Ralph, don’t panic yet. I have not have any slip warnings and I hooked up my OBDLink, which can access the ABS ECU. Same 1150.9. I tried the secondary PID and got 1163.7 No change in steering angle reported as I went lock to lock, at least sitting in my garage. Kid’s car is behind me so I’m not going for a drive right now. Point is it may not be an accessible PID on these, or it might be a different PID hex code that is not in the software Toyota is releasing these days. Since this is an early CAN bus for the ABS ECU only, it may not conform to the current standard. And I don’t think there is a place to access analog data. Interesting problem. I’ll let you know if I come up with anything else.

 

Ralph, I think I would look more at the wheel speed sensors. Remember the problems OveurEngineer had with low signal?

 

Thanks, seeing that you don't get a (reported) steering angle change is very helpful!

The wheel sensors are readable through techstream (easier than a 'scope) and a weak sensor will cut out at say 7mph or higher while a good one will read down to 1. Mine are good.

 

I will dig out the bi-directional scanner and see if I can see some life.

 

Same result with the bi-directional.

 

Some additional internet sleuthing reveals that non-responsive steering angle as reported in techstream and other diagnostic tools appears to be a bug that affects at least a few Toyota vehicle types. There's a procedure to probe the sensor and harness, but I don't see the need to do this, yet.

I repeated the ZPC, this time disconnecting the 12V battery first and hitting the brake pedal. All good so far, but I can't declare victory on this one for a while.

update: shit. Never mind. VSC activation problem persists. I guess next step is to try to reproduce the problem while logging live data.

 

I drove the car with techstream running. First surprise was that the steering angle sensor came to life. All four wheel sensors are working well.

I wasn't able to find anything obviously amiss in the data stream. I reproduced my dry pavement VSC activation while recording, no smoking gun.

My -guess- is that I have some minor fault in the yaw/deceleration sensor that isn't bad enough to set a code.

 

This continues to bug me, so I took a drive to a local parking lot, techstream up and running. As far as I can tell, this car has the yaw rate and decelerometer(s) in one package. (edit: McGeorge is wrong... the part in my car is 89183-48010, which McGeorge says does not fit my vin #!). Yaw rate and steering work fine in my parking lot test. The decelerometers (1 & 2) puzzle me. Both appear to read (more or less) lateral acceleration. In a steady state turn left or right I get steady state readings that are very similar on both "channels", and right about where the butt dyno says the lateral acceleration should be. Seems like one of these should be longitudinal. I went through the ZPC procedure again, just for giggles, watching closer this time. Hmmm. Neither decelerometer channel zeros (they are reasonably close, but don't change). The yaw rate sensor did zero the first time I tried this and the offset value changed by 1 degree. I located a salvage unit for short $$. Tempted to just give it a try.

Last time I took a front seat out of a car it didn't require me to defuse a bomb first . I have the factory procedure, it doesn't look horrible, just makes me think about the days when cars were truly easy to work on (and lasted about 1/3 as long, there's that).

 

...and doing a bit more homework, I discovered some conflicting info on the yaw sensor part number... and location. So I decided to have a look. Out came the passenger seat. Hmmm, no wonder I couldn't find the darned thing with aa flashlight and mirror, it ain't there! Then out comes the interior at the passenger side B pillar. Dang! Not there either. So where the hell is it (2006 3MZ AWD)?

 

On my 2008 RX350 (same sensor location as 04-07 HL I believe) I not only had to remove the passenger seat, but also the door sill kick panel, and some tabs holding in the carpet to peel it up a little and reveal the bracket and 10mm bolts holding it in. It was really close to the air vent under the seat

 

Found it, thanks Low. For $25, I'm going to drop a salvage unit in there and see if it solves my problem.

 

Salvage sensor received, installed, zeroed, absolutely no change. The one interesting thing I did discover is that "Deceleration Sensor 1" and Deceleration Sensor 2" do have axes 90 degrees apart... but both are 45 degrees off axis to the centerline. The less interesting thing I found is that the rubber plug in the floor pan has been leaking water (and also on the driver side), so those are now sealed up and carpet pads dried out.

 

I've poked at this a little more from the driver's seat. Here are my observations....

All of the following must be present to trigger the unwanted VSC on dry pavement nuisance:

1. Left hand turn
2. Speed close to 40 mph.
3. lateral acceleration in "sporty" territory, but well short of sliding around,
4. Descending grade. (update -- I've managed to trigger the VSC on an uphill left hander once)
5. Stop light switch on.

I have a collection of four curves where I can reproduce this condition at will. It's very specific.

I've captured a very limited amount of data, but what I've got seems to show that VSC is activating brakes on all wheels except left front.

The vehicle is more or less stock. Quick struts and a Cardone new manufacture steering rack installed two years ago. Tires are new, brakes are all up to date, work great, nothing drags. Alignment is not quite stock. I have crash bolts on all four corners (because QS more or less require this) and I took the opportunity to dial in zero camber at the rear for better tire wear. Front has just a little cross camber so the vehicle sits in right lane with little steering effort. Steering wheel is within a degree or two of perfectly centered. Nothing is moving around on the chassis in an undesirable way. All the bushings, links, control arms are solid.

I've got enough experience with this now that I am quite certain this is just a nuisance, not a safety concern. There are no codes.

I've got a ton of other things I need to work on so going to leave this be for now, just wanted to write up my notes for future reference.

 

I'm still poking at this problem, no resolution. To recap, I have no codes and no dash lights. On certain left hand curves at 35 - 42 mph, brakes applied lightly, the skid indicator and beeper comes on and the ABS system kicks in applying additional braking. The severity of the ABS application depends how fast I'm driving.

I've cleared the memory of the VSC/ABS system and done the zero point cal both via tech stream and the OBD TS + CG jumper procedure, I've even tried detaching the neg battery terminal for 10s first.

In all cases I get the slow blink (memory cleared) then fast blink (ZPC accomplished).

I've even tried re centering my steering wheel by throwing in a half, then a full turn on each tie rod. Somewhere in there I replaced the yaw sensor, no change at all.

The wheel sensors all read good in techstream and on my MX+ down to 1 or 2 mph. They never cut out. I've had the front sensors out and cleaned. The back sensors, unfortunately, can probably not be extracted without destruction at this time (300K miles yesterday!). They've both been out and cleaned, but it's been a couple years.

Assessment: because there are no codes, the system believes it's working normally and actually correcting a skid. There's definitely no skidding going on.

I guess I am back to wheel sensors. They are the least reliable sensors in the system.

I can read the skid control solenoids on my MX+. Maybe I can use this information to figure out which wheel the skid control system thinks is the problem(?). Along those lines, does anyone know what the various solenoids called out in the following PIDs do?

 

Ralph, as I recall the 2004 New Car Features manual has the details of the solenoids. I think it is in the manual too. Should be available on the TIS site or I can look it up when I get my manuals back in about a week.

 

So it looks like my skid control events involve abs braking on all but left front.

 

Hmmmm. Looks like VSC fighting oversteer to the left to me. Which is what you said.
First thought- Remove and reseat the ABS ECU connector. Use some of Ivan’s favorite CRC contact cleaner and Deoxit. A touch of silicone grease on the seal. Make sure the CAN bus termination resistances are good. Just trying to think of things that do not involve the SAS. I’m a little surprised that it does it going uphill at all.

 

Collecting more data. It looks like the order of events goes left rear, then right rear, then right front. In some events (depending on how fast I take the corner) only one or two wheels get triggered. I haven't looked at the ECU connectors, etc. Problems there would normally generate DTCs, but I agree that it deserves a look. It would be nice to get that left rear sensor out for a look too, but that is just not happening... I gave a pretty good try, don't want to break it.

 

This may not be germane to this specific issue but here's an interesting animation on how the VSC computer detects and attempts to correct a skid in a hypothetical situation. In this specific situation, which looks like a moose test, the animation describes the sensors involved, the sequence that the sensors are polled and the braking action that the VSC takes (on a specific wheel, no less) to correct the situation. VSC Explained - YouTube.

 

Ralph, Just some ideas. I’ve been thinking about this. The RF is pulsing, like the ECU is detecting a speed differential between left and right. It is not a smooth brake application like the rear. I wonder if that is normal or if it is a clue. The question I have is which algorithm is operating - ABS or TRAC. If I am thinking about this correctly, if it is ABS then the ECU is detecting impending lock-up and is reducing brake pressure on the RF. If TRAC, then it is detecting a faster rate and applying more force. Seems like some of the other valves are involved. I am back home and will do some research. In either case it seems like a front speed sensor differential. How are your front hubs? Maybe look in to see if there is a problem with the RF tone ring?

 

The RF axle is the only one not OE. It's a Cardone, new manufacture. The tone ring, last I looked, appears less corrosion resistant thatn OE, but not horrible. The RF knuckle is also non OE. It's moog. I elected to replace the entire knuckle assembly when the RF bearing gave out. All the ABS issues post date those repairs. I did have an issue with the RF speed sensor after the knuckle swap. It had low output, was cutting out at about 5 mph, I was getting ABS on approach to stop. I cleaned it all up, reseated the sensor, that problem has not recurred. the data stream from the wheel speed sensors is perfect, no dropouts, no weird values.

When the current skid control issue comes up, the system is applying noticeable additional braking.

 

Another thought. I went back and did some re-reading. You are getting the skid light and warning buzzer. So either the VSC or TRAC is activated.
Is it possible that you have the front end set up close enough to neutral steering that the VSC doesn’t like it? Are you using stock tire pressure?
Maybe Low can comment. He has been working on neutral steering.

 

At long last, this issue appears to be (95% confidence) solved! Root cause appears to have been the worn out spherical bearing I replaced in RR knuckle last weekend.

This is interesting because this vsc thing has been going on a long time. Likewise, that bearing didn't get as worn out as it was in a day. However it's only been in last 5k miles that there has been -obviously- something amiss (rattles and shakes).

Edit: 100% confidence now. I've run my "test track" several times, only one vsc activation. And that one doesn't count because I -was- sliding around (got a little carried away, lol).