Knock Retard hell! Desperate for help

[itsdaveonline]

I need some help. I have thrown ridiculous money to tuners and on parts to fix this but still no resolution. I keep on getting knock retard. It seems KR is most prevalent when throttle is greater than 40% but it's also kind of random, it does not matter if it's in PE or running stoich, i get KR. I'm running 93 octane and this has been happening for a while regardless of where I gas up.

Mechanically speaking here's what I've done & checked

- Stepped up to 7.425 pushrods to quiet down valve train. Measured /verified length on ea pushrod with tools (now running .60xx preload)
- Replaced 1 knock sensor (original looks fine, wiring looks good to both sensors)
- Changed fuel filter
- Upgraded to AFR500 wideband AKA ?NGK AFX?
- A few valvestem seals unseated a while ago. All have been replaced and verified good
- Engine Seafoamed afterwards (brake booster method) although borescope only showed minor carbon deposits if any
- Fuel injector cleaner has been pushed through the system

In regards to the tune, the new wideband says i have been pretty dialed in with the exception of startup and first 30 sec of driving? Everything else looks good in the tune IMO. My wideband is connected to A/C press sensor. Custom PID = 9+[PID.7101]*1.400 (according to BMR motorsports site). PS it?s also calibrated. And yes, VVE tables are tuned properly too.

I don't see any smoking guns and I desperately need your help. I'm tired of driving like a grandpa to avoid KR. My timing is pretty much stock too. Please help!!!

Car specs
2006 CTSv
Stock LS2 w/ stg 1 tick performance cam
longtube headers
LS7 clutch
NGK TR5IX plugs

[GHuggins]

Under Spark - Idle Spark Advance - Coast down - put the stock table back in here... Only had access to a ls2 vette file and it's much lower in the decel areas. Looks like your spark knock nearly always happens when or right after you let off the throttle and the ignition timing jumps up in the 30's. You might also want to correctly tune your ve table. Lean spot in the 3000 rpm area can also cause issues during fueling transitions. Some things worth trying anyway

[10_SS]

I only saw maybe 3 spots of slight knock in the entire log.

[itsdaveonline]

Thanks guys

@greg - I attached the stock tune, but the Idle adaptive spark controls I'm running are already stock values??! VVE table has been tuned and should be dialed in. Are you suggesting that the VVE table needs more work? Not trying to argue with you just curious

@10_ss - correct, KR is not terrible in the datalog. But if I lay down on the throttle more , the KR will occur more frequently and will generally be more severe. I didn't run the car that hard on this run and didin't want to risk the motor. I also monitor my laptop during the datalog, and if I see KR i back off the throttle immediately. You guys may be seeing that in the log too.

[GHuggins]

Dave, I only had a ls2 vette for comparison. Curious though. It's got longtubes and longtubes are known for causing picked up "false knock". Is this actual or just seen on the scanner knock? I only noticed it in the log happening during decel or nearly immediately after letting off the throttle.

Does pulling timing out in these areas actually remove the perceived or seen knock? You may just need to alter your knock tables a little if they haven't been?

Not sure if you have beta and the ability to look at your VE table or not, so I'll post a pic for you how it shows it when I look at it...

[schpenxel]

That looks... interesting...

[itsdaveonline]

Dave, I only had a ls2 vette for comparison. Curious though. It's got longtubes and longtubes are known for causing picked up "false knock". Is this actual or just seen on the scanner knock? I only noticed it in the log happening during decel or nearly immediately after letting off the throttle.

Does pulling timing out in these areas actually remove the perceived or seen knock? You may just need to alter your knock tables a little if they haven't been?

Not sure if you have beta and the ability to look at your VE table or not, so I'll post a pic for you how it shows it when I look at it...

What do you mean by actual knock vs scanner knock? I mean I definitely see perceived KR in the cars computer. It went ape shit pulling timing when Burst knock was enabled. But audibly speaking I can't hear any detonation or anything obviously wrong if that's what your asking

I suppose I can try pulling timing in spots where KR is seen, but i mean I'm running stock timing and a degree less than stock in some spots. Only KR related items touched are burst knock disabled and max step went from 6 to 4 otherwise factory config.

[itsdaveonline]

That looks... interesting...

Indeed. Here's the stock one for reference.

stock ve.png

I thought i was only concerned over the "manifold switch open" VE as i don't have DoD. below is a pic of manifold switched open table. The manifold switch closed table is the weird one with craters in it. So which one(s) do i adjust?

ve_current.png

[schpenxel]

What are you using to change SD coefficients? 2.25 beta or blue cat?

[schpenxel]

I just downloaded your file, smoothed it out, calculate coefficients, and it barely added a bump at all..

[itsdaveonline]

Using the 2.24 beta. Maybe it was if i copy paste coefficients themselfs from my "VE tuning tune" to my real tune is when that's happening. IDK for sure. Either way which VE modes should i be tuning lol. I will look into fixing those craters either way.

[schpenxel]

I just make all 4 the same so I don't have any chance of it using one I didn't change.

I don't really understand why you'd need to do that with a "real" tune vs a VE tuning tune? The VE editor updates the coefficients automatically. Just update the VE table like you would normally on any other tune, hit calculate coefficients and you're done.

If you're copying old coefficients out of an old tune and putting them in a new one.. then that's why you keep getting the old VE table every time. Those coefficients ARE the VE table.

What I do is after I hit calculate coefficients on the first one I select the whole table, copy it, then go through the other 3 tables, paste the VE table, hit calculate coefficients on each one. Sounds like a PITA but doesn't take long and makes all 4 the same

[itsdaveonline]

Gotya - okay so based on the feedback from both of you this should be my plan of attack to combat my KR issues

1) use the good VE table, copy it into the other "VE modes"
2) pull timing (hesitant to do since im at stock values and below but i'll test)

anything else i should consider doing? thnx!

[GHuggins]

Dave just looked at the two tunes. There's not a whole lot different in your "paid for" tune vs. the OE? It shouldn't be spark knocking at all unless your cam is something out of a stock blower motor or something? By that I mean it has "no overlap"... If it had overlap and 15 or 20 degrees or more, you should need to add considerable timing at least in the 1200 to 3000 rpm range. This would be because of the static and dynamic compression losses... Now with that being said cams sometimes do better with it removed in this same area - typically this has more to do with incorrect fueling than it does anything else - by this I'm referring to the fact as to whether or not the O2's were dialed in for "their" new locations and whether or not the injection timing has been dialed in to keep the fuel in the combustion chambers... Requires a dyno to know for sure whether adding or taking out timing helps... With all that being said I really think your just seeing false knock... Whether it be due to the fact that they're sensitive from the factory or it's due to the "noise" coming from the headers, I get the impression that your engine isn't actually knocking from what your telling us? Again only way to know for sure is on a dyno... But with the two most obvious facts your telling us "it's got a cam and headers", it shouldn't be knocking... Again that depends on the cam, but 70 to 80some percent of the time cams typically like timing because people go for "lopey" cams... Then headers as long as they're not wrapped or coated usually get heat out and backpressure down which usually further aids in being able to add more timing...

Only real thing I can advise is to dial in the O2's for the headers and injection timing for the cam if they haven't been - this will change fuelling a little, so you'll need to dial in your VE and MAF tables again... Then on a dyno or using a road simulation virtual dyno of some sort, dial in your timing and knock sensors... There are lots of methods for doing this. I multiply the multipliers for the individual cylinders by .05 at a time until it still will show knock just before it's peak power output. You can also raise the initial level "this is what most out there do" or modify the knock attack rate gains... Again can not prioritize this enough - Do Not do any of this without a dyno or some sort. Engine Damage can and most likely will occur!!!

Hope this helps...

[CactusG8]

Dave just looked at the two tunes. There's not a whole lot different in your "paid for" tune vs. the OE? It shouldn't be spark knocking at all unless your cam is something out of a stock blower motor or something? By that I mean it has "no overlap"... If it had overlap and 15 or 20 degrees or more, you should need to add considerable timing at least in the 1200 to 3000 rpm range. This would be because of the static and dynamic compression losses... Now with that being said cams sometimes do better with it removed in this same area - typically this has more to do with incorrect fueling than it does anything else - by this I'm referring to the fact as to whether or not the O2's were dialed in for "their" new locations and whether or not the injection timing has been dialed in to keep the fuel in the combustion chambers... Requires a dyno to know for sure whether adding or taking out timing helps... With all that being said I really think your just seeing false knock... Whether it be due to the fact that they're sensitive from the factory or it's due to the "noise" coming from the headers, I get the impression that your engine isn't actually knocking from what your telling us? Again only way to know for sure is on a dyno... But with the two most obvious facts your telling us "it's got a cam and headers", it shouldn't be knocking... Again that depends on the cam, but 70 to 80some percent of the time cams typically like timing because people go for "lopey" cams... Then headers as long as they're not wrapped or coated usually get heat out and backpressure down which usually further aids in being able to add more timing...

Only real thing I can advise is to dial in the O2's for the headers and injection timing for the cam if they haven't been - this will change fuelling a little, so you'll need to dial in your VE and MAF tables again... Then on a dyno or using a road simulation virtual dyno of some sort, dial in your timing and knock sensors... There are lots of methods for doing this. I multiply the multipliers for the individual cylinders by .05 at a time until it still will show knock just before it's peak power output. You can also raise the initial level "this is what most out there do" or modify the knock attack rate gains... Again can not prioritize this enough - Do Not do any of this without a dyno or some sort. Engine Damage can and most likely will occur!!!

Hope this helps...

[CactusG8]

Would a cooler plug help?

[itsdaveonline]

Thanks. Lots of useful info here. I'm running the stg 1 tick performance sns cam. so it has 7 deg of overlap 227/231 on 111 LSA, around .59x & .61x lift. Definitely not a blower cam by any means lol. And yes, I don't hear any knocking, but my exhaust and valvetrain is so loud it's hard to tell. Headers are not wrapped or coated either, so shouldn't be any heat "trapped".

To be safe ill be running conservative timing until I can figure out a tuner / dyno. Here's a diff comparing the stock and revised tune. Note how im either at or below stock values.

timing diff.png

EOIT tuning seems like alchemy. I have seen plenty of spreadsheets which don't work IMO. I've attempted to adjust it into the ballpark but i have no means of confirming that it improved the car. I even called tick performance at one point and they state to not mess with injecting timing even after installing their cam. Mind sharing a link or article on dialing the 02's for headers?

[schpenxel]

Did the longer push rods help with valve train noise?

EOIT is, in fact, alchemy, lol. Ghuggins has a way of getting it going in the right direction. I'll let him post it if he'd like

[itsdaveonline]

Yeah i was running hardened 7.400 pushrods and had maybe .35x preload - approx 3/4 a turn till 22 ft/lb. 7.425 got me to around .60x preload and did quiet it down a little bit. It still has some sewing machine noise, but not nearly as bad, I don't want to run ultra-high preload or deviate much from stock pushrod length. Im thinking the base circle of the aftermarket cam was slightly smaller than stock. I did not measure to confirm though!

I may just try to tune with him, seems like from some light reading he has a good shop. But i still am running colorado plates with expired tags, so gotta figure out this inspection crap which is new to me. Then maybe i can make the 5hr trip to cashiers NC.

My EOIT is advanced 23 degrees currently (boundary set to 497 as opposed to stock 520) my thought process is that I want to emulate the stock cam and injection values to where it injects fuel on the intake valve right as it opens for best atomization. With new cam, the intake valve event happens sooner, so injection is set to happen sooner but not too soon to where it blasts right outta the exhaust during overlap. But that's just all theroy at this rate though! I read for days and those are the values i came up with.

[schpenxel]

Ghuggins knows more about injection timing than me....

I'd try setting it up to inject just before (or maybe even just after) the exhaust valve closes and see how that does

Typically if it runs richer (fuel trim values will go negative) then that means it's making better use of the same amount of fuel, which is the way you want to go, in my opinion anyways. When I was on the stock cam increasing the boundary table to I think I ended up at 530 or 535 made an obvious impact on fueling.. I had to bring the VE table down a good bit as it was going quite rich. Mid-range torque felt better as well

My cam has no overlap (blower cam) so I haven't spent much time messing with injection timing since it can't blow fuel out the exhaust even if it tried (unless I REALLY break something, lol)

Here's what mine did after increasing the boundary table about 15*. This was with the stock cam.