Knock PID's on e40 LS2

[MadMaxHSV]

Can anyone confirm an assumption of mine, before I sh*t myself that i've actually been seeing a fair bit of knock. (perhaps cue that famous phrase around assumptions)

I've been logging GM.KNKRET (Knock Retard) and GM.KR (Retard due to knock)

My assumption based on the fact I've seen some activity on KR and lots on KNKRET is that KNKRET includes burst knock etc, and activity recorded on KR is 'actual' knock from the knock sensors.

Is that right?

Although i've noticed another one which is KNK_E40_DMA.....? Not tried logging this one yet.

Thanks

[MadMaxHSV]

Ok, a little worrying no-one seems to know what PID shows actual knock.

Surely this should be somthing monitored during tuning and getting the right one would be quite critical....

[MyM8V8]

A LITTLE worrying???

[gmh308]

Can anyone confirm an assumption of mine, before I sh*t myself that i've actually been seeing a fair bit of knock. (perhaps cue that famous phrase around assumptions)

I've been logging GM.KNKRET (Knock Retard) and GM.KR (Retard due to knock)

My assumption based on the fact I've seen some activity on KR and lots on KNKRET is that KNKRET includes burst knock etc, and activity recorded on KR is 'actual' knock from the knock sensors.

Is that right?

Although i've noticed another one which is KNK_E40_DMA.....? Not tried logging this one yet.

Thanks

Try zeroing out burst knock and look at the PIDs again to see whether they are showing the same KR. Usually best to zero/disable burst knock if you a chasing knock questions as burst knock will pull timing on most throttle movements and confuse things.

GMPX could probably advise what the DMA E40 Knock PID does. Maybe it is actual knock sensor activity.

[MadMaxHSV]

Ok, i've had a chance to play with a few parameters and I think i've sussed it.

Firstly I reduced the burst knock values, and the KNKRET value dropped for a similar run.
I also can get KNKRET to show timing pull just by blipping the throttle when stationary (in one test tune I had 12.5 degrees timing and 14.2 AFR at that moment, don't think this is true knock....)

I also tested the KNK_E40_DMA PID. And here's my theory. The attached log should bear this out.

KNKRET shows burst knock retard (plus any real knock recorded on GM.KR) based on change of airflow seen by the MAF.

KNK_E40_DMA shows burst knock retard based on change of airflow seen based on the APCYL_E40_DMA PID.

Neither is real knock on its own, although as said KNKRET does include real knock if present, but GM.KR shows this in isolation.

At the start of the attached log you can see a few throttle blips with a KNKRET peak matching the TPS output, and a KNK_E40_DMA peak matching the APCYL_E40_DMA peak.
8963

I have a log somewhere where GM.KR was showing to demo how KNKRET includes it.

[joecar]

Ok, summary:

KR = real knock
KNKRET = real knock and burst knock
KNK_E40_DMA = burst knock

have I got that right...?

[ I deleted my other post above because I was guessing and incorrectly at that ]

[MadMaxHSV]

Yes I reckon so,
its possible that KNK_E40_DMA also includes real knock as with KNKRET, but would have to see some to say for sure.

[MadMaxHSV]

Ok, bit more logging today and saw a little knock. Sh*t way to prove a point but hey :)
Final findings are as follows

GM.KR = Real knock (detected by the knock sensors)
GM.KNKRET = Real Knock + Burst Knock (based on MAF changes)
GM.KNK_E40_DMA = Burst Knock (based on APCYL_E40_DMA changes)

A log which shows this 8966

Hope can help someone else remove some uncertainty

[joecar]

Ok, thanks for info :cheers: I added a link to this thread from the collected tutorials thread.