Potentiometer design

[3DDave]

I'm looking for a link to describe how the construction of a GM throttle position potentiometer works.

I came across this in relation to "Reduced Engine Power" because there were a lot of videos about the engine being artificially crippled by the software in response to some problem. A lot of people replaced the throttle body, some found wiring to the throttle body had a broken wire, bad ground, et al.

One of the iterations has a 4 wire potentiometer - +V, Ground, Pot 1, Pot 2. If there is a disagreement between Pot 1 and Pot 2 voltage the software cripples the engine. Another iteration uses 6 wires, 2X V+, 2X Ground, Pot 1, Pot 2, so that losing a ground wire or a V+ wire can allow the computer to continue in "Reduced Engine Power Mode." I think the 4 wire is a cost-reduced version of the 6-wire.

My interest is in the 4 wire version, which seems like it should be simple, except the wiper doesn't have an external connection and the layout of the wiper has 4 segments. The video that shows a teardown is:

https://youtu.be/C5BEPmFB6cw?t=113

The view is only for a short time and it required me to freeze it pretty often to see that there are four tracks of resistor. I had to set the video to 720P. This is the clearest spot I can get is just after

https://youtu.be/C5BEPmFB6cw?t=118

(YouTube doesn't capture a frame #)

The first link should start about 1 minute in and the author moves the unit around to show a little bit, but not a 100% clear view of the printed resistors of which it looks like there are 4 segments. It looks like there are also two smaller resistors, possibly for matching purposes. Near the end of the video he shows the minuscule brushes, of which one has what look to be defective finger.

What I am trying to understand is how the brushes pick off the voltage and return the signal. I see gold arcs but it seems unlikely the brushes run on them. I think part of the layout is to put each short arc of resistor in series with a long one so that the difference in radii and therefore resistance difference is cancelled out, allowing each path the have the same overall resistance.

I've spent a half a day looking for patents on redundant potentiometers, for potentiometer design, et al, but I don't see this configuration, The main sticking point is that if the gold arcs don't return the voltage, and the brushes don't have leads, then how does the voltage get back, and if the gold arcs do return the voltage how is voltage applied to the segments?

 

[roydm]

Many of the newer Throttle Position Sensors are wiper-less Hall Effect type, these should be more reliable than the old potentiometer type. BTW they are not a simple potentiometer they have extra resistors in there to provide a fail value if the wiper goes open.

 

[Compositepro]

I will also point out that it is common in many variable resistor applications, like a volume control, that the wiper contacts are electrically connected to one end of the resistor. That way loss of wiper contact does not cause an open circuit but max resistance instead.

 

[jraef]

I had an issue with my Ford earlier this year and in researching it, discovered that they are now using an absolute encoder (or maybe a resolver), not a pot of any sort. The "Reduced Engine Power" mode that I was forced into was not related to a failure of the sensor from the throttle, it was from a failure inside fuel injection system, which basically had the computer ignore the throttle information and put me in that "Reduced Engine Power" mode of creeping along at 7MPH max. with no response from the throttle.

My wife had the throttle go out on her Prius a couple of years ago and it was a pot, and I was able to replace it by finding out the values and adapting a COTS pot to the mechanism, it wasn't that hard, and the one Toyota waned her to buy was $400! She still drives it on mine, cost me about $10 and maybe 8 hours of my life. But in the discussion, I heard that the newer "Prii" are using the Hall Effect versions now, so "bush fixes" will be a bit more difficult.


"You measure the size of the accomplishment by the obstacles you had to overcome to reach your goals" -- Booker T. Washington

 

[IRstuff]

But, the Hall sensors shouldn't wear out, so there's that.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[roydm]

The Hall Effect sensors I have are compatible with the old pots, besides they seem to be cheaper.

 

[itsmoked]

Pots of any kind are just so wrong for accelerator pedals only noob idiots would consider such a wear prone, problematic, flaky device, in that application. Throttles should all be non-contact and non optical. Either LVDT or use one of the many TI inductance-to-digital converters.

TI Inductance Converter Brochure

Keith Cress
kcress -

http://www.flaminsystems.com

 

[VEBill]

"...only noob idiots would consider..."

You spelled many major automobile manufacturers incorrectly. Same thing of course.

 

[ScottyUK]

Cheapest solution which lasts longer than the warranty period will always be selected over best available technology.

 

[MacGyverS2000]

The reason they want you to replace the entire throttle body when the TPS goes out is it's an emissions item (BTDT). $0.25 part goes out, but it costs $400 to replace the (included) complicated TB.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[3DDave]

I think these throttle bodies are around $70 and 30 minutes labor. A lot of the problems are actually in the wiring harness leading to the throttle body.

I was just hoping someone had come across one of the GM 4 wire throttle bodies and looked at how the potentiometers were built. For example, setting up the tracks to account for the radius differences to make the two resistor paths the same or if they even they are the same.

Lots of other sensor alternatives, but this is an interesting puzzle.