Knock Sensor Voltage

B8 S4
#1

All the tune knock discussion had me looking into alternative ways of analyzing how tuners were tweaking the knock response loop and I came across VAGCOM blocks that log actual knock sensor voltage vs. timing pull.

http://www.ross-tech.com/vag-com/m_blocks/020-029.html

Do you know if anyone has tried to log both the voltage and the timing pull so see how they are correlated for a given vendors tune.

Since “desensitizing” the knock sensors in application is reprogramming how the ECU responds to a given voltage value it would be really interesting to see at what voltages trigger how many degrees of pull under given conditions and compare that across OEM, GIAC, APR, REVO.

I thought that might be interesting, but I obviously have very little experience in this area so putting it out there to see if I am totally off base or if you guys had any thoughts or insights.

#2

I had a similar thought a while ago, but there’s one major problem: the frequency at which knocking or pinging occurs would require sampling well beyond the rates of VCDS. Here is a great background on the subject: http://www.ti.com/lit/an/spra039/spra039.pdf

Here is a quote about the frequency spectrum in which the knock occurs: [quote]The shock wave
excites a characteristic frequency in the engine, which is typically in the 5 kHz–7 kHz range
[/quote]
If you want to detect a signal on the order of 5 kHz, then the nyquist sampling theorem states you will need to sample at least twice this rate, so 10 kHz, which is equal to a sampling every 1 / 10000 = .1 msec. Even with turbo enabled, you’ll be lucky to sample at .1 sec with VCDS.

So really the only way one could do this would be to hook up directly to the ecu/sensors with your own microprocessor or spectrum analyzer/scope. Unless I am confused about this…

Another approach would be to look at the “knock totalizer” signal. I would assume revo isn’t changing this measurement. They are instead modifying how the timing pull reacts to knock detection. In theory, one should be able to log the knock totalizer to determine whether revo is allowing predetonation to occur despite the timing pull from knock measurement showing no movement.

#3

Are there any “must have” logging settings you guys want for Revo on a dyno, when I do that? Otherwise the guy is just going to tell the tech “log everything” which might be unwieldy.

#4

If the operator is willing, and/or you are, then doing back to back pulls to the point that it heat soaks enough to pull timing would be interesting with all the standard parameters A/F ratio, retard, IAT, etc.

#5

It might be better to have more information not less. Then just sift through it all and use what’s helpful

Knock stuff can be hard to really to just take all the possibilities based off the logs. I think the ecu is able to sample much faster then vag com reports. When you look at knock what’s kinda hard is unlike AFR and timing knock can be effected by so many things it would be hard to actually pin down what they might have done. Back in the day when I was logging almost every car I saw they all seemed to act differently even when modification wise they were all the same.