Paul (and all other RX-8 owners)...
This is a bit long, but I want you all to understand what's apparently
happening, and I'm not good at keeping a description short and still
conveying enough to make my message clear.
I have been playing around with an apparent random steering/braking problem
on my RX-8. I couldn't put my finger on it for awhile but have finally
determined how to reproduce it.
The symptom...it would apparently, randomly, and intermittently, steer hard
in one direction or the other when first applying the brakes. Sometime it
would go for days without doing it, and other times it would happen several
times during one trip...sometimes to the left, sometimes to the right,
sometimes so light it was not noticeable unless you were looking for it, or
took your hands off the wheel, and sometimes it would pull me over into the
next lane if I wasn't paying attention. Didn't matter if the car was cold or
had been driven several hundred miles. I first noticed it around 1000 miles
(now have over 6000), but, we now know, it was happening before then (more
on that in a minute).
The service department couldn't duplicate it, and I think the mechanic and
the service manager were beginning to think that I didn't know how to drive!
So, two weeks ago, I was determined to find out the cause and be able to
duplicate it for them. I spent the day driving around and applying the
brakes until I was able to find the conditions that caused the problem, and
was able to reproduce it in varying degrees of severity about 1 out of three
times. Enough to take them for a ride and prove it.
Here's what is happening. It appears to be acting as if the electric
steering AND the ABS system are somehow interacting. This is my theory
only, as I have no idea about the programming in the ECU of either of them,
or how they are wired into the car.
The way I reproduce the effect is to apply the brake at the very instant
that the electric steering assist is getting a signal to correct. Then, I'm
guessing, that the correction signal is somehow interrupted by a brake
signal, and the assist then either holds the current position of the front
wheels (that may no longer be straight ahead, due to the depression in the
road), or feeds back a reverse correction signal which actually moves the
steering in the same direction the wheel is trying to go and creating a
severe pulling to that side. Note...trying the same thing with non-RX8s on
the same road at the same spot does NOT produce this effect.
I reproduce this by finding a stretch of road where there is a patch or
shallow crack or depression in the asphalt that goes in the direction of
travel for some distance. Driving with one wheel on the crack or the patch
edge producing a condition where the wheel is trying to be drawn slightly to
one side or the other. The electric assist is correcting that pull and the
car goes straight as long as I stay off the brakes. Applying the brakes
(watch for cars behind you...the sales manager and I almost wiped out one of
their new RX-8s trying this <G>) while the steering assist is in this
condition can produce this pulling effect about one out of three times. As I
said, sometimes severe, sometimes very light. I think the reason it's
intermittent is that applying the brakes depends on a very small time frame
where the assist signal is changing and the signal can be interrupted. The
effect at it's worse, feels exactly like a frozen caliper does.
We tried it with my 1987 RX7, same road, same day...couldn't do it at
all...so we know it was not just the road conditions. We also took out a
brand new Mazda 6 and couldn't do it. But, with a new RX8 they had just
taken off the truck with 1 mile on it, we were able to reproduce it three
out of four times, just as with my car!
One of them was a very severe jerk to the right that pulled us within inches
of a car in the next lane. The service manager said that he almost left a
"load" on the seat of the new car (I was driving <G>)...whatever that means!
<G>
So, we took the info to the Mazda district manager, and he has forwarded it
to Mazda engineers somewhere.
Has anyone else witnessed this effect? If we get a bunch of us with the same
symptoms, they might look a little closer than they would just from my word.
Has anyone else noticed a random and very intermittent, uncommanded turning
of the wheel, very slight most of the time, but sometimes quite severe. It
appears to occur only when the tire is going through a small dip in the road
or a pothole (something that would make the wheel try to turn without a
command from the steering wheel..even for just an instant) at the same time
the brake is applied. If driving over the same spot without applying the
brake, the steering stays straight.
Or is it just happening with the RX-8s here in Rochester?
Or is it just me that the cars dislike? <VBG>
Harley
I was watching for it and it happened to me at least once.
Nothing too serious on my part. Harley, Try turning the auto stability
control off.
For those of you familiar with CPU's it probably has to do with type
of processor being used in the car not being real time. In other words the exact
time it takes the processor to complete an instruction is indeterminate.
The Motorola 68000 had a 32 gig linear address space while the Intel processors
had a "paging" address space. We used to call it by its real name
in the old days. Nasty "bank switching" :) If the program crossed a page
boundary strange things could happen time wise.
The RX8 processor is very powerful with a lot of stuff in there including
two separate network protocols. OBD II and CAN.
Paul Lamar
The AirCraft Rotary Engine NewsLetter. Powered by Linux.
ACRE NL web site. http://home.earthlink.net/~rotaryeng/
Copyright 1998-2003 All world wide rights reserved.