gregturek

Is this possible, highly unlikely, or just stupid:

My thought is to use brake pressure sensors to detect when ABS is activated and use that as a tool to dial in brake bias adjustments. I'm guessing/assuming the brake pedal pulsing that driver feels when ABS activates is a pressure change in one or both the brake lines from the master cylinder. If I can capture that change with my MXL2 I might be able to "see" on which end of car the brakes are locking first.

In my case (944S2/SP3) I have a brake pressure sensor for the front brakes and one for the rears. I have a bias adjuster "in front" of the rear sensor so it should record the modified pressures. If (big if) I sample fast enough (100Hz?, 500Hz?) is the AIM pressure sensor fast enough to capture the pressure spikes caused by the ABS activating on one end or the other?

Thoughts? Is there a better way to do this?

Thanks,
Greg

ProCoach

Better way is wheel speeds. Then you could do a "slip" math channel to light the LED's.

I've never tried this before...

gregturek

My concern was seeing a notable difference with only a single pulse per rotation? Would I need more granularity? Like the ABS has with 30 or more pulses per rotation?

I currently have one wheel speed sensor on the RR. (It is interesting to see the "speed" difference just from tire deformation differences when accelerating vs. braking when compared to GPS speed.)

Greg

Matt Romanowski

You would not be able to do it with a one "tooth" wheel. I'm guessing you would need 30 or more. I believe some cars are now north of 100, lots 64 or more. Slip is what the ABS uses to figure out when to work the pump. I would have to think about, but I'm not sure if you could do that in the dash with how it will let you calculate math channels. I know you could do it on a MoTeC system.

ProCoach

Matt, you'd need four wheel speeds, but you could build an alarm if the speed on an axle or a corner went out of range, could you not? Easy in MoTeC, with Ground Speed and Drive Speed.

Greg, how does the ABS work if there is only one wheel speed? How does it trigger the pressure release if it can't know which corner (or axle) is slowing more than the others?

The same trigger for the ABS in modern cars is the trigger you would need to build or see to integrate it into the logger, I would think.

BillNye

It sounds like you have an OE abs system w/ wheel speed sensors, but no way to splice into it w/ the aim, hence the single RR wheel speed sensor? Not being an expert in AiM systems, i can't answer specifics, but purely from an instrumentation/signals point of view I think I can offer some insight.
Not sure if you're trying to do this on a quick diagnostic basis for setup or to "permanently" install something and have an in-car readout to adjust bias while driving. For the former, the existing pressure sensors and manual data review would probably be easier, looking for a ripple @ X hz under braking on either channel. However, reliably teaching a computer to recognize that waveform would be another thing.
For the latter, like others said wheel speed is probably your best bet. Do some research on the wheel speed sensors used by the ABS system and what sort of signal those are sending and how to interface that with AiM. Alternatively, looking at some pictures of a 944 ABS pump, it looks like there is some external wires to the solenoids for front and rear that you can tap into. These should give you direct feedback of which axle the ABS activating without any of the data noise from the rest of the braking system. Heck, if you do that you could probably bypass the AiM altogether. Just wire an LED into each of those solenoid circuits.

gregturek

The ABS has 4 wheel sensors with multiple teeth/pulses per rotation.

I only have 1 AIM wheel speed sensor today on the right rear. One pulse per rotation.

gregturek

Thanks Bill,

"It sounds like you have an OE abs system w/ wheel speed sensors, but no way to splice into it w/ the aim, hence the single RR wheel speed sensor?" -- Correct

I'm looking for a method to set the brake bias to insure that the front tires lock up before the rears. I now have the same (larger) size tires on all four corners and added Big Reds to the front this winter. At the very least I now have way more front traction with the bigger tires than the stock tire sizes, so the rears are more likely to lock. Even though the piston sizes are the same on the new calipers the larger rotors might add some brake bias to the fronts. So I'm looking for a method to set the brake bias without flat spotting a set up tires testing, or risking a high speed spin with the ABS disabled.

Thanks,
Greg

"

gregturek

Thanks Bill,

"It sounds like you have an OE abs system w/ wheel speed sensors, but no way to splice into it w/ the aim, hence the single RR wheel speed sensor?" -- Correct

I'm looking for a method to set the brake bias to insure that the front tires lock up before the rears. I now have the same (larger) size tires on all four corners and added Big Reds to the front this winter. At the very least I now have way more front traction with the bigger tires than the stock tire sizes, so the rears are more likely to lock. Even though the piston sizes are the same on the new calipers the larger rotors might add some brake bias to the fronts. So I'm looking for a method to set the brake bias without flat spotting a set up tires testing, or risking a high speed spin with the ABS disabled.

Let me think about monitoring the ABS solenoids. That sounds easier.

Thanks,
Greg

"

924RACR

Your effort to ensure proper biasing is notable and worthy; many screw that up when modifying brake systems on track cars with ABS, ruin the brake performance, and blame it all on the ABS.

For on-track use... yes, monitoring wheel speeds is the correct way to go. If you simply track brake pressures (and brake pressure sensors aren't cheap!), you're not taking into account any erroneous operation of the ABS unit, likewise monitoring solenoid activity. Nor would solenoid/relay activity tell you which end locks first.

You also wouldn't be removing possible influence of the ABS unit on brake proportioning, if it does that (most do, these days).

If I were doing it without wheel speed logging - I'd probably be inclined to pull the ABS fuse and do some straight-line threshold brake applies on a skidpad, service road, or comparable test surface. Of course, you'll want to have some finesse with the brake pedal - maybe do some with ABS first, see where the threshold is for actuating ABS, then once you can just brush that level consistently, do it without ABS and see if the car starts to wiggle as you get close to the limit...

ProCoach

Matt Romanowski

Depending on the rear end, they used to do ABS with 3 wheel speeds. 90's Ford Ranger pickups had 1 wheelspeed off the ring rear that activated the ABS. Certianly not optimal, but could be done. Not sure a Bosch or Teves unit would let you do that now though.

The challenge with the AiM math channels is that we're limited to bias and linear corrector. You could probably do it with a few bias channels, but it wouldn't work as well as it does in a MoTeC dash. The math on ABS gets a little little deep when it's selecting the wheel speeds.

IMHO, for the average DE drivers, save a few sets of worn out tires and then do some braking exercises. You'll quickly learn how to get near lockup, lockup, and how to peal out of it. Lucky for us in the NE (and you this week!), snow is awesome for learning vehicle dynamics with the lower mu.