Calibrating knock sensor (Evo 9

When dialing in the knock sensor I'm trying to decide whats noise and whats not. I'm getting a spike around 7300 rpm most of the time (but not all). This is with barely any timing in and 9-13 psi (depending on the log as i changed the wastegate spring). I have the timing from what I remember in the 9-12 degree range, which should be plenty low given the psi. I even put in some C12 just to make sure it wasn't a fuel quality issue. On one log I did three runs back to back. The first one had the spike, the rest did not. It gets up to around 2.1-2.3 volts. The average is around 1.5 volts.Its a built motor, so I expect it to be noisy.

I tried attaching the log cals, but I can't attached .daq files (AEM Data file). I took a screen shot of a couple of the runs if someone can take a look. I pulled timing in those cells on the timing map, however it doesn't seem to change anything. I believe its noise, however I'm not 100%.

The car is an Evo 9 if that helps.

Edit: I just realized that I didn't take a screen shot with the proper pressure channel selected. All these runs are at 9-12 psi, so again, nothing excessive.

I'm just trying to see if its normal to have voltage spikes from time to time due to noise. The car has really stiff motor mounts, forged bottom end, 272 cams..etc so its not going to be a quiet motor. it always seems to happen around the 7300rpm (give or take) mark no matter the timing, AFR, or fuel used. As mentioned these pulls were all done on very low boost.

I'll look over the car and make sure nothing is touching the block. The down pipe might be a bit close, so I may move it a bit or extend it so it clears better. Baring that I guess I'll have to work around it?

Just looking for feedback from someone with a bit more experience than I.

I’d stuff and secure some heat wrap between that down pipe and whatever it might come in contact with just in case. Also, does it occur under only WOT and did you try various pulls with various TP %. Process of elimination.

If it still occurs, well, then rely on more tenured tuners on here. Good luck and I hope you solve it.

Thats a good idea use the heat wrap. The DP is actually wrapped, I just can't remember how far down it goes. The wired thing about it is that when I'm doing back to back runs on the same log, the first pull always seems to be the "noisiest" and the follow up pulls are quieter. Anyway, from what I read thus far I've came to the conclusion that its noise and I'll keep my eye on it as I move forward with tuning the car once its on the road. From some of the webinars I've watched I could just have one cylinder that's louder than the others. I'm using an older AEM S2 at the moment so I don't believe I can trim the knock sensor on any of the cylinders.

From what I can gather from searching around the past few days is real knock seems to jump out a lot higher than .5 or .75 volts above the rest of the noise. I mean the car had C12 and SUPER conservative timing it in, I can not see anyway that it was knocking.

I'm not too familiar with the knock control strategy you're using but are you just looking at raw voltage from the knock sensor? If so this will be quite limiting because you're essentially looking at all the noise the knock sensor is monitoring. Most modern ECUs run the raw signal through a DSP and allow you to focus on just the frequency where knock is likely to occur. I'm not sure exactly how this system works but if you can't define the frequency range then you may find it difficult to get really good control.

Hey Andre, thanks for the response.

This is on an older AEM Series 2 ECU. I'll probably upgrade it next year, but for now its capable enough.

The knock voltage that I'm logging is filtered through the ECU's DSP. The cal in the ECU is set up for an Evo, so the voltage should be filtered correctly for the engines harmonics.

I guess my main question is, due to my engine being nosier than stock, is some voltage spikes "normal" at times? I'll see spikes of .5-1 volt above the rest (not bunch together, just individual spikes). However, removing timing, adding fuel, using race gas all does nothing. I've ran the car at 9psi and 13psi...no change. The car is barely breathing with a 64mm turbo at 13psi. it always seems to do it around 7000 or 7300 RPM. It seems to be regular enough and not that high above the normal noise that I believe it should be ok. I was going to set up the knock threshold conservatively for now and go from there I guess.

I know you've tuned a lot of Evo's so I was just curious on your experience.

Side note: If I'm going to use a link knock block, where on an Evo9 is the best spot to put the links knock sensor?