fireroasted

iTrader: (3)

I am working on a problem I am having with a CEL posting P0325 continuously. It happen during a track day at Fontana. I have posted in advanced, but this is NOT a cross post; this is a very pointed ECUFlash troubleshooting question. Since I am trying to log to figure out if the answer to my problem is in the log data.

I have attached 3 short/simple logs for review. My basic question for the gurus is:
If knock is detected, in this case 9 from 0, does the ECU bleed off boost as a precaution or does it only impact timing?

I see in my logs that I go from 0 to 9 counts of knock instantly, and it stays there. I don't here any in the car. I see that when this happens the boost is killed off. So is that ECU action or does that indicate a boost leak?

I am trying to see if I have a boost leak somewhere, that isn't obvious that I haven't found yet, that is actually causing my knock values instead of the other way around.

Thanks for any help or troubleshooting direction.
M

infini_IV

iTrader: (4)

Im not so sure "knock counts" is being used/logged correctly. The mut 3 does not pull "knock count" data from the ecu. It reads "knock CA" OR knock retard crank angle

Im interested to see if evoscan has been incorrectly set up and the reading everyone believes is "knock counts" is actually degrees of ignition retard, anyone able to shed light on this??

Boosted Tuning

iTrader: (12)

Knock counts cant pull boost, unless your using a Tephra modded ROM with the drop boost feature, which drops boost on a user set knock sum

EvoScan pulls the straight "knock count" value from the ECU. If you look at EvoScan "knock count" formula, it is just "x", which means it get the exact value from the ECU.

infini_IV

iTrader: (4)

what im saying is the data coming from the ecu is not "knock counts" its knock retard. so a reading of 2 from the ecu is 2 degrees of retard not 2 counts of knock.

GST Motorsports

iTrader: (8)

It's knock counts, not knock retard.

It's complex algorithm, but 3 knock counts usually equals pulling of 1*

- Bryan

Boosted Tuning

iTrader: (12)

Once again, your wrong. "Knock count", actually called "Knock Sum", is just that, knock count/sum. Its not degrees of retard. Please stop posting mis information.

infini_IV

iTrader: (4)

Well these drive recorder files from an evo x on the mut 3 show it as knock retard

GST Motorsports

iTrader: (8)

There is knock sum on the Evo 10 as well

- Bryan

infini_IV

iTrader: (4)

its the same for all, there is learn knock which express's in a percentage how much of the knock curve the ecu has learned and then there is knock CA or knock retard

GST Motorsports

iTrader: (8)

Ok let's just make sure we are all on the same page here.

Are you talking about the Evo 10 or the Evo 8/9?

Just FYI this is the Evo 1-9 section, so it has to do with the 8-9. There are actually a couple pretty good threads on knock detection in the 8-9 ECU in this area, do a search and read away.

The Evo 10, as I think you are presenting, is slightly different. There however is knock sum as well in that ECU.

- Bryan

infini_IV

iTrader: (4)

doesnt matter which evo 7,8,9,10 none of them have a knock sum calculation that is read via the MUT they only read knock retard in degrees and learn knock as a percentage

GST Motorsports

iTrader: (8)

ok

- Bryan

Boosted Tuning

iTrader: (12)

Lo, he beat you into submission.

03whitegsr

iTrader: (8)

You know, the term "knock sum" is slightly misleading. Being a "sum" it could lead somebody unfamiliar with the term to think it is a summation of the knock events? Maybe that is what he is saying, which in that regard, no, there is not a "summation of knock events" channel. Although, "octane" would be a moving average of sorts.

mrfred

iTrader: (50)

You really should do some reading in the Evo 1-9 ECUFlash forum. They are right, and you are wrong. I'm actually a bit surprised you haven't been pummelled by more people. Guess they figure what's the bother.

BTW Bryan, it is ~1/3 deg per knock count. Its just that the ignition timing MUT we have been logging all these years doesn't have the resolution to see ~1/3 deg increments.