ABS Tone Ring,Sensor or Controller?

[kraichle]

I have a 2003 RX300 that falsely activates the ABS motor very lightly when coming to a stop.  It doesn't start till your down to about 5 to 10 mph and doesn't do it every time.  It makes a chattering kind of sound.

I pulled some waveforms off each sensor as I turned each wheel, and found the left rear to be much different than the other 3.  I was back probing the connector at the controller.  At first the left rear was almost hard to find the waveform in all the noise.  Although, as I rechecked some of the others again and came back to it there seemed to be less noise.  The noise also seemed somewhat intermittent too as you can see in the pics.   I did not turn on the ignition when performing the tests.  When I unplugged the connector from the controller and observed the waveforms, all 4 seemed to have additional noise.  I'm thinking that when connected to the controller there is extra circuity that may help reduce the noise.  The left rear is the only one the noise wasn't reduce when connecting back to the controller.  All the resistances were similar between 1000-1400 ohms, although at first the left rear did seem to have 800 ohms, but didn't see it again.

I looked for a glitch pulse on each of the waveforms which would indicate a cracked or broken tone ring, but did not notice any.  The waveforms were all smooth for at least a complete rotation.  

The only thing I could conclude was because all the sensors looked the same when unplugged from the controller, and only the left rear did not change when plugged in that it was the controller.

I replaced the controller and the problem still exist.

Now I'm left with either a bad sensor, or broken tone ring.  I have to believe I could detect a bad tone ring by the waveforms, so I suspect it's the sensor.

I'm wondering if anyone has observed similar waveforms with a lot of noise and can give me some feedback.  I hate to change the sensor and find it still didn't fix it.

Any feedback would be appreciated.

I attached pics of the Right Front, Left Rear, and a comparison of the Right Front with the Left Rear.  The Right Front I'm assuming is good.

[TrevC]

Hi Kurt...welcome to the Club

Great waveform screenshots!  Good DSO software to capture it all so clearly, most aren't as clear.

The fact that you have sensor output is the important thing...I think the scope would be better with shielded leads as it looks like a whole load of noise being picked up as you say (especially if you are in the vicinity of a metal garage door acting as an aerial, etc).

However, the important thing for a passive analogue sensor is for it to generate an on/off signal for comparison with the other sensors to calculate the individual wheel speeds, which is what is occurring in normal operation.

One thing I would check is the rolling diameter of each wheel...tyre pressures, tyre wear across the axles, tyre sizes being the same. If any of these are out slightly that can cause the symptoms you describe.

Let us know how you get on with it all.

Cheers   Trevor

 

[kraichle]

Thanks for the feedback Trevor, but I find that hard to believe.  The tires are fairly new, and my wife will drive the car with nearly flat tires and the ABS never kicks in then.  I'm pretty anal about keeping the tire pressure exact and even across all tires.

The problem just showed up a month or so ago and if I'm not mistaken, seems to be getting ever so slightly more and more consistent at every stop.

As far as the noise on the waveforms, it's real.  It's a pretty high-end scope I grabbed from work to capture the waveforms.  Initially the noise was so bad (as you can see in the first pic of the left rear), I had to set up a complicated trigger to even catch me spinning the wheel.  Otherwise it just trigger continuously on all the noise.  I repeatedly inspected the connections etc as I didn't think I had a good one based on the noise.  I did 10x the work trying to get a good connection than I did on the other 3.  The other 3 seemed to be consistently nice waveforms.

I'm wondering if there might be a partially broken wire in the sensor loop, but then the ohms reading seemed to be consistent with the other 3 so that kind of debunks that theory too.  As a matter of fact I initially thought I saw it at 800 ohms, which may be pointing at a bad sensor if it seems to change resistance.

I'll go back in tonight and see if I can get some waveforms with the tire spinning a little faster as it's a beast to try spinning the AWD tires by hand.  Although, it'll kind of suck being an AWD as I'll have to lift both a front and rear tire, and back out of the garage so I don't end up in the family room, lol.

Any more advice will still be appreciated.

 

[TrevC]

If the resistances are the same then they sensors should be good unless there is a break in the wire as you say.

The other thing worth checking is the reluctor ring/toothed rotor as this may be cracked or badly corroded...equally, there may be a build up of debris over it or the sensor itself.

It is especially the case with AWD and particularly on the new GEN2 and GEN3 wheel sensors that all rolling diameters are the same...e.g. flat or worn tyres, etc. as this will throw all manner of problems up with transmission and steering/suspension settings.

Let us know how you get on with your additional checks

Cheers Trevor

[pj8708]

Hi Trevor.  good to hear from you!

Hi Kurt and welcome to the LOC.

I know almost nothing about your issue, but that hasn't stopped me before.

Many moons ago, I recall a posting regarding a similar "sound" mystery. I can't find the archived post, but the problem was in the the steering wheel column.  It appeared that the engineers ran out of space under the hood so they just stuck the wiring in a open place.

Also, do the TPM and ABS share the same  frequency

 

Paul

Don't worry about hurting  my feelings,...i had them removed when I took this job.

 

[TrevC]

Hi Paul

It's worth looking at as sometimes faults can manifest from other areas within the car due to noise being generated from other components.

Its certainly worth you looking into the aforementioned thread Kurt.

I think I have also had my feelings removed over the years, but that was by two ex-wives !

Cheers   Trevor

[pj8708]

I know what you mean Trevor. My third has been a charm. This one has been great for 33 years.

Paul

[kraichle]

I took some more waveforms the other night and am still stuck on how much different the LR is compared to the other 3.  I spun the tires using the engine so I could get a faster time/div on the scope to see if there were any glitches over a full 360 deg rotation of each tire.  The waveform is smooth for all and even for the noisy one. The reason I can say this is because as you increase the rotation speed the amplitude of the sensor signal increases too.  With a much larger amplitude I had do adjust the volts/div on the scope to accommodate the whole signal and when I do that it squashes the noise too.  All tone rings look good.

My theory is that there's circuity on the sensor inputs of the controller that automatically scales the signals (similar to what I did with the scope) as rotation speed increases.  Usually they're called AGC circuits for Automatic Gain Control.  With higher speeds and higher amplitude the signal to noise is good, but as you slow down the signal gets lost in the noise and that's when the ABS starts to kick in.

The left rear is the fartherest away from the controller, but it looks like the wire goes over to the passenger side to meet up with the RR and head up to the controller.  It could only be a couple more feet so wouldn't think that much noise would be added.  I also thought that maybe one of the wires got shorted to chassis and was almost sure of it for a while unit I figure out I had an earth ground loop.  Lesson to anyone using a professional lab scope.  Those scopes tie the probe ground reference to earth ground.  Believe it or not I had made a connection from one side of the sensor that the probe ground was connected to, through the scopes earth ground to the wall outlet, then back through my trouble lights earth ground tied to the metal bulb shield with a hook on it that I had hanging on the glovebox metal latch hooky thing.  when I used a DMM to measure the sensor leads to chassis, it followed that funky path and F'd up my brain for a while.  After I figured that out I found I had full Meg ohms to chassis, so no chassis shorts.

I decided to buy a $20 ABS sensor to test outside of the hub to see if the noise is still there.

I will look into your suggestion Paul about the routing around the steering column, but not sure why that wouldn't affect more than just the Left Rear.

Thanks for you support guys.  Why do I always have to get the freaking out of ordinary problems.  I should tell you my Town & Country torque converter lock problem someday.

Kurt

 

[kraichle]

OK, I'm guessing maybe my problem can be added to the majority statistic of what fails when the ABS is falsely activated.  I believe the Right Rear is the problem.  

Take a look at the waveforms attached and let me know if you'all think this is a typical Tone Ring failure.  

I didn't catch it earlier as I was looking for something more obvious, although maybe since the ABS was barely being activated I should have been more thorough. Another thing that made it easier to see was to use the engine to spin the tires.  There's no way to rotate the tire fast enough or consistent enough to see a good waveform.

If you think this is typical and needs to be replace, is the best way to fix it, is to replace the whole driveshaft?

Thanks

Kurt

Note, my previous noise issues had to do with not using a fully isolated differential probe.  Seems like if I alternated which side of the sensor I connected the scope ground would either give me a clean signal or one with noise.  Crazy though, as both sides of the sensor seemed to be isolated from the cars chassis (GND).  Oh well, I don't need to figure out the stupid noise thing if the problem is the RR tone ring.  Maybe someday when I get bored I will.....NOT.

[TrevC]

The extended duration could be a crack in the ring which is longer between the teeth (much like a missing tooth on a flywheel or crank sensor) and if you can separate the ring from the end of the driveshaft and source another ring (eBay) it is just a case of removing the old one and fitting the new one.

[kraichle]

Good answer Trevor, it was a cracked tone ring, and yes the crack extended the distance between two teeth which is evident in the scope waveform.  The trick is spinning the tire faster using the engine so the Time/Div is set smaller so it stretches the waveform out exposing the diff in spacing.  It's harder to see if you have to set the scope Time/Div to a larger number.

I got an axle on ebay for $55 and it took couple hours to change it out.  Fought more with getting bolts loose to swing out the hub than actually changing the axle.

Thanks for all your feedback.

Kurt