Bolts on C7 GS : big knock retard above 6200 rpm (10+ !!!!)

[Sachs]

Guys,

C7 GS with :

AFE CAI/ARH LT Headers/MSD Intake Manifold/Katech ported Throttle body

I ajusted MAF/Torque management/PE/COT. Long terms trims are really good, even short term. Car runs good. Stock timing.

LM2 Wideband sees 0.83-0.85 Lambda at WOT (don't look at WO2 table can't get my LM2 work with HP tuners yet)

But, and big but, I get a HUGE knock retard starting at 6250 rpm....Like 10-12deg....It climbs from 0 to 12deg in 100 rpm !!! Obviously it triggers knock learn and the car switches to low octane table really quick....

I tried to add more fuel in the PE enrichment table (from 1.18 EQ ratio to 1.21 from 5800 to 6600) and adjusted the MAF table at the end (8100 to 9000 hz range I add 5g/s, but I already adjusted it before by 3% to run a little bit richer)

I don't understand what is going on... Do you see any clues in my log file or tune ? Could this be parasitic knock ? I know my X pipe is close to touch the metal brace under it, if it touches would you see that ?

PS : I run 91 octane at 2700 ft elevation (and drive pretty often at 8000-8500 ft)

Thanks

test18THROTTLE BODY OL.hpl

Last Tune PE MAF TQ.hpt

[Ben Charles]

May be fuel knock... that?s too rich on a NA setup...

Didn?t look at tune, but I see fuel knock on these engines if fuel is off

[Sachs]

May be fuel knock... that?s too rich on a NA setup...

Didn?t look at tune, but I see fuel knock on these engines if fuel is off

Too much fuel can cause knock ?? I thought more about the contrary but I'm new to tuning so perhaps I'm wrong..normally we try to add more fuel/remove timing to remove knock for what I knew... ??

[TriPinTaZ]

I agree with too much fuel. Your MAF curve is adding a lot of fuel. the AFE needs about 10% more over stock to be in the ballpark when using the stock PE table. In addition to that(you're over 10%) you have left Cat Overtemp enabled but the settings look wrong. Maximum and Minimum enrichment should be set to 1.00 so that it doesn't mess with fueling. Basically you're probably dumping even more fuel on top of already being rich via the MAF in the 8000+ Hz area.

In fact you can see the car is commanding .85 or richer (11.8-12.0 AFR) the whole time. Not only is that too rich, but its getting more fuel than that because your MAF values are too high in that area. Add COT enrichment of 1.4 on top of that and you're dipping into the way too rich area even further.

You should be logging PID's such as COT Enable, Injector Pulsewidth, all of the Spark and timing multipliers and a host of other things. You Datalog is too bare, I Was going to look at the Injector Pulse Widths but you're not logging. They really shouldn't be any higher than 3.9-4.0 during WOT on a stock car. I know you're modified but they shouldn't be much more than 4.2ms probably.

[Sachs]

Thanks for your insight.

But I'm afraid I had the same knock up high with stock MAF and PE....

My fueling is very rich because I tried to enrich around 8000hz to solve this issue....I added only initially 3% over stock MAF curve before really seing this KR, then on top 5g/s all above 8000hz as a second attempt and PE enrich at 1.21 instead of 1.18 above 6144 rpm.

I only changed the COT enable temps (bumped 100C)

You're right I need to log injector pulse to see what are the values during the pulls.

I don't think I have an air leak in my manifold as it would not be so consistent at 6250 rpm and i would see AFR changes on my wideband...

As far as WOT lambda 0.85 is fine for NA or do you want to be higher ?

I know my Xpipe Cats were touching my tunnel plate earlier, added one washer on each bolt to clear more space, almost never rubs again, but wondering if it does a high rpms as sometimes in high G corners I still hear it rubs.... Will add one more washer to see if it is not false knock.

[TriPinTaZ]

Are you sure your Wideband is reporting the proper lambada? Was it recently free air calibrated? .85 lambada x 14.1 stoich = 11.98 Air Fuel. You want to be between .89-.90 lambada.

What I was saying about the MAF is, specifically for the AFE intake, is if you multiple the entire MAF curve by 1.10 (10% increase over stock) you will be in the ballpark of where the MAF is accurate again for that mod. Meaning if the car asks for .85 lambada you will actually get pretty close to .85 lambada. You are well beyond 10% in the area you are seeing knock which is contributing to your .85 lambada reading. If your MAF is accurate at WOT then your PE table will be accurate. If you have not changed the min and max COT enrichment values to 1.00 then you could also be adding even more fuel here at times. But you won't be able to tell unless you add COT enable/disable to the PID list and data log it. Yes injector pulse widths will help tell the fueling story.


Have you messed with any of the Knock Sensor settings in the tune?

Once I get home I can compare some of your changes to my tune and share a datalog config file so that you log the appropriate PIDs.

[Sachs]

Are you sure your Wideband is reporting the proper lambada? Was it recently free air calibrated? .85 lambada x 14.1 stoich = 11.98 Air Fuel. You want to be between .89-.90 lambada.

What I was saying about the MAF is, specifically for the AFE intake, is if you multiple the entire MAF curve by 1.10 (10% increase over stock) you will be in the ballpark of where the MAF is accurate again for that mod. Meaning if the car asks for .85 lambada you will actually get pretty close to .85 lambada. You are well beyond 10% in the area you are seeing knock which is contributing to your .85 lambada reading. If your MAF is accurate at WOT then your PE table will be accurate. If you have not changed the min and max COT enrichment values to 1.00 then you could also be adding even more fuel here at times. But you won't be able to tell unless you add COT enable/disable to the PID list and data log it. Yes injector pulse widths will help tell the fueling story.


Have you messed with any of the Knock Sensor settings in the tune?

Once I get home I can compare some of your changes to my tune and share a datalog config file so that you log the appropriate PIDs.

I can assure you I only bumped the MAF table by 3%. Stock MAF Table was getting me 0.87-0.88 Lambda. Thought it was safer to run 0.85 for now.
I calibrated my Wideband 4 weeks ago in fresh air, though I can do it again to be sure.
I'm having a strange behavior with the MSD intake manifold like I said in a previous post. When I had only the AFE /LT Headers the car ran significantly leaner. When I added the MSD intake manifold the engine ran same as stock or close as AFR... I checked it everything is fine with it mechanically...Took it apart 2 times.

I didn't touch to the knock sensors in the tune yet. Was planning on bumping knock learn sensor down table as it is slow to 'de-learn"

[TriPinTaZ]

Actually you're not too far off on the MAF. I was reading the differentials backwards. I wonder if you don't have enough timing for your mods. And maybe you need some knock sensor help. I'm attaching my tune from my C7 Z51 M7. I have the same mods as you except ARH Mids and ported stock intake manifold vs your Longtubes and MSD. My car runs real hard, no knock. My tune is pretty much finished. Maybe you can look at some diffs and zero in on what might be causing you issues. Check the Knock sensor tables and maybe the timing as well. I've made a lot of adjustments in the WOT timing and VCP timing as well as the multipliers. You won't want to just copy one of my timing tables over without the rest, or at least thinking through the logic of them all as a whole.

Taz.015-2.poppop.ddtest.hpt

[Sachs]

I'll take a look thanks

I solved my knocking at high rpm mostly it seems. My Xpipe was rubbing on the aluminium plate under it.
I also removed COT protection and removed what I did on the PE table and the +5g/s I added on the MAF table

I still have small knock at some points but much less (2.5 max) as you can see and not always at 6250 rpm like before. Injector pulse width is 3.6-3.7ms

test23THROTTLE BODY OL.hpl

[Sachs]

Which version of VCM are you using ? I have VCM Editor 3.6.26 and get an error trying to open your file....

[TriPinTaZ]

Which version of VCM are you using ? I have VCM Editor 3.6.26 and get an error trying to open your file....

I'm using the 3.7 Beta. You have to email support and request access to download it from your Customer Center. Any files you edit or create in 3.7 are not backwards compatible with 3.6.

[Sachs]

So I looked at your tune.

To see I copied your MAF table and your timing /PE enrichment. Didn't take your DD table.

I looked at your knock sensor threshold, you bumped everything to 5000 instead of 3200, no wonder you see no knock as you made them less sensitive. But I read in this forum the knock sensors were pretty prone to be too sensitive or seing false knock ???

As expected seeing your MAF values, I run crazy rich with your table (0.83-0.84 Lambda). Which lambda are you running with your tune on your car ? What is the target lambda a DI engine like the LT1 should run a WOT ? On Corvette forum people say 0.85 is good and safe & makes the most power, some say it has to run leaner like 0.88-0.89...

Otherwise being rich it runs pretty good... I will lower my MAF table by 3-5% tomorrow to see if that makes me pick knock or not..

Almost no knock retard at WOT (like 1.5deg max), seems the more I push the less knock the car had... Something to investigate though, I get sometimes knock retard when I release the throttle.... Just once and during the first datalogging I saw like 10 deg of knock when I lift the throttle... What should I look for to get rid of that all the time ? It didn't do it after, like if the PCM was learning...

When you write calibration does it reset all PCM parameters ? This would explain why first pulls are less good than following one as far as transient ?

Here are my last 2 logs

test27THROTTLE BODY OL.hpltest28THROTTLE BODY OL.hpl

[TriPinTaZ]

As for the knock sensor tables, once you deviate from stock parts, especially exhaust, headers, reduce the amount of Cats etc., the stock settings are way too sensitive. I would suggest copying my Knock Sensor and Burst Knock settings so that you are not chasing your tail on Knock.

As for the MAF values, every car is a little different. I'm running a ported factory manifold and ARH Midlengths vs your MSD and longtubes so you may not get the same AFR as me if you're in a different part of the MAF curve than I am at WOT. My target is about .85-.86 which I know is a bit on the rich side but I didn't really see much gain by leaning it out. Once I get my permanent wideband in the car I will street tune it to be a bit leaner and will aim for .88-.89 lambada. The fact that you are still seeing knock on my MAF and PE table means its not a fueling issue. I really think you first step should be to change your Knock Retard tables. Don't worry, true spark knock still works but it is less likely to transition to the low octane table and it will ramp in and out of knock retard quickly. I have faith in my tune so I am not worried. Pretty much any pro tuner does the same thing to the Knock Tables.

Again, I did see the 1-1.5 degrees of retard in your logs. You AFR is safe and your spark is actually lower than it could be for more power. I'm at a full 22-23 degrees all the way to redline. But again if you copied my spark table you have to make sure you mirror the multipliers and adjustments as well as the variable cam values and VCP spark. I also sometimes see a little KR when i lift off the throttle but only like 1-2 degrees and not very often. Actually, it's not very often at all that I see it. Again this could be your factory spark retard tables.

When you write the calibration it should reset your Long Term fuel trims. It does reset the throttle position learning and what not I'm sure too. I only manually reset my LT fuel trims after a write calibration when I change my MAF tables, but I only do it to be 100% sure they are reset. I Can see the LTFT's at zero every time I start the car right after a write calibration so again, I don't think it's needed but I'm from the old school days of LT1 and LS1 edit where we had to reset LTFT's after every flash.

The logs don't look bad but the knock and spark isnt optimal for full power. You are leaving a little bit on the table there. Strangely I noticed in your logs that your injector pulse width is only reaching 3.6ms. I typically hit 3.8-4.0 during WOT. So now I wonder if I am running way richer even when you used my MAF curve, or if I am making more power due to my Spark Curve. I need to get my Wideband in the car but I won't be able to do so until next month.


By the way are you catless? If not, where did you mount your wideband sensor?

[Sachs]

Guys,

Car runs strong but I still get this damn knock retard above 6250 rpm....it doesn't make sense because sometimes I have a lot, sometimes nothing...

Take a look at 2 of todays logs...

I will start to monitor cylinder misfires also to see if I see something

I tried to pull 7 deg of timing above 6200 rpm, and it seems to remove the issue...at least on the 2 pulls I had the time to make...

It seems cylinder 7 has more knock retard than the others...

But why suddently at 6250 rpm...

test34THROTTLE BODY OL.hpl
test32THROTTLE BODY OL.hpl

[Higgs Boson]

put race gas in it or disable the knock sensors as a test to see if it is real or false knock. retard without knock sensors tells you it could be your torque tables pulling spark out. removing the lead manually could just make delivered and predicted torque match up when all you need to do is raise predicted to match actual.

[TriPinTaZ]

Return your peak torque table to stock. Use my knock sensor settings. Your timing looks fine, and if your wideband is right then fueling should be good, right? I'm at 3.9-4.0 ms injector pulse width by 6200 RPM. It's weird to see you are 3.4-3.6ms so I wonder if your wideband is calibrated properly?

My knock settings do not eliminate knock detection, only reduce the sensitivity of the overly sensitive factory settings with aftermarket headers/exhaust. If there is true knock, you ECU will still pick it up and retard timing. If you are not willing to do so, get your injector pulsewidth closer to 3.8-3.9 up at 6000+ and see if that helps.

[Sachs]

I tested my wideband calibration again last week. (fresh air must show 20.9% O2) And it shoes 20.9 % O2 when in engine breaking which tells me it is good.

My Wideband is on my Xpipe just after both headers merge (ARH X pipe). My wideband shows 0,85-0,86 Lambda all the way.

On this logs my torque table are back to stock and I used your Knock sensors configuration (5,00 threshold and burst knock settings). It didn't change anything at all with my issue, which makes me think it is not real knock??

Higgs Boson does virtual torque table pulling timing is seen as knock retard ?

I'm a little reluctant to disable my knock sensors if I have knock for real .... I heard one of my colleagues saying there is a pump with 100 octane in the area, I'll probaby try.

[Higgs Boson]

in my experience, yes, the knock retard pid will show retard not from actual knock. I am not sure what you call it, whether "torque management advance/retard" or "pcm spark lead adjustment."

there is definitely pcm spark control that we don't seem to have access to directly in one table or disable drop down menu (probably a lot of things like that with the E92(A)). i just try to get fueling right, spark right, and fiddle with the virtual torque tables, peak torque, and driver demand until it doesn't try to take action over my goals. sometimes you have to go back and fine tune after getting everything dialed in.

gm genV pcms are very annoying, but at least we have access to these. I don't think it will be much longer and I will only own old cars. :-(

[Sachs]

Hey by the way about the injector pulse width... I live at 3000ft, so my baro is lower. I'm assuming you live at sea level, so you have more air pressure, so need more fuel...around 7-10% more so your 3.8-4.0ms pulse makes sense

[Sachs]

So guys I had only one hour today to test but here are the results...

First test was a try with virtual torque table increased by 3% above 6200 rpm... no change on the knock retard, as you can see at the end of the record

test38THROTTLE BODY OL.hpl

Then I tried something, putting my PE fuel at 1.23 instead of 1.14 above 6144 rpm (its a shame we can't change the scale on this one)

No knock anymore, but runs crazy rich above 6144 rpm (0.77-0.78 lambda, and 0.85-0.86 before 6144 rpm)... See the log ...

test39THROTTLE BODY OL.hpl

So what does this tells us... It seems I need to add more fuel to get rid of this knock...I need to see how much, went pretty high with this trial, will try less.

So now what would be the best way to deal with that ? Adjust the MAF at high rpm or just play with the PE table like I did ?

My second thing is, is it normal to have this behavior ? I know the MSD flows a lot better at high rpm but it is pretty strange to pick knock like that sometimes and sometimes not at all...