Checking throttle position sensor

Choose the one which has a 'smoother' resistance curve. (No erratic readings or sharp transitions over the full range).

JB

Agreed, the absolute value is only of relative importance - you want to make sure the pot is mechanically sound and doesnt provide spiked output. Id go with TPS 1 as its values are mid-range - TPS 2 has higher values, indicative of greater wear on the wipers and this more likely to have erratic output (though both should be checked to be sure) and more likely to fail again in the near future.

HTH J

Those are only rough checks. You really need to know the voltage (and spec) it delivers to the ECU under both conditions.

I'm not sure about your system but on mine you adjust the TPS to give a specific voltage output at idle.

Essentially the same thing with a pot, assuming all other connections are correct and theres no high resistance joints due to corrosion, the voltage is proportional to the pot resistance.

Yes, but measuring the resistance with the pot at either end of its travel isn't that reliable since it depends on mechanical things.

Its as reliable as measuring the voltage out with the pot at its extremes. If you want to do it properly you'd get a digi 'scope and trace the voltage out at the ECU pin during operating conditions and check the output was withing specs and not spikey but its not practicable for a normal DIY person :-) Also a lot of manufacturers dont give the voltage specs, only the resistance. Therefore watching a smooth sweep of the wiper across the full range will give the best indication for Joe Normal.

Thanks for the replies, both resistance readings seemed smooth during the opening throttle test but I've gone with #1 sensor which had lower readings (and I think it came off a lower mileage car) but will keep the other for spare. I don't think this is the kind of tps that can be adjusted separately on the throttle body shaft as the sensor can not move with the 2 screws loosened. I've give it a quick test and all seems fine.

If the pot is non adjustable and fitted in place I'd agree. However, it's unlikely to be right at the end of its travel when fitted if you think about it.

Easy way is to wire it up as a volume control and feed audio through it.

Some TPS's use the end of the slider as the idle location, a point of zero output (or max), they arent normal pots like you buy in farnell.

Again, not exactly an easy proceedure for a DIY person who may have no electronics background. Very good method of inrregularity detection though, good thinking - it would show up loud and clear especially with a single tone output.

I've not seen that type. It would have to be pretty meaty to act as a throttle stop.

Funnily, RS do a suitable replacement for the early Rover EFI engines where the TPS is NLA although you have to make an adaptor plate.

The message from "Dave Plowman (News)" contains these words:

It's a shame they don't use optical shaft encoders which'd be far more reliable and probably these days not a lot more expensive.

Trouble is a shaft encoder requires electronics to give a usable analogue output while a pot gives a straightforward voltage output which varies with position.

I agree about shaft encoders, though. The Calrec sound desks at Teddington Studios use them and are still working at about 20 years old. Desks with analogue pots are likely to be clapped out beyond economic repair in less than half that time.

Which is then digitised as soon as it hits the ECU. I would be astonished if any modern (>1980?) ECU used the voltage directly.

(there are problems with this - an optical encoder needs at least 3 wires in addition to power and ground. (zero position, and quadrature), for a simple encoder with no intelligence)

I'm in the process of using several USB mice to read various car parameters, from wheel rotations to throttle position. Each 2.99 mouse has 3 rotary encoders, and 3 buttons.

The Rover EFI units did until into the '90s.

The message from "Dave Plowman (News)" contains these words:

But you don't want an analogue output - it's going into the ECU where you can bet it's converted to digital.

The message from Ian Stirling contains these words:

You can do it with fewer if you send it serial! Rather than a slotted opto I'd use a Gray-code. Five bit would be enough - perhaps six. Built onto a PCB with a row of LEDs reflecting off a mirror behind the encoded sector. No need for zero position or quadrature then. Three wires would be fine - supply, ground and signal. It'd have one moving part, no contacts or tracks to wear, nothing to get out of alignment. A lost-motion arrangement on the connection to the pedal would take care of the foot off position not being the same as the zero output from the encoder.

In the context of this thread, you do.

I'm not sure most modern set ups have a real TPS anyway. Some just a switch for idle and maybe one for full throttle.

Modern systems are entirely a variable resistance actually- though the ECU tends to use this more for degrees of accel enrichment, WOT and idle position. Actual load is measured by the air mass meter, or MAP.

We're all going to DBW throttles now anyhow, troublesome as they are!

Tim..

Or use a 1-wire (+ ground) bus which gets power from the signal line...