Interesting problem with piggyback MoTeC install

[anthonym]

I installed a MoTeC M48 in my MR2 over the weekend as a piggyback system to the stock (gen 3 3SGTE) ECU. The MoTeC is spliced into the factory RPM, cam position, air temp, TPS, MAP, speed and stop sensor voltage outputs (all of which remain connected to the stock ECU as well), but has it's own power supply. Currently everything including fuel and timing is still being run by the standard ECU. Everything is working and the MoTeC is receiving credible signals from all sensors but here's the strange thing: when the MoTeC is plugged into the loom, the idle RPM increase to approx 1400 Unplug it and idle RPM returns to normal, unplugging just the RPM and cam position sensors from the MoTeC had no effect (idle stayed high). I'm wondering if anyone has any ideas as to what might be causing this?

[fivebob]

Engine Temp Sensor?

[vvega]

is it connected to the air bypass solinid??

oh and what five bob said

v

[fivebob]

From looking at the wiring diagrams it appears that you can't piggyback any of the temp sensors unless you connect them to an AV input (not a temp input which has a pullup to 5v), everything else should be ok though.

[anthonym]

is it connected to the air bypass solinid??
v

Yes, the stock ECU is, and the MoTeC is affecting it's operation, which is not what I want.

[anthonym]

From looking at the wiring diagrams it appears that you can't piggyback any of the temp sensors unless you connect them to an AV input (not a temp input which has a pullup to 5v), everything else should be ok though.

You are right, disconnecting the coolant temp sensor from the MoTeC solves the problem, looks like I'll have to dupilcate all the temp sensors .

[krad]

If you have some electrical skills you could probably buffer the sensor that is causing you problems with a simple opamp follower circuit.

eg: sensor -> to normal ECU input.
-> opamp circuit (gain of 1) -> motech.

The opamp would effectively isolate it from the motech's 5V pull up.

This way you could just use the one sensor.

[fivebob]

You are right, disconnecting the coolant temp sensor from the MoTeC solves the problem, looks like I'll have to dupilcate all the temp sensors .

Or upgrade to the M400 and use the AV inputs which can be assigned to ET/AT unlike the M48 which has fixed assignments

[fivebob]

If you have some electrical skills you could probably buffer the sensor that is causing you problems with a simple opamp follower circuit.

eg: sensor -> to normal ECU input.
-> opamp circuit (gain of 1) -> motech.

The opamp would effectively isolate it from the motech's 5V pull up.

This way you could just use the one sensor.

My knowledge of electronics is somewhat limited but I don't think that would work as both ECU's have the 5V pull up so I can't see how the Motec would receive the correct signal.

From what I understand the ECU measures the voltage drop at the pin to determine the resistance of the sensor, which is just a thermistor, so you must supply the 5v to determine the resistance. The Motec and the Stock ECU have a pull up to 5v so if both of them supply 5v then the overall resistance value of the circuit would appear higher to the ECU's as the circuit has three resistors in it instead of two.

That is why you should be able to use the Aux Voltage pins on the Motec and get the correct value to both ECUs as only one is providing the 5v source, of course this relys on the stock ECU having a 1k Ohm resistor, or else you need to recalibrate the Motec (not very hard to do).

Anyone with some more in depth knowledge of electronics care to comment? I'm just a poor dumb computer geek so it's not my field of expertise.

[anthonym]

If you have some electrical skills you could probably buffer the sensor that is causing you problems with a simple opamp follower circuit.

eg: sensor -> to normal ECU input.
-> opamp circuit (gain of 1) -> motech.

The opamp would effectively isolate it from the motech's 5V pull up.

This way you could just use the one sensor.

Thanks for your input, it's an interesting idea, but I'd be concerned about the stock ECU receiving the correct signals and there is no way of checking this. It would be pushing my electronics skills to the limit as well .

[anthonym]

You are right, disconnecting the coolant temp sensor from the MoTeC solves the problem, looks like I'll have to dupilcate all the temp sensors .

Or upgrade to the M400 and use the AV inputs which can be assigned to ET/AT unlike the M48 which has fixed assignments

Actually that's not entirely true, the latest software for the M48 enables you to use aux v for engine temp, interestingly to solve this exact problem (sharing the sensor). I would upgrade if I could get a reasonable price for my M48, but few buyers recognise the value of the wide lambda option, and fewer still are prepared to pay a premium for it.

[anthonym]

That is why you should be able to use the Aux Voltage pins on the Motec and get the correct value to both ECUs as only one is providing the 5v source, of course this relys on the stock ECU having a 1k Ohm resistor, or else you need to recalibrate the Motec (not very hard to do).

I can use aux v for ET, but there is no provision for air temp, which I presume is similarly afflicted since it also has a pull-up resistor. Nevertheless, adding an additional air temp sensor is not a big deal. The ET sensor does need to be custom calibrated, how do you do this easily?

[fivebob]

The ET sensor does need to be custom calibrated, how do you do this easily?

You'll need a reference source for the ET so you know what the temp is, then it's simply a matter of connecting the Motec and the A/D value will be displayed, then enter that value in the appropriate cell. Other than that you'll need the resistance curve of the sensor and a least one voltage reading at a known temperature to calculate the A/D values.

[MartinNZ]

Nemesis...

Just wondered, why are you doing this? Why not just let the Motec completely control the engine? Are you just using it for datalogging?

If you can point me at the wiring info somewhere on the web I can try suss it out for you. I work as an electronic design engineer.

Martin.

[FLAWLES]

i ve got te opiste problem with my piggy-back comp my idol dosnt stay at a constant rpm it goes from 1000rpm to about 1200rpm just bounch back and forth
treid smoothing it out with the tps adj no go

[anthonym]

Nemesis...

Just wondered, why are you doing this? Why not just let the Motec completely control the engine? Are you just using it for datalogging?

If you can point me at the wiring info somewhere on the web I can try suss it out for you. I work as an electronic design engineer.

Martin.

Mainly to see if it can be done. And yes, initially I'll be using the MoTeC for data logging, as well as checking AFRs etc. I also want to be able to swap between the 2 ECUs quickly and easily for tuning purposes. Thanks for the offer to look at the wiring, unfortuantely I don't think Toyota or MoTeC share circuit diagrams for their ECUs (if that's what you'd need).

[anthonym]

You'll need a reference source for the ET so you know what the temp is, then it's simply a matter of connecting the Motec and the A/D value will be displayed, then enter that value in the appropriate cell. Other than that you'll need the resistance curve of the sensor and a least one voltage reading at a known temperature to calculate the A/D values.

This is not fun at all as the A/D values are not displayed for this configuration, but I've managed a crude calibration so far by matching the sensor voltage to a standalone temperature gauge with a probe stuffed down the filler neck, followed by on the fly adjustments .

[fivebob]

This is not fun at all as the A/D values are not displayed for this configuration, but I've managed a crude calibration so far by matching the sensor voltage to a standalone temperature gauge with a probe stuffed down the filler neck, followed by on the fly adjustments .

Not even displayed in the View screen? Yet one more reason to upgrade to the M400

[krad]

Fivebob you're right, the double pull up issue makes my idea pretty much redundant I didn't realise both had pull ups.

Wonder if some diode trickery would help out?

Go the A/D if you can.