I took Al's advice and changed my knock sensor.

I also went ahead and wrapped a section of my DP that was near the crossmember.

I've datalogged for the last week and besides a couple of knock counts here and there I'm pretty much knock free.

I have to thank Al for going out of his way to help me solve this problem. He has contacted me repeatedly and helped me troubleshoot this from the beginning.

I've been using the dataloglab software and been making some power pulls this last week. It's been pretty cold here and my car seems to like it!

Average hp: 322
Average trq: 312

Pretty nice powerband:

over 275 trq from 3100rpm to 6500rpm.

I still have an issue where my boost drops off a little too much. Going to be installing a Forge WGA in the near future. Once that's done I should have a little more powa! LOL. I'm not sure yet if it's because of the changed VE with my cams or a faulty WGA.

Going to test and tune this Sunday so I'll post my track times. This will be my first time at the track so my ET may be low. I'm more interested in my trap speed.

Thanks to all that helped me troubleshoot this problem.

 

In retrospect I now realize that experience (not only skill) can be major factor in tuning.

Many of the modifications that we do to our cars can cause "phantom knock."

Stiff motor mounts. DP's and exhausts. Stiff suspensions.

All these factors need to be taken into account when diagnosing a problem.

 

I'm glad you got it worked out, I always tell people to check their downpipe clearance because thats the biggest contributor to phantom knock, since the engine can shift and resonate through that pipe, which with most aftermarket downpipes, will transmit the noise directly to the engine block.

An engine damper also helps keep the engine from shifting and keeps the downpipe from making contact also.

I doubt the knock sensor was the real issue, but its cheap enough and worthwhile to replace if your in doubt.

 

We had a car in the shop this week which had a improperly clocked balance shaft (after a cam job at a UNNAMED 3rd shop) - thats a big knocker

 

Uh, imagine that. A mechanical fix to a problem. So you mean what the computer is telling me is knock may not REALLY be knock all along. You mean I may be able to get away with more timing if I can make sure that the knock readings are accurate?

/sarcasm



I'm glad you got it figured out. I'm not trying to be a jerk in this thread towards you mind you. I'm just very glad that someone was not close-minded enough to take what the computer's datalogs were saying as 100% Fact/end of story.

 

What I took from this experience is this:

Real knock will/can pull timing.

Phantom knock will/can pull timing too. It doesn't matter what the source of the knock is - the ecu will do what it wants to do regardless.

Just another layer of difficulty one must face when tuning cars.

You have to log like crazy and analyze the numbers. I've learned a lot from this experience.

 

It is entertaining however, to go back to your original thread and read what some of the -evoM- "experts" suggested was the cause

I shouldn't but,
https://www.evolutionm.net/forums/sh...0&page=3&pp=15

Glad you fixed your problem and thanks for posting the results.

 

Hey that wasn't directed at me was it? I actually mentioned the downpipe and motor mouts in my posts.. LOL..

 

hell no! It was directed at machuang

I'm realizing I'm turning into an a__hole. This clutch install is driving me nut's! It's in the garage on ramps, no help, one jack and the OEM jack, no transportation to work, and so on.

whaaa

Allmost done though

 

Yes you are, Mr. The exhaust manifold studs have oil flowing through them...

 

Well the knock sensor replacement was most likley not needed. Let me explain:

5 years ago I got my DSM logger and chased what was being called "phantom knock". Nobody knew what caused it back then, only that some cars had it and some didn't. I ran race gas, I ran 94 octane pump gas, I ran toulene, I changed my plugs, I changed my wires, I replaced my stock downpipe, I replaced all the bolts. Somebody recommended replacing my knock sensor and I did. What a PITA! With bloody knuckles and a sliced wrist I went for a drive, knock was the same if not worse. So a 12 year old sensor at the time wasn't the culpret.

The problem for me was my HLA's (Hydrolic Lash Adjusters) were ticking. That with a mobil 1 10w/30 oil and a pure one filter cured 99% of the problem.

I can tell you one thing the evo and DSM can put up with a fair bit of mechanical noise. My eclipse just got done running some mid 11's at 130 on half a downpipe that is slamming off the subframe, 3"cutout after the O2 housing, full poly motor mounts, full energy suspension suspension bushings etc.... No phantom knock driving to the track, racing, and driving home.

Sometimes, but it is very rare, I'll get a little blip of knock in the evo when driving normal. The knock sensor seems less sensitive however when I'm full throttle.

Real knock typically increases with the rpm's. False knock typically spikes and slowly decays with rpm's. That is something that has held true 98% of the time for me.

 

Dan -

That is exactly what was happening to me.

I don't know if you want to search for my posts regarding this, but most of the knock was starting off boost. Around 2500rpm. Sometimes it would spike high - like 20 counts or so, then decrease as the pull contiuned.

So, for you, when you got _real_ knock it would typically occur like this:

rpm knock
3000 2
3050 2
3100 3
3150 3
3200 4
3250 4
.
.
.
.
.
5000 10
5050 11
5100 12

Is this what you are saying?

I'm researching/learning so when I start to fiddle with the tune myself I understand what's happening and will be able to distinguish between the two.

 

I still haven't been able to find the answer, but what is typically considered too much knock?

Some say 1 ~ 3 counts are acceptable.

I would think - with all the knock filters and protection in place - once too much knock is encountered the system protects itself and pulls timing. So, the ecu is pretty robust and it can protect itself.

 

Anything over 5 should be avoided, 0-3 are pretty normal, anything over that will alter your octane number or pull timing anyway..

Honestly, although I make a big deal of minimizing knock, it takes quite a bit of knock to do damage thanks to the really good protection the ECU offers.. But you want to minimize the risk.. Also, you will get knock in daily driving, its the nature of how the fuel system works, just focus on open loop high load areas that get knock..

 

Wtf you talking about, I hope you not talking to me. I was just throwing suggestions which said nothing about the tune as seen in the link you posted so if it was directed at me I'd appreciate your keep my name out of your sentences.