Why am i still getting knock retard?! this is insane

[itsdaveonline]

As for SD mode my VVE is only tuned till about 4500 RPM, since at that rate the MAF will be doing all the fueling calcs...

But from my experience it does have alot less or no knock in SD & OL mode. I'm just weary to do knock testing & WOT pulls in SD & OL i assume PE will still function as normal and will kick in with no issues?

[schpenxel]

Yeah it will work fine in SD and OL assuming VVE data is accurate

[itsdaveonline]

Cool. I'll muck around with OL & VVE stuff later.

I think first I'll try logging the individual cylinder knock data first. Probably won't do much good with numbing the sensors down and running the 100 octane. But who knows!?!

[mourat]

Turn off transients, turn on OL & SD I expect no KR. If that's the case start there. Make sure your bleed off hose is not clogged (the one on the front of the heads joining the two) also helps to add a second one to the back bleed steam off the front and back in case you have a hot spot.

What car is this?

[jtrosky]

Wow - that is a lot of KR - especially being that you are looking at average values! It would also be helpful to review the Max KR values and the counts.

Also, one more question - are you only reviewing data from the first run after making tune changes? I swear that I get significantly more KR for the first few trips after making a tune change - I know that Knock Learn will help avoid KR in the future, but even when Knock Learn is not applying any corrections, I seem to get more KR on the first few trips.

You might want to try at least a few different runs (after cycling the car off an back on) after making the changes. I can't explain why, but like I said, I always seem to get more KR during the first trip or two after applying tune changes.

Also, since fueling is automatically "kept in line"while in closed loop (via fuel trims), fueling tuning shouldn't really have any impact on KR in closed loop, should it, since the fuel trims ensure the proper fueling? I mean it's not like you can be running super rich or super lean while in closed loop... At least that's the way I understand it. Although, I guess if your commanded AFR is way off, your fueling could still be messed up even in closed loop...

[GHuggins]

Wow - that is a lot of KR - especially being that you are looking at average values! It would also be helpful to review the Max KR values and the counts.

Also, one more question - are you only reviewing data from the first run after making tune changes? I swear that I get significantly more KR for the first few trips after making a tune change - I know that Knock Learn will help avoid KR in the future, but even when Knock Learn is not applying any corrections, I seem to get more KR on the first few trips.

You might want to try at least a few different runs (after cycling the car off an back on) after making the changes. I can't explain why, but like I said, I always seem to get more KR during the first trip or two after applying tune changes.

Also, since fueling is automatically "kept in line"while in closed loop (via fuel trims), fueling tuning shouldn't really have any impact on KR in closed loop, should it, since the fuel trims ensure the proper fueling? I mean it's not like you can be running super rich or super lean while in closed loop... At least that's the way I understand it. Although, I guess if your commanded AFR is way off, your fueling could still be messed up even in closed loop...

The knock retard after flashing is a great point. Give it 15 minutes before logging anything... Some tune changes can and will make it log excessive knock after flashing.

As for the closed loop vs open loop vs sd, this is why I was wondering about base fueling. If you see differences while in sd or open loop then most likely your O2's aren't dialed in correctly for something such as headers and O2 relocations... It may "appear" that air fuel is something close to what it should be when in fact it could be cycling .1 off from where it should actually be and cause more knock... Seen that one quite often...

[schpenxel]

Turn off transients, turn on OL & SD I expect no KR. If that's the case start there. Make sure your bleed off hose is not clogged (the one on the front of the heads joining the two) also helps to add a second one to the back bleed steam off the front and back in case you have a hot spot.

What car is this?

I've been thinking about adding rear steam vent hoses on mine (LS2). Hmmm

[itsdaveonline]

hmm even if my tune adjustments were off for longtube headers and o2 relocation, that still shouldn't contribute to KR when in PE right? That would definitely make sense outside of PE if the o2's "swings" were being incorrectly deciphered. Seems i get KR both in and out of PE.

I adjusted the closed loop proportional a while back (multiplied by .8) but my PCM does not have integrator delay available. So i don't think i can truly dial in the LT's 100%.... I honestly cant say that the proportional settings have helped or hurt.

For the other guys helping vehicle is 06 ctsv w/ ls2. I'm not familiar with the bleed off hose? Unfortunately the 93 log was not after a fresh flash, but good info on expecting more KR after flashing... i will keep that in mind.

I really appreciate evreyone batting various ideas. This is the biggest pain in the rear to tune ever.

[mourat]

My guess is you have a bad coil / spark plug or wire because that's too much KR. Try swapping coils if you know anyone that has a set. Do plugs at the same time I run TR8's here in Canada no problem.

[itsdaveonline]

Hmm i mean the car seems to fire on all 8, spark plugs seem to confirm that, and mysteriously enough i still see no evidence of detonation / "peppering" on the sparkplugs. The motor seems smooth, but not dismissing the possibility of a weak coil or maybe intermittent fire. But if that were the case wouldn't I see the same amount of KR with the 100 octane?

I'm having a really difficult time finding how i can test each coil for dummies. I've found some overly complex guides, but they assume you have an OEM scan tool. I don't

[jtrosky]

I think somone already mentioned this, but maybe trying logging knock per cylinder. I don't know much about your engine, but my LFX V6 can log knock per cylinder. If you are only seeing knock on certain cylinder(s), that might point to a plug/coil issue like mentioned above.

[itsdaveonline]

So bringing this thread back. I logger per cyl knock retard for all 8 cylinders. It just shows knock on all cylinders... Im not sure what that means. But im having a hard time finding any meaningfull data from this. Placing the data in histograms did not sow any trends either. I also am now -1 & -2 degrees from stock timing.

[sssnake]

Dave,

A couple of questions:

1.) Have you checked the knock sensor torque values? 15 ft lbs is the spec. How much confidence do you have in the torque wrench accuracy/has it been calibrated lately? Over torquing this sucker will definitely make it more sensitive. If you are using an HF torque wrench bite the bullet and buy a good one. They are horribly inaccurate a low torque settings (My son worked for HF for a while during college).
2.) Have you checked to see if the exhaust is touching the frame? At one point I thought this was the case. If it still is, you MUST fix this to get accurate knock sensor readings. The knock sensors are acoustic so any noise that fits the frequency profile of a ping/pre-detonation/knock will register as knock.

I know the second potential issue should not change with the addition of torco but the first definitely would.

On the exhaust pipe issue, I had the drivers side header touching the steering shaft occasionally under big throttle mashes. It drove me nuts for a LONG time. In this case there were specific conditions that had to be met before it was an issue (it REALLY looked like honest to God knock - low RPM high load - but was actually false). I would check the clearance there as well. Are you running the Kooks 1 7/8th"? If so, they are very close to the steering shaft. You might want to place a go pro under the car to see if you have anything odd going on.

The only other thing I can think of that might be generating false knock would be the pinion bushing. With the amount of deflection we get under hard acceleration you could be getting the rear end housing into the floor pan. This would be a LONG SHOT - IMO but easy enough to check with the go pro.

Well if you don't get this thing figured out by fall let me know (which I certainly hope you do). We will be making another Dragon trip and I could head on out to Raleigh to see if anything was obvious.

Charles

[GHuggins]

After you verify it's not something rattling, you should probably pull 3 degrees at 4800 then 2 at the two neighboring cells and then smooth just until you can get it to a dyno. This is the rpm range where your going to see peak cylinder pressure and the highest dynamic compression and is where all of the knock was in the log. You may just have a really good "filling" cam and setup with this thing?

[10_SS]

Yeah im running stock timing in bulk of the areas and in some spots 2 degrees less timing than stock. Removing 10 deg more could be devastating lol. I actually did inspect the cats not long ago. They are not clogged and clean as a whistle.

Took care of a Hemi with tons of knock. I found by removing 1/2 of the knock the system was reporting, all knock went away, which makes sense since once you start knock, the excess heat and pressure start a chain reaction. SO your logs show 3deg retard occasionally.. So pull 1.5 from those cells, see what happens. This guy has been driving the truck for about 1-2yrs with all this knock, now he swears he's getting noticeably better MPG, and much more power which I did notice. Point is, dont be afraid to pull 1-2 deg timing, or less. Drive a couple week and keep an eye on it. In his case, it was bored once as part of a dealer funded rebuild which I think increased timing, and while he had ~5 deg of timing in most all driving conditions, only 1.5-3deg was required to make all knock go away.

[jlpearl]

Took care of a Hemi with tons of knock. I found by removing 1/2 of the knock the system was reporting, all knock went away, which makes sense since once you start knock, the excess heat and pressure start a chain reaction. SO your logs show 3deg retard occasionally.. So pull 1.5 from those cells, see what happens. This guy has been driving the truck for about 1-2yrs with all this knock, now he swears he's getting noticeably better MPG, and much more power which I did notice. Point is, dont be afraid to pull 1-2 deg timing, or less. Drive a couple week and keep an eye on it. In his case, it was bored once as part of a dealer funded rebuild which I think increased timing, and while he had ~5 deg of timing in most all driving conditions, only 1.5-3deg was required to make all knock go away.

We run into this scenario often too, it is common when doing a custom tune for something that had a handheld installed. It's amazing how many handhelds just throw a bunch of timing at the motor and let the knock sensors kick in. We often power by removing commanded timing.

TLDR; don't worry about removing a little timing