Hal Fisher

Just a quick post on what I'm going through to find a source of false knock that usually happens over 4k rpm. I used TunerCat for OBD1 adjustments. For data recording I used Datamaster (obd1), and AutoTap (obd2). I tried both ODB1 and ODB2 PCM, no difference. I dropped WOT advance to below 15 (stock is 33) on the OBD1 pcm, no help. I tried a new OBD2 sensor but both PCM still showed knock (knock counts went up each time by over 300). I then used a box that adjusts the sensitivity of the knock sensor and had to go to 30% less before it stated tapering off (would see 1 or 2 degrees less advance sometimes). I know there must be something causing this but nothing can be noticed when driving, it doesn't even feel that dramatically slower though I've never timed them.

I'll check the motor mounts one of these days, not really looking forward to that.

Hal

MikeGyver

Try unscrewing the knock sensor from the block and plugging the hole (which has coolant behind it).
Hang the sensor in the air so it wont hit anything and electrically ground it to the block using an aligator clip.
This way it's still 'there' but cannot physically detect anything. See if it still knocks and go from there.

Injuneer

Are your data logs showing knock retard? If so, how much?

Changing knock counts are not "knock retard". Its very normal for knock counts to increment without causing knock retard.

Hal Fisher

I'll try disconnecting the sensor in the block and connecting to the other sensor I have. I can check this tomorrow.

Injuneer, I think I have one recording out of a hundred with high knock counts and no retard, it's that rare on my car. OBD1 with Datamaster or Freescan record a high of 36 at WOT while Autotap and OBD2 record the highest of 34. I don't have the version of OBD2 Autotap that shows actual retard but OBD1 and Datamaster does show actual retard. I have TunerCat for the OBD1 as well and see advance going down sometimes 10 degrees even with WOT set to only 15. on I'm pretty sure both pcm might record a max of 36 but the OBD2 was still seeing minor knock even with my desensitizer circuit at it's maximum of 30% (two adjustable pots allow adjustment from 1% to 30%). The OBD2 might need 35% for some reason but is close enough.

I really do want to track this down instead of probably rendering the knock sensor worthless with 30% desensitizing. It could just be lifter tap that I've been chasing as well. This might cause just the right vibration at high rpms to make it think it's knock. The tap isn't that bad but noticeable at idle. I've changed lifters (LS7), springs, push rods, adjusted and tightened the rockers many times, but nothing seems to completely solve it.

Hal

Injuneer

I've never seen a PCM tune that allows 36-deg of knock retard. Even a factory tune limits to a max of 15-deg. Who tuned it?

Hal Fisher

Oh, I meant 36 degrees advance, not retard. The desensitize circuit was letting it get to the 36 degrees of advance.

Hal Fisher

I hooked up the extra ks and did not see any retard during WOT. So it is certainly something the ks are hearing.

The real test for the ks is to set timing way up to something that causes actual knock at WOT and see if the ks desensitized by 30% sees it. While the engine is still stock (no supercharger yet) it should be able to handle it without blowing a ring land.

Hal

96LT1TX

I have had "false knock" for ~ 3 years now (since I bought my new cam). I gave up chasing it down and turned "max knock" in the tune to 0. I realize you are trying to go FI, but for me it got old trying to track it down after a year or two. I tried everything too... I had max knock at idle though and it would drop when I pulled my foot off the gas... no sense.

Casey_SS

I've been through the ringer on this one as well although the specifics are little different. In my case, I was getting KR (both counts and actual retard) only at 3400 rpm under all conditions that included more than 20% TPS. Spark advance didn't affect it in the least....got the same results from 8 degrees total advance all the way up to 30 degrees and everywhere in between. In my case, the culprit seems to be resonance in the headers (no way to confirm but that's the only viable theory left after going through great pains and expense to eliminate everything else). My final solution was to cross my fingers and disable the knock sensor from 3200 - 3600 rpm but leave it intact everywhere else. I've been roadracing it on 93 octane in the Texas heat for well over a year since then with no issues. To put that in perspective, figure 8 20 minute sessions of mostly full throttle abuse per weekend and I run roughly once a month...that's well over 30 hours of the harshest use case I can think of for one of these motors and I'm doing it in temps ranging from 85 - 105 degrees (not counting a couple winter events). If there was a problem I would certainly have found it by now.

I'm not at all suggesting this is your issue, just a data point to consider if you're running headers. Not sure if you are or not... Mine are Pacesetter longtubes for reference but are on a 396 so the exhaust volume needed to create whatever frequency pisses off the knock sensor will undoubtedly be different. If you buy into this theory, it would make sense that on a smaller motor this particular frequency would occur at a higher rpm, all else being equal. Again, I'm not saying this is the culprit in your case nor am I suggesting you disable the knock sensor in any way. Just one possible explanation out of many. Take it for whatever it's worth...

96LT1TX

This is a good post. Thanks for the reply. I am not trying to thread hijack by any means, but my false knock seemed to be from my headers / exhaust as well. I believe my y-pipe may have been very close to the under side of my floor pan. I recently upgraded to some pacesetter longtubes and a pacesetter y-pipe and have yet to run a new data log. I will run this and see if the knock issues have changed any.

Previously when I was getting knock at idle and it would decrease past 2500 rpms and increase again past ~4500 rpms, I was running Mac shorties.

I will repost here when it gets warm again and I can drive my car.