JohnBradley

iTrader: (30)

I have been getting some phantom knock with EVOscan. The other day I did a 20-130 pull, and had 0 knock in 1-3, 2 counts in 4th (6500-7000), and then I got 10 counts in 5th 6300-6700. I richened the top end a little bit just to be sure (9.7 to 9.3) and am awaiting another chance to make a pull.

On work computer so I dont have the log with me, but the only thing that I can remember is during the pull, coolant temp started at 204.26 dropped to 201, and by 5th gear had hit 212. Intake air temp never got over 71*, with a starting temp of 58.

I had 14* of timing (which is dead on the map at 240-260 and 6500rpm) so it didnt pull timing that I can see. I noticed it also measured 1 count or so everytime I shifted. So it obviously picks up alot of background radiation as well.

TTP Engineering

iTrader: (465)

I have got 35 counts at 3200rpms, cruising 30% TPS which pulls timing to -10* which is likely the max retard. I think with the lopey cams and stiff mounts all around it is making the knock sensor extra sensitive.

Its just another reason EGT is beneficial. I can see the EGT's climb at cruise at the RPM that likely vibrates the knock sensor.

l2r99gst

iTrader: (2)

JohnBradley,

I'm curious what your timing map says in the load column at 6500RPM just to the left of the 14. Is it 17?

I'm just curious to see if the Evo ECU acts like the DSM ECU. With 10 counts of knock, approximately 3 degrees of timing would be pulled. Perhpas you were in a lower load column than you expected?

Also, speaking of coolant temps, the DSM ECU would pull one degree of timing at anything over 206*. I don't think any map like this was discovered for the Evo, yet, but I'm curious if this is the same, too.


Eric

l2r99gst

iTrader: (2)

That is exactly on par with the formula for the DSM ECU of .35* of timing per knock count. I forgot the 2G maximum, but the 1G maximum was 43 counts, or 15* of maximum timing retard.


Eric

dudical26

iTrader: (17)

Hmm... not a bad idea. I mean if it was knock then yes it would probably go away.

It is just kind of hard to isolate variables here since the behavior is fairly erratic. Multiple runs result in different behavior. From no knock, to knock with timing pulled, knock without timing pulled, and no knock and timing pulled anyway. I may have to try and do several runs to try and get some sort of reliable base line. Or at least enough runs that I can come up with some averages for.

TTP Engineering

iTrader: (465)

Actually Eric I think it is more than that.

When I say -10, I am saying below 0* timing -10, not -10 degrees off the programmed ignition timing figure.

I need more info to be clear.

KOEvo

iTrader: (5)

The question is more, how many volts of knock is showing on your counts? Every car knocks, pretty much all the time, it's just when it gets in the 3-4 volt range that is becomes detrimental.

l2r99gst

iTrader: (2)

Ahhh....I see. Yes, I misunderstood your statement.

I'm still curious to see exactly how much timing the Evo ECU pulls per knock count, the max retard, etc.

Hopefully I can start some research soon.


Eric

JohnBradley

iTrader: (30)

Eric thats the thing, I am very careful now when I make claims. The load cell directly adjacent to the 14 degrees was also 14*. Unless I was running in the 220 column at 21psi? That is the first location where there is more than 14*, and it is 16*. I backed it down to 13* so we will see what happens. The coolant temp hit 212 for a logged peak of 3 seconds before it dropped to 209. Right after that I had to lift, so coolant temps only spiked it wasnt running sustained that high. The other thing is it only came up with that high of knock in 5th at nearly 130mph.

l2r99gst

iTrader: (2)

No, that's not true. The knock counts reported by the ECU already takes the knock sensor voltage as well as many other factors into account. The knock count is calculated by the ECU already for you. This is basically when the ECU says that you are knocking.

Knock voltage alone doesn't tell you much, unless there is a dramatic spike above a background threshold. And that threshold is definitely not any particular number, like 3 or 4 V, as the background noise is roughly proportional to RPM, etc.


Eric

l2r99gst

iTrader: (2)

I would like to know the results after your dropped the 16* to 13*. Keep us posted. Also, about the coolant temps....it doesn't matter if it is sustained. A spike above a certain temp (at least for DSMs) would pull timing.

Also, I would expect you to knock at the top of 5th moreso than at the top of third or fourth, just because of heatsoak to the IC, coolant temps, etc.

I'm just trying to figure out if you actually had timing pulled or not.


Eric

JohnBradley

iTrader: (30)

Okay, I understand what you mean. Underhood temps (just going by the IAT) raised the before mentioned 13* during the run at wherever the temp sensor is. I would imagine the rest of the system was warmer than that as well.

If I can get away from work early I will make a run to my secret spot and make a pull or 2. Without name dropping, I was told today that IX's dont like/need timing like an 8 will/does. So if it doesnt need it might as well take it out right?

dudical26

iTrader: (17)

ok here are some logs. First are the logs from my Evo IX. There is one log that is knock while bone stock. And another log with knock while at 14psi on an intake and a MBC running on an E-flash

Next group is an Evo VIII with an inake and stock ECU.

http://www.skidmore.edu/~a_blaker/Car/Evo%20IX.zip

http://www.skidmore.edu/~a_blaker/Car/Evo%20VIII.zip

stingerbtry

Those are about the same numbers as I saw on my stock evo IX on a hot afternoon.

Ludikraut

iTrader: (17)

^ What year was the Evo 8 that was logged?

l8r)