TPS sensor drift with temp

[reggid]

hi has anyone else encountered TPS drift with temperature?

if i set the 0 and WOT throttle with engine fully hot and heat soaked etc and then check the TPS cal when cold it reads 2% with throttle closed. since im using Alpha-N with ITB this causes a rich condition upon cold running and warmup. presumably as the sensor heats up then the sensor drifts back to 0%. its rather annoying as with alpha N the fuelling changes rapidly with only a couple % change in throttle position and it takes some time for the engine to heat up to a point where the sensor itself gets hot much longer than the time for coolant to reach 80C

[grom_e30]

Yeah I had that problem at first with my bike was a bit annoying I had no iac valve either so was using idle advance to keep it running till the engine warmed up and some times the throttle drift would cause it not to engage and the engine would stall. Changed over to itb mode and the slight tps drift didn't matter any more and was smoother to drive also.

[reggid]

Yeah I had that problem at first with my bike was a bit annoying I had no iac valve either so was using idle advance to keep it running till the engine warmed up and some times the throttle drift would cause it not to engage and the engine would stall. Changed over to itb mode and the slight tps drift didn't matter any more and was smoother to drive also.

i'm trying to use a contactless TPS from penny and giles which is supposed to have next to zero drift with temp to combat the issue, but i cant even get it spit out a signal when its connected to the harness. works fine with a simple battery and multimeter when not on the car.

for some reason my +5V supply voltage at the harness drops when the sensor is connected to it and i dont get any signal change as the throttle is rotated. its got me stumped

[grom_e30]

what sensor is it? some can be powered by up to 30v, so you may be able to power it off the battery.

[reggid]

what sensor is it? some can be powered by up to 30v, so you may be able to power it off the battery.

the hall effect is a penny and giles.

it seems i have a supply issue with the 5V, when i plug my normal resistance style TPS into the harness (sensor from a bmw) my 5V regulated ref voltage falls so im not getting the full range of voltages (0.9V closed throttle to 2.15V WOT) and hence its more susceptible to drift.

if i use a 6V battery when bench testing it off the car i get 0.8V closed to 5.7V WOT yet with the 5V at the harness im getting much less.

[kjones6039]

if i use a 6V battery when bench testing it off the car i get 0.8V closed to 5.7V WOT yet with the 5V at the harness im getting much less.

6 volts is a bit low for doing any meaningful voltage testing, me thinks!

Ken

[DaveEFI]

what sensor is it? some can be powered by up to 30v, so you may be able to power it off the battery.

the hall effect is a penny and giles.

it seems i have a supply issue with the 5V, when i plug my normal resistance style TPS into the harness (sensor from a bmw) my 5V regulated ref voltage falls so im not getting the full range of voltages (0.9V closed throttle to 2.15V WOT) and hence its more susceptible to drift.

if i use a 6V battery when bench testing it off the car i get 0.8V closed to 5.7V WOT yet with the 5V at the harness im getting much less.

Measure how much current it is drawing? Although I'd be surprised if it's enough to worry Vref.

[reggid]

if i use a 6V battery when bench testing it off the car i get 0.8V closed to 5.7V WOT yet with the 5V at the harness im getting much less.

6 volts is a bit low for doing any meaningful voltage testing, me thinks!

Ken

actually not (its not a tiny watch battery lol), it gives the same results as a properly working ecu (obviously voltage readings at closed and WOT are slightly altered proportionally being 6v vs 5v ref but it certainly all functions properly with the battery). i certainly have a supply issue though my 5V vref drops to 2.2V with a sensor attached. couldnt get a reading for current for some reason

[reggid]

what sensor is it? some can be powered by up to 30v, so you may be able to power it off the battery.

the hall effect is a penny and giles.

it seems i have a supply issue with the 5V, when i plug my normal resistance style TPS into the harness (sensor from a bmw) my 5V regulated ref voltage falls so im not getting the full range of voltages (0.9V closed throttle to 2.15V WOT) and hence its more susceptible to drift.

if i use a 6V battery when bench testing it off the car i get 0.8V closed to 5.7V WOT yet with the 5V at the harness im getting much less.

Measure how much current it is drawing? Although I'd be surprised if it's enough to worry Vref.

its a 4K TPS so i should get 1.25mA right? any less and there is another source of resistance in the line somewhere?

[quan3165]

I think I am having a similar issue with my Hyundai TPS retrofitted on to my BMW M20 throttle body. It is especially annoying when I get current for the 0 TPS value and later it reads negative 5% or worse . Next week I am going to machine a plastic thermal isolator/adapter plate to try to resolve.

Do you think something like this would work for your build?

[reggid]

I think I am having a similar issue with my Hyundai TPS retrofitted on to my BMW M20 throttle body. It is especially annoying when I get current for the 0 TPS value and later it reads negative 5% or worse . Next week I am going to machine a plastic thermal isolator/adapter plate to try to resolve.

Do you think something like this would work for your build?

not really the thing will heat up eventually, either by conduction through the D-shaft or just from the ambient under hood temps

[baz-r]

Itb mode may help as it's like SD & AN in one so you can get the best of both.
Should be fine unless you have very low vac at idle

[kaeman]

what is your itb setup, I was having problems like that when I first installed my itb setup because all 4 butterflys in each bank had a single d shaft, the problem was that the manifold was expanding when warm and would bind up the butterflys, I had to cut the shafts and install flexible couplings into each shaft to allow for thermal expansion, then recalibrated the tps and never had drift problems again. the problem was not the tps sensor. it was manifold expansion between hot and cold temps.

[Yves]

Since I went to the new ms2 code, I can see tps position in .1 increments. It goes from 1.9 cold to 2.3 hot.

IMO it has to do something with the linkage expanding, not so much the manifold, but each manifold is different.

I have been trying to resolve the issue with CAI, but I was unable to the get the rpms up.

[reggid]

What voltage should I get probing with a multimeter the ECU side of the harness with ignition on between the “signal pin” and “ground pin” with NO sensor plugged into the harness?

Should I get:

5V, between “signal pin” & “ground pin”

As well as

5V, between “5v pin” &“ground pin”

???

Or should there be no voltage between “signal pin” & “ground pin”?

I have come to be aware that some ECUs will have 5V at both the “signal pin” and “5V pin” without a load on it. Subsequently when the sensor is plugged in the voltage in “signal pin” is pulled down to a lower value say 0.5V closed throttle which then typically increases to 4.5V WOT depending on ECU and sensor characteristics.

For example a Siemens ECU on an OEM BMW harness registers 5V between “signal pin” and “ground pin” and also between” 5V pin” and “ground pin” there is 5V. The difference being between the” 5V pin” and “ground pin” its always 5V regardless of closed or open throttle.

[reggid]

after fixing the harness supply voltage i managed to get the penny and giles TPS to work. no more drift and no more noisy signals. these aren't supposed to wear out either. happy days so far.