Knock sensor code

[eboggs_jkvl]

I tossed a P0327 Circuit low Input Bank 1. Do we even have a PID to read and monitor the freq or output from the Knock Sensors?

Elmer

[joecar]

Elmer,

I didn't see PIDs for either of those; PIDs for actual knock counts and frequency would be very handy indeed.

I'm hoping Paul chimes in here...


Joe

[eboggs_jkvl]

Yep, it got mad and didn't give the correct frequency to the PCM and I tossed a code. I cleared it and I'll see if the code pops up again. I'd like to see what caused it and under what conditions.

Elmer

[Blacky]

While the knock sensors themselves operate around a given frequency, the knock value (i.e. strength/severity) processed by the PCM is done in volts (i.e amplitude of the signal from the sensors).

There are two components to the knock sensor control, hardware and software.

We cannot log the frequency side of things because the PCM never gets any frequency information.

The calibrations listed under Spark->Knock-Hardware are programmed into a seperate processor in the PCM. That processor filters the data coming from the knock sensors (based on frequency) and passes a voltage back the PCM that indicates the severity of the knock. The higher the voltage the more severe the knock. The PCM never sees the "frequency".

Once the voltage is sent to the PCM the software filtering takes place to (hopefully) remove false knock and to set and learn the base knock-noise level of the engine.

The only two PIDs that might help are: {GM.ESC1DNL} and {GM.ESC2DNL}

Regards
Paul

[eboggs_jkvl]

Well I knew they were transducers and we were talking voltages. The stupid code refers to frequency too low.

http://www.ls2.com/boggs/dtcs/DTC%20P0327.htm

[Blacky]

From that description:

"When an engine operates, the PCM learns the minimum and maximum frequency of the noise the engine produces. When the PCM determines that this frequency is less than or greater than the expected amount, a knock sensor DTC sets"

I would read "frequency" to mean "amplitude". The PCM learns the voltage thresholds of the knock sensor background noise, not the frequency range of the knock signals (that I am aware - I could be wrong).

If it really did set a code based on frequency, then a lot of false DTCs would go off due to there being a lot of high amplitude (i.e. loud) noise in an engine block that is outside the frequency range of the knock signature of an engine.

Maybe the hardware filters (that I can't "hack into") do some learning of the frequency ranges. But I don't see why they would - it makes no sense.

Not sure why (or even how) then PCM can determine that the frequency is too low or too high. It could be written up that way to make it easier to understand conceptually.

I guess if the frequency is too high or too low then no voltage will get past the hardware filters and the PCM won't see any knock background noise (i.e. below the learned voltage threshold) and will assume that the knock sensors have failed. I guess calling it "frequencey too high or too low" is reasonable - (at a stretch).

Regards
Paul

[joecar]

[eboggs_jkvl]

From that description:

"When an engine operates, the PCM learns the minimum and maximum frequency of the noise the engine produces. When the PCM determines that this frequency is less than or greater than the expected amount, a knock sensor DTC sets"

I would read "frequency" to mean "amplitude". The PCM learns the voltage thresholds of the knock sensor background noise, not the frequency range of the knock signals (that I am aware - I could be wrong).

If it really did set a code based on frequency, then a lot of false DTCs would go off due to there being a lot of high amplitude (i.e. loud) noise in an engine block that is outside the frequency range of the knock signature of an engine.

Maybe the hardware filters (that I can't "hack into") do some learning of the frequency ranges. But I don't see why they would - it makes no sense.

Not sure why (or even how) then PCM can determine that the frequency is too low or too high. It could be written up that way to make it easier to understand conceptually.

I guess if the frequency is too high or too low then no voltage will get past the hardware filters and the PCM won't see any knock background noise (i.e. below the learned voltage threshold) and will assume that the knock sensors have failed. I guess calling it "frequencey too high or too low" is reasonable - (at a stretch).

Regards
Paul

I don't feel so dumb now! I read it a few times and it didn't help me either. It appears that my background noise has/did reduce(d) and it didn't like it? Well, I'm going to give it another chance to get grumpy. I nuked the codes and if it comes back, I'll put one of the sensors to sleep.

Thanks for giving it a whirl.


Elmer

[Doc]

E,
Are you running the original 98 sensors still? Or have you replaced them? If you have replaced them, with the 98 part # or the 99+up # ? You are still running the 98 pcm and OS correct?

[eboggs_jkvl]

I have replaced them with 99+ sensors when I put in the LS6. I am running the 98 PCM and OS.

Elmer