

| Subject: | Re: O2 sensor coax wire testing |
| From: | David Beierl |
| Date: | Tue, 22 Jan 2002 03:17:48 -0500 |
At 02:45 AM 1/22/2002, Laurence Smith wrote: >I am looking for advice on how to test the green coax wire coming from >the O2 sensor. In the archives there is discussion about this wire >and the possibility of an internal short that causes the ECU to get >upset, run the engine rich and cause poor idling. However, the >archives were not very clear on a test procedure for this shorting >situation. The coax braid is grounded -- if it shorts to the internal wire you'll see a constant low voltage on the sensor and the van will run *very* rich. If you see this condition, unhook the sensor from the green wire and measure the voltage at the green wire with engine running -- it should go to about a half volt and stay there. If it does, prolly bad sensor. If it doesn't prolly bad wire. >Another question...the O2 sensor voltage is definitely produced by the >sensor - correct? Not the ECU. The ECU picks up this voltage and >looks for a nice consistent flip-flop between .45 and .55 volts - >correct? The ECU supplies a very high-impedance (easily overpowered) bias voltage to hold the system at about a half-volt until the sensor comes on line. The sensor is actually an odd sort of battery that supplies close to a volt (still at high impedance) when its temp is above about 300C *and* there is oxygen on one side of it and not the other, which is the case when a rich mixture burns up all the oxygen in the charge. The ECU notices that the voltage has moved away from the initial half-volt, and starts paying attention, i.e. goes into closed-loop mode where it adjusts mixture based on the sensor reading. If the sensor indicates rich (> half volt) the ECU leans things out until the sensor indicates lean (< half volt). The ECU then richens until the sensor flops over to rich again, repeat forever. If you look at the sensor signal with a scope you can see that it makes a sloppy square wave between about .2 and .7 volts in response to the ECU's fiddling -- however with an ordinary digital meter all you will see is that the reading jumps around above and below a half volt with no visible pattern. You must use a high-impedance meter to measure this signal, or you'll distort it and make the engine run richer than it should. Ditto when checking the green wire with the sensor disconnected. david -- David Beierl - Providence, RI http://pws.prserv.net/synergy/Vanagon/ '84 Westy "Dutiful Passage" '85 GL "Poor Relation"
Have something to add on this topic today? Revive this thread in the forum »