Well, I had another one of these sensors fail. OEM part. I did a tear down on a prior one of these I had fail and posted it here but this time, I believe I understand the root cause- the proverbial smoking gun.
It’s a simple design with the conductive elements printed onto a circular circuit board. It’s a dual element potentiometer with both tracks in parallel using two, four fingered wipers for redundancy all around.
I expected to find significant wear around the common dithering point at the center of the arc. It seemed to produce a nice consistent output when viewed through the OBD diagnostics but I suspect the actual software monitoring it is more sensitive to drop outs. Shame on them for not including some software filtering but that is another story.
Looking at the tracks, there is significant erosion of conductive material where the fingers ride. This looks more like a sensor that has 200k+ miles than one that has <80k miles, mostly expressway:
The whitish areas are completely eroded away.
I did some investigating and the remaining material is properly adhered and resistant to scuffing as you’d expect. So the attention turns to the wiper(s). The design of these are equally critical as is the pressure they apply when in use. The general view:
I see something I do not like, let’s zoom in:
Ah HA! Notice there is accumulation of material that has been scraped off- but it’s only on one side of the wiper!
I know exactly what has happened because I have seen similar issues on things like this before. The wiper has been manufactured using a shear to cut out the shape. This leaves a sharp edge on the trailing edge of the shear. This sharp edge will erode the conductive material prematurely. Normally, there will be either an alternate way of cutting out the shape or there will be a post process step to remove the edge. Both of these can be victims of cost cutting later.
I have no doubt that the original design called for the elimination of this sharp edge through some means that was later omitted by the production group. It’s a costly step and if they don’t understand why it was implemented, it appears as low hanging fruit for the CI/Lean guys later.
I would guess either sand/file that edge prior to production assembly. Or the shop/installer tries to, and hope it can be put back together before installation.
It’s too late for this one bing, the resistive material has been scraped off down to the substrate. Had I done this teardown prior to installing the new one, I would have taken the new one apart and chamfered the edges of the wipers to eliminate the problem. It’s not too difficult to open one but there are gears that need to be aligned and the wiper orientation to the housing is important. It comes with a locking pin the hold it in position while installing that you pull out once complete.
Looks like photos of Pluto! … lol … Reliability-wise, requiring mechanical wiper contact like that, seems like a questionable design. Some sort of optical method would be better, similar to shaft-encoders.
Of all of my inexpensive flashlight on/off switches, portable radio volume & station-selection controls, or anything that uses mini-audio-connectors, if more than 2-3 years old, none are reliable b/c of mechanical/electrical contact problems. Nearly all remain useable, but I have to fiddle with them first, and usually more than once.
Compare to my more expensive radios that use optical volume- and station-selection-control methods, all remain working like new.
Yes. The data display on your scan tool is much slower than the input the ECM sees. I’ve run into this where the data display shows no abnormalities but tracing the signal with a labscope shows it’s full of dropouts.
It’s actually a very reliable design until it wears out. Properly fabricated, that should be a long time, maybe even the life of the vehicle depending on use case.
The issue with optical encoding is one of direct versus indirect measurement. The potentiometer is a direct measuring device. Its absolute position is tied to the steering shaft and will survive power loss and changes in angular displacement that occur when the power is off. For example, you disconnect the battery and turn the steering wheel. When power is restored, the potentiometer immediately indicates the new absolute position of the steering wheel.
In contrast, an optical encoder does not store absolute position but can indicate change and direction. So if you disconnect power and turn the wheel, it cannot detect the new position directly when power is restored. It works on a stereo for example because the position was stored electronically before power loss and turning the encoder while power is off does not change the actual tuning position.
That being said, there are ways to accomplish direct measurement but it requires additional electro-mechanical structures in the design. That is more complicated and so more expensive. It also introduces additional failure modes. You’ll see these sensors deployed in autonomous driving applications where the steering is electronic and controlled by computer input/output. Turning the steering wheel will also be indirect, just like the stereo knob and throttle by wire is today.
I totally agree regarding contact oxidation on audio equipment. I rarely use my old equipment and have to fiddle with knobs and switches to clean off the oxidation before they work normally again. But bear in mind, those potentiometers are a different design and inherently less robust than the steering angle sensor is designed to be. Even so, I agree with you, they can never be as good as optical encoders in that application.
Properly fabricated, that should be a long time, maybe even the life of the vehicle depending on use case.