PSA EGR Emulator

[Badazdz]

Helo, I am trying to create a circuit that will replace EGR valve on a car and make an ECU think that everything is ok.
I have already read this post https://maker.pro/forums/threads/help-to-design-egr-valve-simulator-cheater.244193/ but after assembling it and connecting to a car, it shows 2.8V constantly on the sensor output so to the ECU it seems stuck.
I have connected it this way. You can see what each pin means below.






This is how the EGR looks like (part num. 9671398180):






This is how it looks inside. There is a spring that pushes the valve to closed position so it automatically closes the valve when no power is given to the motor.






It has a motor inside driven by PWM:






This is how it is connected to the ECU:
1 - PWM-
2 - PWM+
3 - Position out
4 - Pos GND
5 - Pos Ref 5V






What I have already found out about my EGR is that the voltage range from the position sensor is from 0.7V (closed) to 4V (open).
I have tried measuring the PWM when the EGR is fully open (engine idle) and the position sensor was at 4V. The frequency of the pulses is 2000 Hz but I don't know how to put this information into a circuit.






Can someone help?

 

[kellys_eye]

All the potentiometer is doing is reporting the position of the motor (flap) back to the ECU. It's effectively a very basic servo system (as might be used in any radio control model).

The flap (motor) position isn't decided by the potentiometer rather than by some other signal (probably engine temperature, maybe exhaust temperature?) and the flap will recirculated engine exhaust until the engine reaches some specific temperature range and then (gradually i.e. with increasing temperature) close the flap.

In theory you could feed the engine temperature signal (which itself will be a variable voltage) straight into the position signal point of the EGR pot.... of course the signal may require some 'attenuation or adjustment' to fit with what would normally be expected from the EGR feedback pot i.e the signal may require 'inversion' (volts DROP as temperature rises rather than rising with temperature etc).

 

[Badazdz]

The EGR is opening/closing even when the engine temp raises.
It is closed when DPF regen is active or the car is accelerating otherwise the ECU open/closes the valve as needed, probably based on readings from an oxygen sensor (lambda) which checks how many oxygen remains in exhaust gasses.
I think that the solution from other thread I posted is good because it actually converts the PWM signal to the position of the valve (as voltage) but as I don't understand what each component in the circuit does I don't know what to change/adjust to make it work for this particular EGR.

Attached is the image of the hall effect sensor which is on the cover side.

 

[kellys_eye]

A PWM pulse stream will have an average DC value depending on the width of the pulse irrespective of its frequency so all you need do is rectify (if it was 'ac' - but in your case it isn't so no rectification needed) ) and smooth the pulse stream using a resistor/capacitor filter. This will give you a DC voltage that you can either use directly or 'scale' using a non-inverting op-amp setup.

See this link for some more details on what's happening....

Arduino LM358 PWM LowPass Filter Voltage Converter

Arduino analog output with low pass filter and amplifier for voltage control with PWM signal and capacitor. We use the LM234 or the LM358 with PWM signal from Arduino and low pass filter. See more...

electronoobs.com

The use of the op-amp gives you 'buffering' and the second op-amp in that package can be used to 'scale' the voltage if this is found to be necessary.

Try to plot the DC volts against the position of the motor (fully closed to fully open) and get back to us with the results. You can rig up a simple RC filter and connect it straight to your test meter.

 

[Badazdz]

I have previously measured the DC voltage for the PWM and when the EGR was open it gave me 2.3V and when closed 0.5V but I guess this does not correlate to the oscilloscope image I posted in the first post where peak pulse is 1.5V.
However I will check the article you sent, thanks.