recently got an evo 6 rs out of japan, and it came with some very severe solid mounts. the front and rear insulators are the polyeurethene type and whereas the left and right transmission are actually solid metal with a small rubber core and sleeve. as to be expected there is tremendous vibration in the cab. i'm using evoscan to log and i've noticed that just revving the motor in neutral i'll see knock sum as high as 9. i've never seen any higher sums. this is an oem ecu from another identical modded (piping, exhaust and intake) evo 6 rs and was reflashed and good for up to 1.7bar and a 12.1s 1/4 for many years (so it's not the mapping). in other logs i've seen as much as -4 degree timing, and the car feels very lazy for the boost it's running at this point. please see logs below for 1.2 and 0.77 bar respectively. any suggestions or questions?

1.2bar




0.77bar

 

I was having a problem like that and i I thought it was the mounts I installed. It turned to be a cracked header.

 

yes they can

 

1.7 bar on pump and you're not expecting it to be the map in there? thats 25 psi.

 

please read the post properly. i am running 1.2bar, i said the other car this ecu came out of ran a 1.7bar for several years with no issues whatsoever. that motor is actually still running today with perfect compression and over 60K miles

for your piece of mind here is log from a 0.77bar run. knock still occuring and timing still being pulled.

 

can all agree though that the knock being read is why timing is being pulled? and is 9 the highest sum possible, as i find it strange why it's always this value even when running 0.8bar?

 

so wait, its the solid mounts causing the ecu to read false knock correct? the car isnt actually knocking.

 

that's what i'm desparately trying to find out short of switching all the mounts out. i've already ordered four ralliart mounts. they are a harder compound rubber than stock, but should not vibrate to hell like these solids. would really suck if knock is still there after swapping 4 mounts, so i'm looking for alternative causes.

 

That can very well be phantom knock, since it is the same amount and same place time after time even at different boost levels.

One way to test for sure, would be to reflash and lower the timing 3 degrees where you are knocking.

If it is phantom knock, it could very well be the stiff motor mounts. Other culprits could be noisy lifters, downpipe hitting something, etc.


Eric

 

Knock counts can go way above 9 counts I beleive...

 

what does the knock look like? really high on the graph or really low?

 

This statement doesn't prove its not the mapping. Every car is different, even the same car with the same mods. Just because you map worked perfectly on another evo 6 rs, doesnt mean its perfect for your evo 6 rs. get a new tune, one made for YOUR car.

 

i knew there would be a couple of people that would suggest this, but i don't think it's the case.

how do you explain the knock sum in neutral with rpm and no load?
reflashed ecu's aren't perfect for every car of the same model, but can be pretty darn good as per ralliart ecu's sold for almost every year of evo.
the original conservative oem ecu that came with the car exhibited the same problem at boost levels below factory, 8psi, which is the reason we switched it for a known good one with a good map.
i think at boost levels as low 0.7bar we can pretty much assume the car variances between like cars with like mods are negligable.


what about a bad knock sensor? do they just go bad and not read, or can they give erroneous readings?

 

not sure what you mean, but the values that would be graphed are included in the attached jpg on the first post.

 

Joey´s hunch could be right. Does your car have the cast factory manifold or an aftermarket tubular job? I´d think that with solid mounts there´d be less chance for false knock because the downpipe and entire exhaust would have less tendency to move and bang the TC and oil pan.