Saturday, March 14 2009
"Oxygen is a poisonous gas that kills all that it touches." This is the sort of thing an anærobic bacterium might be suspected of writing. To us eukaryotes, of course, oxygen is a wonderful gas that allows for a basic necessity of life: respiration. Oxygen is so plentiful and so taken-for-granted that it isn't sold or taxed. It's just there for anyone who wants to inhale it or contaminate it with pollution. Unless, of course, you're in the tailpipe of a modern automobile. Oxygen is difficult to find in there, unless the car's catalytic converter is on the fritz. Evidently the catalytic converter is on the fritz in our accursed hatchback, because the oxygen sensor just upstream from the tailpipe has been detecting unusually high quantities of oxygen. I don't know how much it's been detecting, but it's been enough for the car's on-board computer to throw an error code.
I wanted to explore this issue further, so today I built a testing apparatus out of a cable fragment salvaged from an old oxygen sensor, as well as the cable that it plugs into on the totalled Honda Civic (parts of which I continue to salvage and repurpose). I used these pieces to intercept oxygen sensor signals, starting with our red Honda Civic four door, a car with an unproblematic exhaust system. After the car had been idling for awhile, the voltage output on the oxygen sensor (and it is a voltage output, sort of like a solar cell) hovered around 0.4 volts. It would go up and down from this value several hundredths of a volt.
Next I checked the signal coming off the accursed Honda's post-cat oxygen sensor. This seemed to vary a lot more, ranging from 0.1 volts to nearly a whole volt. It seemed like it might be possible to fool the computer into thinking there was a working catalytic converter if I could do two things: stabilize the voltage (that would require a capacitor) and lower the voltage (that would require a voltage divider, or two resistors in series with a center tap between them).
So I built a little circuit and hooked it up between the oxygen sensor and the car's computer and monitored its voltage output for awhile. Interestingly, though my divider (comprised of two 1.5 kilo-ohm resistors) was supposed to divide the voltage in half, the voltage acted as if it had been divided by nearly ten. This suggests that the resistance between signal and ground without my tampering is already on the order of 150 ohms.
This is the circuit I tried today. The red line segments indicate breaks in the existing wiring into which I inserted my circuitry.
For linking purposes this article's URL is:feedback
previous | next