Reverse Engineering an EVAP System - Part 2

related: embedded , c , firmware , cars , msp430

Following up from Part 1

LD Pump Switcher circuit hooked to LD Pump Simulator circuit hooked to MSP430… on the bench, everything looks good!

LD Pump Simulator on Bench

Powered it up in the Jetta, and nothing happened.  Verified voltage to the simulator and the MSP430, but no signals were detected on the LD Pump lines (any current between LD Pump VCC and LD Pump GND would cause the LEDs to turn on).

Further analysis shows that the balloon that is semi-permanently installed on his LD Pump is no longer inflating, while it used to inflate immediately upon starting the car.  

A bit of Google-Fu later, it has become clear that the LD Pump test only runs when the gas tank is between 15% and 85% full – his tank was near empty.  We’re chalking this one up to a bad starting condition, and will try again soon.

While we were in there, we also verified that the Purge Valve works correctly… how unfortunate; it’s an easy one to replace.

But the good news is the circuit works on the bench!  The wave generator input a 5V 5Hz signal into the LD Pump Switcher, which increased it to 14V and passed it to the LD Pump Simulator, which dropped it back down to 3.3V and passed it off to the MSP430.  The MSP430 immediately kicks on a flashing red light, and after 8 seconds we reach the target 40 pumps, the green LED kicks on, and the 3.3V sense line goes high.  The sense line is stepped back up to 14V, where it would presumably go back up to the ECU.