1 2
AngryCorvair
AngryCorvair SuperDork
8/4/11 2:54 p.m.
ransom wrote: The morning, it is early, and my brain, it is... Affected. But... If you added a proportioning valve to get the bias to where the rears didn't lock so early with the ABS unplugged, would that then raise the threshold at which they try to lock and trigger ABS glide mode?

the problem is because there's so little normal force on the inside rear, it takes very little pressure to cause the wheel to lock, like barely enough pressure to actually achieve the desired trail-brake rotation. if the prop were adjusted so the rears did not lock under hard straight-line braking, it would (probably) still lock the inside rear when attempting to trail-brake through a high-lat turn.

tuna55
tuna55 SuperDork
8/4/11 3:07 p.m.

Wow, what a punch of doubting donnas around here. Next you're going to tell me it's impossible to break into the 10s 1/4 mile time for a challenge car.

I don't know anything specific, but years ago they thought factory FI was untunable too. I like the idea of rigging up the rear sensors on the front end of the car. If it gets really hard, you can probably figure out a way to generate the right pulse at the appropriate speed by Ting in the front signal to the rear sensor. Perhaps it's easiest to jumper the pins at the ECU or whatever.

I am sure there is a way. Don'yt change the suspension to cure a silly ABS issue.

ransom
ransom HalfDork
8/4/11 3:13 p.m.

Hrm... How much logic would be required to have the greater of the two rear wheel speeds sent for both rear wheels?

That does involve an actual device and all the reliability concerns that entails. But it does mean it would essentially ignore the inside rear wheel, while continuing to take the weighted rear wheel into account.

I'm still trying to figure out how going "glide mode" when any one wheel goes to zero isn't a really, deeply, truly horrible default behavior...

AngryCorvair
AngryCorvair SuperDork
8/4/11 6:15 p.m.

when you wreck that berkeleyer out on the road somewhere, be sure to return it to OE before the forensic engineers show up. of course I"m a doubting thomas. for the last 20 years i've been an ABS/TCS/ESC engineer in detroit. i know how they work, why they work the way they do, and what can happen when calibrations aren't spot-on. the E36 M3 you guys are suggesting is downright dangerous, and i will tell you that every time it comes up.

ransom
ransom HalfDork
8/4/11 7:11 p.m.

In reply to AngryCorvair:

I think if it were my car, I'd be unplugging the ABS and using a proportioning valve.

That aside, a question primarily for AC: Apart from reliably doing the same thing every time (which is important, I don't mean to dismiss that), is there any sane explanation for the Daewoo's default one-rear-locked behavior?

Wouldn't that essentially mean that if a wheel sensor died, the car would have no brakes? Perhaps the system is at least smart enough that if the sensor dies while off the brakes that it knows that's not lockup and ignores that channel on the next braking event?

Obviously, I'm not going to manage to work through Daewoo's flowchart here. I'm just so completely baffled by this behavior.

Huh... Does your venue give you the chance to lift a rear wheel in both left and right corners? I'm wondering whether there's already a dead sensor or something, so instead of seeing one locked wheel, the system would go from seeing one to two. I'd assume this would throw an error light or something.

And that's probably plenty of shots in the dark for one day...

codrus
codrus New Reader
8/4/11 7:51 p.m.

Is it actually giving you zero brakes, or just a lot of reduction? I don't think I've ever heard of a single-channel ABS system, worst I saw was the mid-90s Neons which used a 2-channel (diagonal corners). If it were a 2-channel system, then by lifting the inside rear you're also losing braking on the outside front, which is probably the tire that's doing 75% of the work so it might feel like gliding even when it isn't.

Streetwiseguy
Streetwiseguy Dork
8/4/11 8:09 p.m.

Your two options are this:

Unplug the ABS computer.

Remove any traces of subtlety and stomp the brake pedal as hard as you can every time you step. If you are gentle on the pedal when the rear quits turning, the pedal pushes back and decreases braking to the other wheels. If you put both feet on the pedal and push like hell, you won't notice it.

AngryCorvair
AngryCorvair SuperDork
8/5/11 7:56 a.m.
ransom wrote: In reply to AngryCorvair: 1I think if it were my car, I'd be unplugging the ABS and using a proportioning valve. 2That aside, a question primarily for AC: Apart from reliably doing the same thing every time (which is important, I don't mean to dismiss that), is there any sane explanation for the Daewoo's default one-rear-locked behavior? 3Wouldn't that essentially mean that if a wheel sensor died, the car would have no brakes? Perhaps the system is at least smart enough that if the sensor dies while off the brakes that it knows that's not lockup and ignores that channel on the next braking event? 4I'm wondering whether there's already a dead sensor or something, so instead of seeing one locked wheel, the system would go from seeing one to two. I'd assume this would throw an error light or something.

i added the bolded numbers above. here are my responses:

  1. Problem with that is if you mute the rear brakes so they don't lock when the inside rear is very lightly loaded, you are underutilizing your rear brakes everywhere else.

  2. Yes, there is. (Disclaimer: First let me say there is no universal ABS software, but there are some universal concepts.) ABS uses the input from the wheel speed sensors to calculate a reference vehicle speed. Each ABS supplier has their own way to calculate this, ie how far into the past do they look, how much weight they put on each of the history values versus the present value, etc. The reference speed is compared to the individual wheel speeds and used by the ABS logic to determine how each channel should be controlled. When one wheel goes to zero, the reference changes and (again each supplier is different, but universally speaking) the control is adjusted accordingly. THere are too many variables for me to expand on this answer, so I won't.

  3. Absolutely not. A wheel speed sensor failure does not impede the function of the foundation brake system. The ABS light in the dash will come on to inform the driver that there is a malfunction of the ABS.

  4. Your assumption of a light being lit when there's a sensor failure is correct.

alfadriver
alfadriver SuperDork
8/5/11 8:23 a.m.
AngryCorvair wrote: when you wreck that berkeleyer out on the road somewhere, be sure to return it to OE before the forensic engineers show up. of course I"m a doubting thomas. for the last 20 years i've been an ABS/TCS/ESC engineer in detroit. i know how they work, why they work the way they do, and what can happen when calibrations aren't spot-on. the E36 M3 you guys are suggesting is downright dangerous, and i will tell you that every time it comes up.

.

Maybe if this goes to page 3, i'll bump this message again.

ransom
ransom HalfDork
8/5/11 8:24 a.m.

In reply to AngryCorvair:

Thanks!

With respect to item 1, I suggest the proportioning valve not to avoid locking the inside rear when it's unloaded, but specifically because the OP says that with ABS unplugged the car has a tendency to lock the rears in general, not just the unloaded inner rear. This is distinct from my earlier question about attempting to reduce the tendency to trigger ABS "glide mode" by reducing rear bias.

WRT to 2, that makes sense. Still seems odd not to ignore a locked wheel as an outlier to the reference speed when the other three wheels suggest the car's doing 50. But I don't expect to understand the whole thing here and now even if I badgered more explanation out of you, so thank you for this; it's more than I knew before.

MrBenjamonkey
MrBenjamonkey HalfDork
8/6/11 4:52 a.m.
codrus wrote: Is it actually giving you zero brakes, or just a lot of reduction? I don't think I've ever heard of a single-channel ABS system, worst I saw was the mid-90s Neons which used a 2-channel (diagonal corners). If it were a 2-channel system, then by lifting the inside rear you're also losing braking on the outside front, which is probably the tire that's doing 75% of the work so it might feel like gliding even when it isn't.

It isn't a complete loss of braking power, it's more like max pulsing. It's probably a 50% reduction in braking power. I think it might be a diagonal system like the Neon, because I've only noticed it when trailbraking into left hand corners.

And yes, my car locks the rear wheels in straight line braking without the ABS, so I think I can cure the straight line lockup with a proportioning valve. If the inside rear locks up when the normal force is 3 pounds, I'll live with that.

TRoglodyte
TRoglodyte HalfDork
8/6/11 11:41 p.m.

I heard a story once about an s 10 blazer that had an ABS problem, It was taken to a computer that told it it was a K5 blazer and its problems disappeared. Computers run on input.

1 2

You'll need to log in to post.

Our Preferred Partners
9JUimvEokEy9m2m8Q7LkUtf8HIRyMBHGttcwroLNWvnJZKJqLm64g82DB3O6Fve5