Lambda sensor - plus a bit of GuzziDiag

[tris]

I've suspected for a while that the Lambda sensor on my B11 was playing up, but I think I've broken it enough now to properly fault diagnose it 

Initially it suffered from surging at low (presumably in closed loop) throttle settings. After a 20 mile continuous 80MPH blast down the A1 the other day it now appears to have gone rich on me.

I thought that I might be able to make some sort of diagnosis using the Lambda mV displayed in GuzziDiag

Using a lap timer I established that initially (60 deg oil temp) it was switching from 100mV to 900mV about every 0.8 seconds. This isn't fast enough for a healthy Lambda sensor (should be 0.3 second ish IIRC) but I might be seeing the refresh rate of GuzziDiag.

Any comments.

However, when it was proper warm I found that when I held the throttle to 3k RPM for 10-15 seconds it stubbornly stayed at 15mV until I released the throttle when it jumped up to 850mV

So I think that means I have a dead/dying O2 sensor, as the ECU thinks the mixture is lean as its seeing a low voltage so enriches things everything up

Does that make sense before U splash out 70 quid on a new one

Cheers

Tris

PS its an 05 bike having done 35k miles on (presumably) the original O2 sensor

[pauldaytona]

why is it odd that when you had the throttle at 3k and then closed it the mixture goes wild?

[tris]

I didn't make myself clear Paul

It was the fact that it stayed put all the time I had the throttle fixed that puzzled me.

I thought that the voltage should vary all the time between max and min

15mV to me says either
a) the exhaust gas is lean hence the 15mV
b) the O2 sensor is faulty and stuck at 15mV when at full operating temperature which would give the ECU the wrong information

However, its very possible that there is some fundamental as to how the Lambda sensor interacts with the ECU that I'm missing and I'm happy to be educated

[Wayne Orwig]

15mV to me says either
a) the exhaust gas is lean hence the 15mV
b) the O2 sensor is faulty and stuck at 15mV when at full operating temperature which would give the ECU the wrong information

c) The O2 sensor rate is to high to pass anything of reason over the slow serial connector to GuzziDiag.

Since you proved it can swing up to 900mv, I believe you proved it is working. Connect an oscilloscope directly to the actual sensor and get actual data. Report back.

[beetle]

I agree with OMG. Looks like it's working. GuzziDiag doesn't poll the ECU fast enough to get real-time data for the lambda. Going rich on a closed throttle is normal. The ECU doesn't trim unless the RPM and throttle are in a steady state, and rapidly decreasing revs on a closed throttle isn't a steady state.

[tris]

Thanks Guys - I forget that line of action then!