data from that kind of a car but based on my experience with the same kind of thing in Porsche cars, you can pretty clearly see the stability control actions if you look at all four wheel speed channels on the same chart -- as an example you can see it grab the brake caliper on one wheel -- a spike down in that wheel's speed trace -- like a rear wheel when it wants to catch oversteer for example. I suspect the red/blue difference you are seeing just comes down to slip on the driven wheels when there is power applied versus no slip (from power, not steering) when you lift off the throttle.

Pedal position is just your drive-by-wire suggestion to the car. As far as throttle intervention by DSC you can't see it in that trace -- you'd have to look for unusual slope in your RPM trace versus unmolested straight line acceleration slopes or something I think.

No doubt Peter and/or Matt will be along with more complete answers shortly!

MFGJR

Mark--thanks for the thorough reply.

Roger that, I developed pretty bad rear brake problems. Just trying to understand exactly what the car was responding to, and from there see if I can change something up to make it happier.

There are data channels for the traction/stability control, and I'm able to isolate stability control from traction control from ABS. It's the traction control that's the issue on pretty much every corner exit. (And even when asking, seems to me, little of the car...) I'm having a hard time seeing the intervention in the wheel speed traces. What you're saying about slip on the rear wheels under power makes sense, but I wonder if 2+ mph of front/rear difference at over 100 mph is reasonable. Data from my 997 show no more than 0.4 mph of difference.

I'll post some data showing the traction control intervention this evening.

924RACR

That data looks normal/reasonable/legit to me. Wheel speeds look correctly compensated for tire size variation.

(from a guy who looks at racetrack TCS data all day long on cars with well north of 500hp)

Let's see some cornering TCS data - all 4 wheel speeds, engine torque delivered, TCS torque request, throttle position, lat g's.

MFGJR

924RACR,

Thanks for looking at what my "half of 500hp"car is doing! Or what I'm doing to it...

The track was VIR, and here's a representative lap showing where the traction control was intervening:



Here's the measures graph displaying traction control ("TC") intervention (0/1= off/on), engine moment (torque, %) and pedal position (%)--note I don't get actual throttle butterfly position:



Here's traction control and each wheel speed:



Here's traction control, G-sum (don't laugh) and lat-G:



And here's traction control with a combined throttle/brake channel (throttle >0, brake <0, downshift blips suppressed) and steering angle:



Happy to zoom in, overlay select channels, or other provide more specific data.

Frank

MFGJR

Duplicate post

924RACR

Dang, that sucker looks like it's tuned pretty tight. This must be normal key-up mode, not any sort of Sport mode? How about tires? I'm a little surprised to see that, on track, you're not really even hitting 1.0g lat? Maybe just brushing up against it, briefly, peak values? Seems like you're maybe on very much a street tire, or are you not getting everything out of a more track-oriented tire?

The reason I ask - and it may not help here anyway - but usually when the systems "read" the surface (in more technical terms, estimate the mue), they can only work with what you give them. If you're mashing the pedal but only pulling 0.6g (as I see in the one cursor point), you may actually legit be on a wet asphalt surface, in the cool, and it'll be tuned tighter for that region. So committing more to full max lat (0.8g+ for a street tire, even) and then going to the gas might help the system work to a higher operating point, and allow more acceleration.

But if you're still in a basic normal street mode, the tune may still be conservative, more than useful on the track.

I don't see anything incorrect in the data you've shown, on a high-level review - just doesn't look like it's tuned for the track.

So if I were you, even more than changing any technique, I'd be looking for either a Sport mode or Off. Which, honestly, to keep some degree of safety net, you might even be fine for now with JUST turning off TCS, NOT ESC. ESC is usually going to allow you to play with the car a bit, and as long as you keep your steering inputs smooth and avoid countersteering, you'll be impressed by just how much freedom the system will allow before getting involved. As opposed to TCS, which will typically be tuned to be earlier, and give smooth, refined, but conservative control.

But by leaving ESC on, you'll still have the real safety net in place to keep your car shiny, while allowing you to more fully explore the grip available from the tires and chassis...

Hope that helps...

MFGJR

Definitely not a race car, just me doing the occasional HPDE and looking for how best to drive the car I brought. Tires a 200TW Toyo R1Rs (fall between a street tire and an R-comp), stock-ish suspension, open differential. Track was dry, ambient temps were warm.

I share your confusion about the traction control intervention at such low lateral Gs and with such low HP. It’s almost like the traction control is anticipating, rather than reacting.

No Sport mode option. My only choices are to either (1) disable the DSC (over/understeer correction) and slightly raise the traction control threshold, or (2) disable everything.

I’ll definitely turn everything off next time. But the question is will I then avoid a bunch of unnecessary traction control intervention, or will I instead get a bunch of inside wheel spin? We’ll find out. If the latter, I guess later apexes and/or more progressive throttle application would be order of the day…

ledbette

Are you getting TC directly from the CANbus via SoloDL, or is this a computed value?

MFGJR

From the CANbus.