L96 Spark Knock

[Dutch]

I'm not used to hearing spark knock. I only heard it in person on an old integra, some very sad lawnmowers, and on a friend's F150. In those cases it sounded really bad - like rod knock mixed with a cold diesel engine idling.

I may have accidently induced audible spark knock by merging the knock sensor thresholds from an 2013 E78 Chevy 3/4 ton to my 2013 E38 Suburban 3/4 ton.

The L96s with the E78 ECM look like this:
Capture1.PNG

My E38 Suburban 3/4 ton looks like this:
Capture2.PNG

I merge the config to get this:
Capture3.PNG

And also adjusted all other knock retard parameters like transient knock, burst, etc to match the E78, and raised the Max Valid Retard to 10.

And then as a result, I hear this:
https://drive.google.com/file/d/19ZC...ew?usp=sharing

Same video with the audio adjusted:
https://drive.google.com/file/d/1pdv...ew?usp=sharing

When I revert the thresholds back to stock, the noise disappears.

I have not included tunes or logs because I didnt actually log the event - I reverted it back after recording that video and went with a different approach.

My questions are
1) Is the sound you hear spark knock? If it is, how bad is it?
2) What would explain the drastic difference in threshold between the two tunes?
Does the E78 handle knock differently behind the scenes (outside of whats visible in hptuners) than the E38?
Or does GM simply consider this level of knock acceptable?

[Dutch]

You know what, forget everything I said except for the first question.

Is the sound you hear in the video spark knock or not? If it is, how much can a L96 generally tolerate?

[5FDP]

Most stock 6.0's always show a lot of knock retard when running 87 octane. Not un-common for me to see 2-7 degrees on average when under load with stock power enrichment settings and timing curves.

Ive never looked into why the tables are set different and wouldn't have ever messed with them in the first place. The E78 computer uses a whole different logic behind it's entire operating compared to the E38's.

[Dutch]

Ok this thread has gotten way more views than I anticipated.
I guess I should at least answer the first part of my question.

The sound you hear in the video is indeed spark knock - induced by excessive spark advance (added more fuel without checking the PE spark adder), made worse by mild carbon buildup.
We all know what spark knock is: abnormal combustion causing violent spikes in cylinder pressure that, in severe cases, could go beyond the mechanical limits of the piston and cause it to fail. Knock severe enough to be audible over engine noise close to redline is unacceptable for this reason, and of course in the example shown in the video, it has been dealt with accordingly.

But the more interesting question is the second one: what is considered acceptable levels of knock for this engine?

I kinda wanted this to be a discussion. I'm not too familiar with the LS family of engines, so I'd like to hear what you guys were able to get away with in your experience in modifying this platform.

Most stock 6.0's always show a lot of knock retard when running 87 octane. Not un-common for me to see 2-7 degrees on average when under load with stock power enrichment settings and timing curves.

Ive never looked into why the tables are set different and wouldn't have ever messed with them in the first place. The E78 computer uses a whole different logic behind it's entire operating compared to the E38's.

I understand occasional KR in the logs is normal. What Im asking is if and when audible spark knock is normal.

You mention its not uncommon to see 2-7 degrees being pulled under load. If you look at the max retard table, below 3500rpm, thats more or less the maximum amount of timing the ECM is allowed to pull.
maxkr.png

Assuming knock intensity corresponds to the amount of timing that needs to be pulled to suppress the event, if a 12 degree knock event occurs and only 6 degrees of timing is pulled, the intensity would be the equivalent of a 6 degree knock event that the ECM took no action against. This would surely produce an audible knocking sound.

In a way, this means that the E78 knock threshold tables are actually safer than the E38 tables: by raising the floor, the ECM is able to respond to actual knock events, rather than overreacting to noise and then hitting the max retard ceiling when knock occurs.

The way this table is set implies that some level of audible knock is tolerated at low RPMs.

Which again brings me back to my second question: what is considered acceptable?

The service manual defines spark knock as "a mild to severe ping which usually occurs worse while under acceleration. The engine makes sharp metallic knocks that change with throttle opening"
Screen Shot 2021-10-09 at 12.09.58 PM.pngScreen Shot 2021-10-09 at 12.10.16 PM.png

The owners manual is incredibly vague about it. It doesn't use language like "light spark knock is normal", but it also doesn't specifically say its dangerous:
Screen Shot 2021-10-09 at 11.27.03 AM.png

I like Nissan's explanation the best:
Screen Shot 2021-10-09 at 11.38.47 AM.pngScreen Shot 2021-10-09 at 11.38.55 AM.png
Even though the way they handle knock is reckless compared to GM.

A ford dealership told my friend this was acceptable:
https://www.youtube.com/watch?v=iBEfW-1ue88

Obviously I can't just follow other manufacturer's recommendations, but they give me clues as to what is acceptable and what isn't.

So I ask whoever's reading this, in your experience in tuning/building/etc LS engines, particularly the L96 or LY6, what do you consider acceptable in a naturally aspirated build making conservative power?

[cadillactech]

Audible knock is not normal. By definition spark knock(detonation) is abnormal or erratic combustion. You don't want to hear it, stock or modified. Best case you are just losing power, worst case you are damaging parts.

That being said, it happens. In stock form, usually because of carbon build up in the chamber or shit fuel. On a modified engine, usually because whoever calibrated it didn't spend the time to calibrate it correctly or thoroughly enough. Even if you get shit fuel, there should be some safe guard built into the calibration(low octane table).

What is acceptable?

Depends on who you ask and the application. Detonation in a lawn mower engine is much different than detonation in a 1000hp forced induction V8. Cylinder pressures vary greatly between applications so the tolerance of knock before damaging something can vary as well.

I would do all you can to have no knock retard under a full power pull.

Is your engine stock?

[blindsquirrel]

Spark knock/ping won't break anything, usually. Detonation will. They are fundamentally different things. The one that will break things, detonation, will break things quicker than the knock sensors can react and pull timing. Spark knock/ping is something the OEs want to avoid because it's terrible for tailpipe emissions, not because of engine longevity.

Ever owned a stock late-'70s/early-'80s vehicle with nonfunctioning EGR? They would sound like a bunch of marbles in a blender and never put the pistons into the oil pan.

[2xLS1]

Google pre ignition vs detonation. They both have totally different causes and both can break things.

[Dutch]

Spark knock/ping won't break anything, usually. Detonation will. They are fundamentally different things. The one that will break things, detonation, will break things quicker than the knock sensors can react and pull timing. Spark knock/ping is something the OEs want to avoid because it's terrible for tailpipe emissions, not because of engine longevity.

Ever owned a stock late-'70s/early-'80s vehicle with nonfunctioning EGR? They would sound like a bunch of marbles in a blender and never put the pistons into the oil pan.

How much timing were you able to get away with on these L96s on regular (87 AKI) gas?
Does a richer PE mixture (AFR in the mid to low 11s instead of high 11s) help you squeeze a little bit more out of it, or does it not matter?

[Dutch]

Audible knock is not normal. By definition spark knock(detonation) is abnormal or erratic combustion. You don't want to hear it, stock or modified. Best case you are just losing power, worst case you are damaging parts.

That being said, it happens. In stock form, usually because of carbon build up in the chamber or shit fuel. On a modified engine, usually because whoever calibrated it didn't spend the time to calibrate it correctly or thoroughly enough. Even if you get shit fuel, there should be some safe guard built into the calibration(low octane table).



What is acceptable?

Depends on who you ask and the application. Detonation in a lawn mower engine is much different than detonation in a 1000hp forced induction V8. Cylinder pressures vary greatly between applications so the tolerance of knock before damaging something can vary as well.

I would do all you can to have no knock retard under a full power pull.

Is your engine stock?

Have you dealt with LSs that had audible knock? How bad was it compared to what you heard in the video clips?
(The truck in the video is fine by the way )

[cadillactech]

Have you dealt with LSs that had audible knock? How bad was it compared to what you heard in the video clips?
(The truck in the video is fine by the way )

I have heard audible spark knock on L92 and L96 engines quite a lot due to carbon build up in the combustion chamber(from poor quality fuel) and 87 octance being used instead of 92 octance. Filling with 92 octance and doing a top engine clean usually restores them to their former glory and no knock is noted. I have seen anywhere from 2-10 degrees being pulled due to those two factors.

You can use a richer PE ratio to quell knock to a degree. Sometimes you can add fuel and see if adding back in the timing will increase power. Sometimes you make the same, sometimes more but its all something you should play with on the dyno to see what the engine likes.

If your AFR has an 11 in front of it for a conservative NA build, you are too rich. Shoot for mid 12's at the start and adjust from there.

Glad to hear your truck is fine!

[cadillactech]

Google pre ignition vs detonation. They both have totally different causes and both can break things.

You are definitely right, every manufacture that has a small displacement turbocharged DI engine is dealing with LSPI. You can't hear the preignition but it breaks pistons all the time and has been for years now.

[blindsquirrel]

If you can't hear it and it's breaking things it's probably better to call it what it is, detonation. Light-throttle spark knock/ping doesn't break things.

[Dutch]

I have heard audible spark knock on L92 and L96 engines quite a lot due to carbon build up in the combustion chamber(from poor quality fuel) and 87 octance being used instead of 92 octance. Filling with 92 octance and doing a top engine clean usually restores them to their former glory and no knock is noted. I have seen anywhere from 2-10 degrees being pulled due to those two factors.

You can use a richer PE ratio to quell knock to a degree. Sometimes you can add fuel and see if adding back in the timing will increase power. Sometimes you make the same, sometimes more but its all something you should play with on the dyno to see what the engine likes.

If your AFR has an 11 in front of it for a conservative NA build, you are too rich. Shoot for mid 12's at the start and adjust from there.

Glad to hear your truck is fine!

Thanks for the reply guys!

Just to clarify - this is a towing rig that regularly pulls 8k+ lb trailers. The truck's curb weight is 6400lbs.
With the priority being component protection, wouldn't a 12.0:1 AFR target be a little lean for this kind of load? Especially for regular pump gas?

You'll love this extreme example: a stock Dodge 5.7 charger PPV targets the low 9s
allthefuel.PNG
(0.637λ with a 14.534 stoich = 9.258:1)

Dodge generally targets the mid to low 11s in their 5.7 trucks during PE.
So why would they choose to run the PPV so rich?
If the answer is component protection, why wouldn't the low 10s do?
("Weak pistons" and "dodge is dumb" aren't good answers )

Their gen3s use modern hypereutectic alloy pistons like every other manufacturer, but the 5.7s lacks piston oil cooling jets - could that perhaps be a factor?

[cadillactech]

Thanks for the reply guys!

Just to clarify - this is a towing rig that regularly pulls 8k+ lb trailers. The truck's curb weight is 6400lbs.
With the priority being component protection, wouldn't a 12.0:1 AFR target be a little lean for this kind of load? Especially for regular pump gas?

You'll love this extreme example: a stock Dodge 5.7 charger PPV targets the low 9s
allthefuel.PNG
(0.637λ with a 14.534 stoich = 9.258:1)

Dodge generally targets the mid to low 11s in their 5.7 trucks during PE.
So why would they choose to run the PPV so rich?
If the answer is component protection, why wouldn't the low 10s do?
("Weak pistons" and "dodge is dumb" aren't good answers )

Their gen3s use modern hypereutectic alloy pistons like every other manufacturer, but the 5.7s lacks piston oil cooling jets - could that perhaps be a factor?

For your application, why don't you put 92 octane in it and see if it stops with the pinging. I would suggest some form of top engine clean as well to clear out any carbon.

Can you upload your calibration?

Some OE LS calibrations with EQ ratios as high as 1.35, I am sure there are ones that are richer. Why? Competent protection. Catalyst protection.

Most OE's are terribly rich for the PE ratios. But they spend the time on a dyno getting everything right for their requirement list which is quite long to keep everyone happy. Think about meeting emissions, offering a power train warranty, an emissions warranty, being fuel efficient(why do you think they have such a high PE delay?) and meeting torque and horsepower targets. All within budget.

The richer you run the more un-burnt fuel makes it it into the oil and also washes the bore down so that has to be taken into account. You may not gain any advantage going from 11.5 to 10.5 AFR in terms of knock limit.

As far as the Dodge 5.7 Hemi goes, they do have a completely different combustion chamber design so that could have a factor in what AFR they select but I am not sure. I know it affects their ignition advance selection. That AFR could be with catalyst protection on so your rich AFR may not be just the PE ratio but PE plus catalyst protection.
The PPV vehicle may be richer as is it a police vehicle and more fuel will help cool an engine under extreme load during police chases?

Piston cooling is for sure a factor, one benefit from a richer air fuel ratio is cooling the piston crown. Oil squirters on the other side of the crown cools as well so this may be part of their calibration choice.

[Dutch]

For your application, why don't you put 92 octane in it and see if it stops with the pinging. I would suggest some form of top engine clean as well to clear out any carbon.

Can you upload your calibration?

Oh the truck is fine - the pinging you heard in the video was caused by a mistake on my part combined with a little bit of carbon buildup, both of which I corrected. I didn't want to post a tune because, well its not very interesting for starters, but I didn't want the focus to be about my truck.
Instead I wanted to <badpun> spark </badpun> discussion about conventional* knock in these engines: mainly what you consider acceptable and at what loads, and what you were able to get out of this engine on 87 AKI pump gas.
If you had any horror stories about terrible tunes, cracked ringlands, spark gaps mysteriously closing, I'd like to hear those too
If anything I figured it would be a nice break from the "help i changed something stupid and now my car wont start" posts I see whenever I check back in here

Another example: I wanted the truck to chug its 6400lb ass up a hill at 1500rpm so it doesn't constantly bounce between 5th and 6th, and I don't have to hold it in 5th all the time.
"chugging" defined as 0.65g airmass, 96kPa MAP, and absolute load of 72%.
The truck tries to achieve 13? advance under this condition.
But with the stock knock thresholds, it will jump to max retard (6? KR), leaving me with 7? advance and no power.

I see plenty of posts around talking about how this is typical while on 87 AKI, and other posts expressing frustration of how sensitive the knock thresholds are.

Now with the E78 ECMs, GM raised the thresholds quite a bit.
5FDP (awesome band by the way) pointed out that the interpretations of the tables could be different between the two ECMs, but after logging the raw knock sensor voltages they seem to have the same scaling.
It could be that GM acknowledged that the thresholds are too low and pulling unnecessary timing, and thus in order to make competitive power, they found it necessary to raise it.

As I said in Post 4, because GM halved the amount of timing KR is allowed to pull below 3000rpm compared to something like the g8 or caprice ppv, it might actually be safer to raise the thresholds to allow the ECM to better respond to actual knock events:

Assuming knock intensity corresponds to the amount of timing that needs to be pulled to suppress the event, if a 12 degree knock event occurs and only 6 degrees of timing is pulled, the intensity would be the equivalent of a 6 degree knock event that the ECM took no action against. This would surely produce an audible knocking sound.

In a way, this means that the E78 knock threshold tables are actually safer than the E38 tables: by raising the floor, the ECM is able to respond to actual knock events, rather than overreacting to noise and then hitting the max retard ceiling when knock occurs.

So I set the thresholds to match that of an E78 ECM, while also taking into account all the other different spark and knock retard parameters, anything else that could of changed with the higher thresholds, and now its able to achieve 13? advance up this hill, with a couple degrees of reported KR here or there but no audible knock. Awesome.

Obviously theres much more extensive validation that needs to be done here, and a good place to start after seeing what GM considers acceptable is to poll the enthusiast crowd.

*I'd like to keep the discussion around conventional knock suppression, and not waste time debating definitions or theory. I have access to a rather large database of SAE technical papers on the subject, so I'm covered there

[Dutch]

I rented a new Jeep Wrangler 2.0 turbo recently, and heard some pretty good spark knock under WOT:
https://drive.google.com/file/d/19n_...DDcTODLxbHEHwM
https://drive.google.com/file/d/14xn...krgeDRW2ZYm2nh

Keep in mind that my camera microphone thinks the knocking sound is background noise, so its actually considerably louder in person than it is on the video.

I showed this to a mechanic at a local Chrysler dealership, and they told me this was normal as long as it wasn't happening all the time.

Thoughts?

[cadillactech]

I rented a new Jeep Wrangler 2.0 turbo recently, and heard some pretty good spark knock under WOT:
https://drive.google.com/file/d/19n_...DDcTODLxbHEHwM
https://drive.google.com/file/d/14xn...krgeDRW2ZYm2nh

Keep in mind that my camera microphone thinks the knocking sound is background noise, so its actually considerably louder in person than it is on the video.

I showed this to a mechanic at a local Chrysler dealership, and they told me this was normal as long as it wasn't happening all the time.

Thoughts?

I wouldn't consider that normal, especially because the motor is a small turbocharged engine and is prone to LSPI which will break pistons.

What fuel did you have in it?

[Dutch]

LSPI is a special catastrophic mode of knock that occurs at high load low RPMs.
The knock you hear in the video is conventional knock.

I imagine the car was filled with whatever the owners manual recommended, which for this engine is 87 minimum, 91 preferred:

This engine is designed to meet all emission requirements, and provide satisfactory fuel economy and performance, when using high-quality
unleaded regular gasoline having an octane rating of 87, as specified by the (R+M)/2 method. The use of 91 or higher octane premium gasoline will allow these engines to operate to optimal performance. This increase in performance is most noticeable in hot weather or other heavier load conditions, such as while towing.
While operating on gasoline with the required octane number, hearing a light knocking sound from the engine is not a cause for concern. However, if the engine is heard making a heavy knocking sound, see your dealer immediately.

Screen Shot 2021-11-02 at 4.18.08 PM.png

The language of the owners manual implies light knock is normal while on 87, even though I wouldn't consider what I heard in the video light knock.

Yes, this is a FCA 2.0L in a discussion relating to knock in a L96, but for the most part, modern pistons in modern cars use more or less the same metallurgy. Both these engines use hypereutectic cast aluminum alloy pistons.

Anyway my point is if this is considered acceptable then I'm being way too cautious in my approach haha.

[Dutch]

Update to summarize

Audible knock was ultimately caused by carbon buildup. Gradually went away as carbon burned up.
I can say with certainty that GM did indeed crank the knock thresholds on the E78 ECMs, probably to help keep the engine competitive, and therefore these higher thresholds can be safely applied to the E38 provided common sense is used (spark tables are adjusted accordingly, engine, fuel system etc is reasonably healthy, stock configuration, etc).

Below is the final tune file. Feel free to incorporate into your future tunes:
Be advised the shift scheduling is optimized for high ethanol blends
thebeast-tire285-flex3-transflex4-FINAL.hpt

Target Final Spark advance with E85 (Includes VCM advance. Does NOT include IAT/ECT/PE adders):
Screenshot 2022-06-15 220135.png

Log of WOT run with heavy throttle at low RPM:
burb85wot.hpl

WOT 0-80 run:
https://youtu.be/fmfhtQ9wP5A

I'd be happy to answer any questions or concerns