circuit testing and confirming the bad

undefined Question


Got an Audi in with an error on a 'reference voltage sensor circuit C', and a code for a DPF filter being unexpectetly high filled.

When starting cold the system worked and you see 13grams of filling of DPF (ash).. both calculated and expected are equal. As soon as the Sensor reference voltage problem shows up, you will see the filling rate bounce up rapidly, hence the error message for DPF on the instrument panel and code set.

When measuring 5v. circuit you will see the voltage being dropped to 3.x volt. (connected ofcourse, disconnected 5v.). Question now was, is this the sensor or is it the 5v. circuit (ECU). Since the code was set it's not wiring (measured inside ECU) or connectors.

I connected an incondessant (spelled correctly?) testlight and it took the circuit down about 4.3 volt. So the sensor is taking more current then the testlight. In this case, I condemned the sensor immediately, after learning a sensor is taking only miliamps. Also the labscope did not take the circuit down, it shows about 5v.

Is this a standard we can hang on to? Is a sensor always taking only miliamps?

You made me check my 50mA light at 5V :silly:



This is 5V LED light I made for testing 5V circuits.



This one is grounding the CKP signal wire on a JTEC PCM. My 5V light lit up when I tried grounding the wire with it so I wanted to see how much current it'd supply.

The signal wire you can easily drop, since it's a real low current voltage. It's especially about the 5v. circuit, to confirm it's an ECU issue, dropping the 5v. or it's a sensor doing it. With more then one sensor on a line it's quite easy to check, disconnecting different sensors... this specific one I referred to had only 1 sensor on it, so it's a bit harder to condem one or the other.

In your tests it's clear it's taking several miliamps (the green led one is only 7.79ma). That should never take the 5v. down, unless indeed there is an issue. An other idea I had is taking 'a' sensor and put short leads with pins on it so you can plug it in any circuit to test. Ultimate question is still, with how many (mili)amps on average is a sensor putting the strain on a circuit So how does it compare to a test light (led or bulb)