Just to let everyone know, I installed another knock sensor on the rear rotor and just Y'ed them together. My base readings went up from 25 to 50, but seems to be working. Hopefully this will cover knock on the rear rotor better.

 

Are you running a PFC?

 

Yes, highest knock I've seen is around 80 under boost.

 

hehe, why do you ask?? Is it because it doesn't do anything about knock??

Really, it should though, I can't think that it would be that hard to retard timing on knock.

 

Hmm...there are some things you should be aware of with this system:

The knock sensor is just a piezoelectric microphone, especially sensitive over a narrow range of frequencies, presumably in the dominant frequency range of detonation. The sound pressure levels the mic detects are converted into a voltage. By simply Y-ing the two sensors together, you are effectively adding the two voltages together, resulting in an effective 3 dB increase, or a doubling in the voltage. This is why your knock went from 25 to 50. Your readings will be approximately double your normal readings with this setup, providing the two chambers are combusting similarly. Adding a second knock sensor may give you some warning, but the addition of the two signals and the logarithmic nature of sound will tend to mask any obvious differences between the two signals.

All in all, I don't think it will be very effective for the reasons stated above.

What would be really cool and useful, is if you could rig up a system to display the highest knock reading between the two sensors, and not the addition. FJO is doing this with their new dual-channel widebands -- they can even display the higher of two EGT probe inputs in addition to displaying the highest AFR from the two wideband sensors.

 

I realize it raises the noise floor, but I believe it will still pick up the rear rotor better. I am an audio engineer, I have thought it through.

I am going to try logging the outputs of each sensor in FC EDIT.

 

Yeah, you are right, it will still at least include the effects from the rear rotor, they will just be masked somewhat by the addition of the two signals.

I think logging each sensor is a GREAT idea, at least temporarily. Once you see the relationship between the two rotor chambers, you should have a somewhat reliable transfer function between them for tuning.

 

Yep, it should, otherwise it's pretty much a useless display. Unless a person is going to watch it the whole time they are driving (instead of the road), it will be too late to notice a high knock reading.