balaguru

iTrader: (2)

I frequently hear a short run of knock at moderate loads (e.g. ~50% throttle in 6th at ~60mph). I am using passive headphones wired to the knock sensor via an amp. Knock strength as logged by the ROM editor does not increase when I hear this "soft" knock. I was a little concerned to hear as much knock as I was hearing, so I first verified that it was really knock that I was hearing by inducing knock at lower loads (e.g. ~30% throttle in 3rd at ~30mph) by advancing timing across the board and going for a short, slow drive. The sound that I hear sounds like a soft crackle/clipping. Unlike before, when I induce knock the logged knock strength increases with the knock that I hear from the phones. The following sources seem to suggest that some knock is acceptable and "soft" knock might even be preferable.

http://www.contactmagazine.com/Issue...ineBasics.html
http://www.freepatentsonline.com/4337642.html

Any thoughts before I try to tune this knock out?

djamps

iTrader: (1)

I had a similar knock headphone setup and was able to get rid of the audible ping except for high boost where it always pings a little (a little on headphones but not noticeable outside the car). I don't think it's normal to hear it in part throttle, unless you're hearing something else -- VQ's are very noisy. Any way you could post a MP3 sample?

balaguru

iTrader: (2)

Unfortunately, neither my phone nor my laptop want to record properly with a standard TRS phone jack. Gonna try a different app on my laptop to see if I can get a recording.

balaguru

iTrader: (2)

Looks like I'll have to make a TRRS adapter to get it to work since both my laptop and phone use the same jack for mic and speakers. I'm fairly sure what I am hearing is knock but couldn't find any samples to compare to. That's why I induced some knock. The only difference I can remember between the induced knock and the part throttle knock was that the induced knock was a bit louder and more frequent. Qualitatively, they sounded they same.

Resmarted

iTrader: (2)

Phantom knock possibly.

My friend's 2j had some wickedly strange phantom knock outside of boost at very odd rpm, with low load. Did a leakage test and found one cylinder was leaking 50%. Haven't had time to catch up with him in a while but we assume a ring got facked sideways.

djamps

iTrader: (1)

As per resmarted's comment, I'd send an oil sample to blackstone to rule out any serious issues before you spend a bunch of time tuning it.

balaguru

iTrader: (2)

Not a bad idea. I may just do that. Was also going to pull a plug or 2 to see how they look. It's odd though as it occurs in the vacuum portion of my map where I am either running stock timing or slightly retarded timing. I had put a few 1000 miles on the car since the turbo was installed. A built short block was eventually planned but I was hoping to put it off till next year. Car actually drives great otherwise. Pulls hard in vacuum and boost--no smoke, no hesitation. Even the death at idle miraculously disappeared on its own. But still have not really pushed it much because I was waiting to have the knock phones setup before I did. Now I'm back to just keeping it in vacuum.

djamps

iTrader: (1)

Also check A/F and make sure it's not too lean, that could also cause constant knock. Should be 12's between -3 and 0psi and 13's anything over 25% throttle. If it's wandering into or above 14's while you're accelerating that would be a problem. Do your plugs have white tips?

balaguru

iTrader: (2)

Haven't had access to the car yet to pull the plugs but I did lean out the vacuum portion of my map. That could be contributing. Attached is a fuel map made from my datalog. It maps boost, target afr, actual afr and knock on the uprev fuel fuel map.

djamps

iTrader: (1)

Pretty lean in the lower load areas (14-15) and even low boost... that's probably it. Bring anything over 25%-50% throttle and especially near or greater than 0psi to at most 13's. Work up from there.

balaguru

iTrader: (2)

Aww...crap.

I enrichened the lower load cells and redid the timing table and I've virtually eliminated knock via the phones. Also saw knock strength fall from the 1000's to near 0. I had driven nearly 3k miles on that tune. Whatever damage was done is done. I pulled a plug and it didn't look bad. Here it is.



Here's my old timing data:



Here's the new timing data:



Lessons learned:
1. if you are going to tune an FI car for the first time install a knock listening device asap.
2. on at least a 2004.5 ECU knock strength numbers in the 1000's should cause pause.

djamps

iTrader: (1)

Probably fine. My Z's plug tips were solid white and eroded from running lean with a failed knock sensor harness for years before I got it. Luckily it was overly rich in boost (no hard knock) and everything ended up fine after I replaced the knock sensor harness and retuned it. Your plugs look virgin compared to the ones I pulled. Do a compression test to be sure!

balaguru

iTrader: (2)

Abnormal compression test itself wouldn't change what I do with the car in the near term. Just need it to be my daily driver till december when I pick up a tow vehicle which can also serve as my backup. Plan to follow oil analysis for a few changes. Should clear after my next change. Will maybe combine a compression test with some other mod like polishing the plenum or something at that time if indicated by oil consumption, etc..