What are known causes of phantom knock?

I ask because I am desperatly trying to get rid of knock under the following parameters:

Load: between 60-100
RPM: between 2500-4000

I have tried increasing fuel in the said areas (including dropping my open loop load conditions to control fuel), and also have tried reducing timing with no luck.

From analysing the logs the knock occurs regardless of speed, open/closed loop, and the car being cold or warmed up.

I am getting knock counts the whole range up to 36. Additionally, I occasionally get the p0300 error code.

From what I can gather I get these counts rather consistantly easing through 1st and 2nd gear from a light or from bliping the throttle in fifth gear.

 

Common among DSMs was phantom knock due to the lifters. Besides that, anything that can be touching or banging metal, like your downpipe, etc.

On my DSM, I used DSMLink and it had a setting where your could ignore the knock sensor below a certain RPM and TPS setting. I had my knock sensor turned off below 2800 rpm or and 20% TPS.

Maybe that could be the next ROM patch for tephra or mrfred. A patch to enable/disable the knock sensor based on certain conditions, such as load, RPM, etc.

Obviously before anything like this would be implemented, you would have to be absolultely certain that it isn't real knock. How much timing did you reduce in that area? Did you log to make sure that you adjusted the correct cells that were experiencing knock? If you reduced timing by a few degrees and still had the same exact knock count time and time again, then it would be safe to say that you are experiencing phantom knock. If reducing timing, verified from logs, does affect and lower the knock count, then it isn't phantom knock.

Eric

 

Eric,

I smoothed the timing map by 2-3 degrees in the entire area blending over into the adjacent cells. I thought initially that this had a positive effect but after more logging the knock counts reared there ugly heads again.

 

Well, I wouldn't be too quick to disregard it as phantom knock just yet. If it went away with the timing changes the first time, I would be hesitant to call it phantom knock. You may simply need to adjust more cells in that area, since you can be hitting a bunch of different cells depending on conditions, when you start your pull, etc, etc.

My test for phantom knock was always to reduce timing by a good amount and if the knock count stayed exactly the same or repeated under the same exact conditions all the time, then it was phantom knock. I'm not saying that yours isn't phantom knock...it very well may be. But, definitely do some more tweaking to rule one way or the other.

Then, if it does turn out to be phantom knock, maybe start a thread asking tephra or mrfred or someone to make a patch like I mentioned. Or of course, nail down the cause of the phantom knock and fix it.


Eric

 

eric,

Sound advice, I just dropped the timing by another 3 degrees in the entire area from 2500-4500 and blended it into the adjacent cells. Hopefully, this will tell me something. I am now sitting about 4-6 degrees below stock timing in this area.

 

Ok, back to back tests. Decreasing timing about 4 degrees throughout the entire area only made the knock more frequent if anything. Confirmed the timing did lower in the logs.

So back to the phantom knock question. What have others found to be causes? This is so **** fustrating.

 

so you have checked for any rubbing or knocking of downpipes , intercooler pipes, metal underpanel...?

I'd try putting timing back to stock. if that doesn't help, I'd try putting race gas in to see if it is definitely phantom.

 

I feel your pain, I went through this same thing and the fix was retarded. i had to deal with months of frustration, p0300's, money spent on nothing, etc. Im not going to name any specific company but my phantom knock was caused by my oil catch tank. It mounted in the space in the stock strut brace. I chased this problem for months and changed many parts for nothing. The nipples on the tank were bouncing off of the strut brace every time I hit a bump which caused a high knock count, ECU pulled timing, engine stumbled (no acceleration, horrible jerking ensued), CEL came on, I had to let off the gas and coast for a few seconds before I could accelerate again and drive to autozone and clear the code. It wasnt until I noticed when this was happening-which was whenever I hit a bump- that I was able to figure out the problem. I removed the oil catch can and the problem has gone away and never reared its ugly head again. Stay vigilant, it could be anything. Good luck.

 

Have any suspension noises at all?

 

Yes, the rear coilovers make a decent amount of noise going over consecutive, small bumps and the fronts make a slight creak when turning while driving down a slope (like a driveway).

 

I have looked at these before, they are very close clearances. I think that my next point of attack will be to remove the lower ladder brace and beatrush underpanel completly. Possibly, cross my fingers the knock counts will go away.

Currently, (as of last night's test the area of the map in question is stock timing. I wish race gas was easier to get ahold of near me (would have definetly tried it already). If the previous method fails I will be making a trip to the track to get a couple gallons to insure the knock is phantom.

 

Does it also do these knock counts in free rev at idle?

Have you added new parts recently?

 

Maybe, a ghost!? =D

 

Have you modified the cam timing? I've seen this happen on IXs with aggressive MIVEC settings with cams. It was mainly on advanced side though. Just a thought of something to try.

 

Yes, I have seen knock counts when reving in neutral up to 4k or so.

I never really monitored knock in the low load cells prior to getting into ECUFlash. For all I know it could have been knocking for the life of the car.

Nope, still running the stock cam gears and now I am back to stock timing advance in the cells in question. I do have comp cams +264 though.