Inside the Gear Position Switch

[Icecream]

Was doing some maintenance and found some surprises on the way, one of which was my GPS sensor. It was giving me trouble (compounded by a loose actuator) and noticed it was hanging up when sliding the rod in and out, so I decided to get a new one and rip this one open.

Back cover removed. My ham fistedness and impatience cracked the sensors at the edge, so much for fixing this one.


PCB where the magic happens. At first I lost the rear pcb so the only way I could make sense of the rotational part of the sensor was some kind of inductive sensor for that part and a standard pot for the horizontal part, but, I found some more pieces and the rear plastic has brushes I didn't pay attention to before.







Lots of dust on the contacts those four brushes glide on. I had already wiped them when I took the picture but they were covered in plastic dust from the carriers sliding back and forth. the plastic is very dry and very tight fitting. My finger came out covered in ultra fine dust after wiping the contacts. I think if you drilled a hole into the housing so you don't hit anything, flushed it with alcohol (maybe even compressed air alone) and lubed/sprayed it with deoxit, it may get them working again. The deoxit may even be enough to keep the plastic lubed so it doesn't wear out again and contaminate the surfaces. I'd experiment but I'm kinda over it at this point so. One thing is for certain, nothing inside warrants a $600 part lol. It seems like about $20 worth of parts inside unless I am missing something.

[sapote]

1) This is the linear and rotational sensor for the SMG shift rod. I wouldn't call it GPS because it is reserved for the screwed in sensor on the manual box that one might need to add when converting from SMG to 6mt.
2) no magic about magnetic field or such; just the simple wipers and tracks or close or open circuit to determine which planes, 5/6, 3/4,...to R and odd or even gears.

How many wires on the connector?

[heinzboehmer]

2) no magic about magnetic field or such; just the simple wipers and tracks or close or open circuit to determine which planes, 5/6, 3/4,...to R and odd or even gears.

Yup! Although to me that looks like it serves the same exact function as a potentiometer (or multiple rather) just that the wiper travels linearly instead of rotationally. Those tracks look to be made out of the same material that tracks on potentiometers are. Guessing it spits out an analog value about the location of the actuator/forks inside the trans instead of a digital value telling what gear the trans is in.

Any pics of the harness? Just by looking at it that seems like an analog sensor so there shouldn't be any sort of processing going on at the sensor. All that would happen at the SMG ECU.

Cool stuff though. Seems like you got it open destructively. Any way to get it open non-destructively so that it can be cleaned and serviced?

[Icecream]

2) no magic about magnetic field or such; just the simple wipers and tracks or close or open circuit to determine which planes, 5/6, 3/4,...to R and odd or even gears.

I was being sarcastic when I said magic

Yup! Although to me that looks like it serves the same exact function as a potentiometer (or multiple rather) just that the wiper travels linearly instead of rotationally. Those tracks look to be made out of the same material that tracks on potentiometers are. Guessing it spits out an analog value about the location of the actuator/forks inside the trans instead of a digital value telling what gear the trans is in.

Any pics of the harness? Just by looking at it that seems like an analog sensor so there shouldn't be any sort of processing going on at the sensor. All that would happen at the SMG ECU.

Cool stuff though. Seems like you got it open destructively. Any way to get it open non-destructively so that it can be cleaned and serviced?

Since I destroyed it and tossed some parts I couldn't see if the lower carrier was attached to anything, wasn't obvious that it did, nor could I tell how the copper piece was attached. So I looked around and found the missing pieces to put it together and realized there were two PCBs, one that rotates on the back and one horizontal.

As for taking it apart without destroying it? It is glued and potted on the other end so it won't come apart easily and the PCBs are very brittle. I think the best way is to go through the side, at the midpoint 1/2" from the flat bottom with the rod fully extended. If you miss you will hit the copper plate (probably not fatal) or the pcb (definitely fatal). Go deep enough to just get through the black housing and internal ribs no more than 5/16". A 1/4" to 3/8" is probably wide enough to get a good flush, the PCB would be visible from this hole (sitting just below it).

[sapote]

Are there 6 wires on the connector?
4 wires for each conductors strips to indicate which strip is touched by the rotary wiper which is connected to a ground wire and a 6th wire for a linear pot for positioning the odd/even gear.

Only the linear pot signal is analog and the other signals are digital to indicate which gear plane is selected.

[Icecream]

Are there 6 wires on the connector?
4 wires for each conductors strips to indicate which strip is touched by the rotary wiper which is connected to a ground wire and a 6th wire for a linear pot for positioning the odd/even gear.

Only the linear pot signal is analog and the other signals are digital to indicate which gear plane is selected.

There are 4 each. I think it just adapts by reading the values for pot 1 and 2 and stores that value for that gear, that's all the information it needs to verify a gear after the fact. When the actuator moves to 1st gear for example, pot 1 can be a value of 2v and pot 2 can be 3v, so long as the pots are healthy, those values won't ever change so when you shift to 1st gear while driving all it needs to do is verify pot 1 is 2v and pot 2 is at 3v (or whatever the values are that it stored during adaptation) to make sure the correct gear was engaged, unless the contacts get contaminated of course. I think this is why DIS warns you not to adapt with faults stored. The GPS is a dummy and has no idea what plane or gear anything is ever in, it moves and stores a value to reference to later.

[sapote]

Ok, so only 4 wires connected. Assuming each wire connected to each of the 4 tracks and each track is a resistor pot with a single wiper making contact to one track at a time corresponding to which gear planes (R, 1/2,3/4,5/6), the value of the selected track will be in the range [0:1000] Ohms, and the not selected tracks have opened circuit of Infinity. This would work but why we see there are 4 wipers for the 4 tracks?
And how the selecting wiper connected if only 4 wires are used?

[heinzboehmer]

This would work but why we see there are 4 wipers for the 4 tracks?

It looks like the left two and right two wipers are electrically connected. So left two tracks get swept by a pair of wires and right two get swept by the other pair.