Stupid question about KR/knock

[jtrosky]

Does the cars comptuer "store" learned KR values at all or are they a "per trip" setting? For example, if the car detects knock at a certain RPM/airmass, will it "store" information about that KR so that it doesn't knock again in teh future under the same situation? Or does it only "learn" KR values during the current drive cycle?

If it does "store" KR informtation, does that information get reset when you apply a new calibration?

I've been noticing quite a bit of KR on my 3.6L LFX engine (especially when really hot out) and I'm wondering if I'm making matters worse by writing new calibrations all of the time (if it resets any stored KR data) - even when not changing any KR-related parameters.

Thank you!

Also - now that I'm really paying attention to KR, it "worries" me when I see it in teh scanner logs - when I see KR does that mean that it's harmful to my engine at all?

[SleepinRegal]

Does the cars comptuer "store" learned KR values at all or are they a "per trip" setting? For example, if the car detects knock at a certain RPM/airmass, will it "store" information about that KR so that it doesn't knock again in teh future under the same situation? Or does it only "learn" KR values during the current drive cycle?

The 3.6 LFX engine might be different than the 3.8, but for me KR (Knock Retard) is not stored just read and corrected as far as I know.

I've been noticing quite a bit of KR on my 3.6L LFX engine (especially when really hot out) and I'm wondering if I'm making matters worse by writing new calibrations all of the time (if it resets any stored KR data) - even when not changing any KR-related parameters.

Heat is a significant contributing factor with KR. And no you're not making matters unless you're increasing your timing before calibrating your fuel trims.

Also - now that I'm really paying attention to KR, it "worries" me when I see it in teh scanner logs - when I see KR does that mean that it's harmful to my engine at all?

Everyone is going to have their different view on what's considered healthy and what's not. I have my aeroforce interceptor to light up if it sees 4* of KR which tells me to let off the gas.

Hope that helps

[jtrosky]

Thank you very much - I appreciate the response. I just wasn't sure if the ECM recorded KR values "permanently" when knock was experienced (when it has to retard timing due to knock) or if just just corrected it as it saw it. Sounds like it just corrects it as it sees it and doesn't actually "learn" from it long-term. At least I know that I'm not "losing" any learned KR values when I write new calibrations now.

My understanding is the KR is just part of the way that the LFX is programmed from the factory and it's considered "normal" on this engine. It's not that I'm seeing 10+ degrees of KR or anything like that, but I do see 4-6 degrees of KR pretty damn often on this engine - especially when the IAT's are up over 95F. Like you mentioned, while monitoring it "live", I can stop it by either reducing or increasing the throttle. I guess I should probably start working on my timing tables where it's really bad (certain airmass/RPM combinations seem to consistently cause KR, so I could probably reduce it substantially by targeting those cells in the timing tables. I'm rather new to this, so I'm learning! :-)

Thanks again for taking the time to respond - I truly do appreciate it.

[Road]

Does the cars comptuer "store" learned KR values at all or are they a "per trip" setting? For example, if the car detects knock at a certain RPM/airmass, will it "store" information about that KR so that it doesn't knock again in teh future under the same situation? Or does it only "learn" KR values during the current drive cycle?

If it does "store" KR informtation, does that information get reset when you apply a new calibration?

I've been noticing quite a bit of KR on my 3.6L LFX engine (especially when really hot out) and I'm wondering if I'm making matters worse by writing new calibrations all of the time (if it resets any stored KR data) - even when not changing any KR-related parameters.

Thank you!

Also - now that I'm really paying attention to KR, it "worries" me when I see it in teh scanner logs - when I see KR does that mean that it's harmful to my engine at all?

Unless you changed your tune your low octane table is the same as your high octane table so if your seeing knock then yes it is not good because you have no cushion in your timing tables. Also your IAT table is zero'd out. Those 3.6 come stock with aggressive timing table's and normally knock right over to the low octane table.
Mike

[jtrosky]

Well, both my stock tune and the performance tune I've used in the past both had different high/low octane tables - I've never run a tune with those tables being the same. However, the aftermarket tune was a 91 octane tune (I always run 93 octane, TopTier gas), so the high and low octane tables aren't as far apart as they are with the stock tune. That being said, I actually had more KR with the stock tune than I do with my aftermarket performance tune (but tey both have significant KR).

As for the IAT table, it was zerod with the factory and the aftermarket tune - I've never modified that table at all. I do have the 87 octane version of the performance tune as well - maybe I'll apply those timing tables instead, just in an effort to reduce the amount of KR... Honestly, outright performance is not even all that important to me - it's not like I'm racing the car or anything - just a daily driver.

Thanks.

[Road]

Well, both my stock tune and the performance tune I've used in the past both had different high/low octane tables - I've never run a tune with those tables being the same. However, the aftermarket tune was a 91 octane tune (I always run 93 octane, TopTier gas), so the high and low octane tables aren't as far apart as they are with the stock tune. That being said, I actually had more KR with the stock tune than I do with my aftermarket performance tune (but tey both have significant KR).

As for the IAT table, it was zerod with the factory and the aftermarket tune - I've never modified that table at all. I do have the 87 octane version of the performance tune as well - maybe I'll apply those timing tables instead, just in an effort to reduce the amount of KR... Honestly, outright performance is not even all that important to me - it's not like I'm racing the car or anything - just a daily driver.

Thanks.

Flashing the ecm does reset your knock learn values. I personally don't like knock retard it keeps timing retarded too much. My goal is to keep spark as close to high octane table as possible. If it were me I would use the stock high octane timing table on your setup and log and adjust it to where your not seeing KR then copy it to the low octane table and subtract @4. Also a heat soaked motor will yield a lot of KR and throw it right over to the low octane table I would use the IAT base table and remove 1 degree of timing starting @86 temp and progressively increase it 1 degree every column above 86 temp and make sure all the multiplier tables are set to 1.00.
Mike

[jtrosky]

Hmm, so the ECM in this case (E39) *does* in fact store information about knock and learned KR levels to avoid knock under the same conditions in the future? That's pretty important to know as a lot of the time, I will flash a new calibration daily while working on something.... What exactly does the computer store - just optimal advance values for each "cell"??

In terms of how to deal with the KR seen in the logs, I've been trying to figure out the best way to address it. There are so many spark "correction" tables that alter the advance values (over 5 I think), so it's a lot more complex than it used to be where you just worried about the main high/low spark tables.

If I posted some pics of my KR histograms and some of my KR-related parameters, would someone be willing to advise me as to the best way to reduce KR? I just cringe eveytime I look over at the scanner app while driving and see it showing KR values between 4 and 6 degrees for a few seconds at a time!

Thanks guys - I really appreciate the assistance!

EDIT: For the record, I never actually *hear* any engine knocking - just see the KR values in the Scanner app. I seem to get it consistently as certain RPMs (1600 and 2200 seem to stand out in my memory).

[SleepinRegal]

Hmm, so the ECM in this case (E39) *does* in fact store information about knock and learned KR levels to avoid knock under the same conditions in the future? That's pretty important to know as a lot of the time, I will flash a new calibration daily while working on something.... What exactly does the computer store - just optimal advance values for each "cell"??

I did not know that. I know for the 3800's that the knock sensor will "attack" the knock retard before you even get the chance to even see it (Which is pretty smart coming from a considerably "dumb" PCM)

If I posted some pics of my KR histograms and some of my KR-related parameters, would someone be willing to advise me as to the best way to reduce KR? I just cringe eveytime I look over at the scanner app while driving and see it showing KR values between 4 and 6 degrees for a few seconds at a time!

I'm no expert but I could take look. The attached image is the histogram that I use if I want to monitor KR. As you can see I have some KR but really 1* of KR is nothing. Heck 4-6* of KR isn't going to pop a piston. I use The Table Modifier Thingy to automatically adjust my KR values. It's a great addition to HP Tuners. Check it out!

EDIT: For the record, I never actually *hear* any engine knocking - just see the KR values in the Scanner app. I seem to get it consistently as certain RPMs (1600 and 2200 seem to stand out in my memory).

You won't hear KR. If your seeing KR at that low of an RPM then it could also possibly by false KR as sometimes knock sensors can be set a little too sensitive. Don't stress it man.

[jtrosky]

An example of the level of KR that I'm referring to - and the IAT for this histogram was only 86F!

And this was with the stock timing tables.

Two pics - one with the Average KR and one with the Max KR for the same drive.

KR_Average.JPG

KR_Max.JPG

See why it bothers me? :-)

[SleepinRegal]

First don't look at the max KR. Concentrate on the average. Looks normal to me man. Also I don't know where you're from but 86*F is pretty darn warm where I come from.

[jtrosky]

Ok, cool - if those levels are not a cause for concern, then I won't worry about it. Just seemed excessive to me. I'm in Pennsylvania and lately the IATs have been up over 100F (outside temps 95F'ish).

When the outside temps are in the low 80F's or high 70F's, the amount of KR is a lot lower (less frequent).

[Dr. Nopps]

I guess you should call that a stupid question seeing as I in fact tried to help you with this several months ago, most likely before you had fully realized it was even happening, but you decided to ignore my hard work and advice and listen to a bunch of idiots that wanted to appear knowledgable. To their credits SOME members actually gave you some sound advice, they know who they are. BUT... IGNITION TIMING IS NOT THE ONLY CONTROLABLE FACTOR CONTRIBUTING TOWARDS DETONATION. And if you think for one second this sounds harsh, imagine how harsh it feels when people just brush aside all of your hard work and advice when you're pretty much handing out for free the best/most usable advice they have been given so far. So ignore my advice again if you want to, or put yourself in a position where you can actually advance or simply smooth out your ignition timing safely for power, MPG's, or longevity.

The "Dr. Nopps Hates Knocks Guide"
Last Post By ZeroBoostBuick 10-13-2014 04:10 PM
2

[jtrosky]

You sure it was I that you are thinking of? I seriously don't remember ever speaking with you about this subject.... I honestly don't ever remember even working on KR before.... I'm rather new to this stuff, so it's entirely possible that what you originally said went right over my head as well.

I've tried searching for the discussion you are referring to, but couldn't find anything.

And for the record, I don't think that anyone is saying that ignition timing is the ONLY source of KR, but that it is typically the most common source of KR...

[JerryPH]

Disagree. The most common source of KR is bad fuel management. That goes as much for stock cars as it does for tuned cars.

Unless you are doing something ridiculously crazy with the ignition tables, I can pretty much get the KR under control with proper fueling. Cars with stock timing tables and getting KR are *not* getting KR because they timing tables are bad, but because the factory runs these cars so lean.

Now... looking at the 87 octane Overkill timing tables, they are mostly way lower down low and a very small amount higher than stock up top. That means that in normal every day driving, they should have less than 5 degrees of KR which is more than acceptable for a stock or lightly modded car.

In the past, I have said it over and over... to get a 0 KR car is not only quite difficult, but usually means timing tables that end up costing you more HP and MPG than if you set an acceptable level of KR. My chosen favorite number is 3 degrees or less. I scanned a bit yesterday when it was 34 degrees celcius (a touch over 94F) and I was seeing 3.5 degrees of KR at 3/4 throttle around 2200 RPM, everywhere else it was less than that. When cooler, it's usually never over 3 degrees of KR anywhere, and certainly at WOT, I am seeing 0 KR.

[jtrosky]

It's funny - you say it's becuase the car is too lean and Dr. Nopps says it's probably becuase the car is too rich! :-)

In my case, I see up to 9 degrees of KR at times, dependng on the IAT's. The hotter the IAT's, the more knock I see even without changing any fueling parameters. I'm not running the Overkill 87 octane tune, I was running the 91 octane Overkill tune, so maybe that's the difference (alone with other difference in temperature, humidity, elevation, etc). Lots of variables. But I can assure you that I was seeing much more that 3 degrees of knock with the 91 octane Overkill tune (running 93 octane fuel).

I think that the reason that I was seeing more KR in the histogram with the stock tune was becuase the Overkill tune returns the timing MUCH quick (10 times quicker) than the stock tune does, which made the stock tune appear to have more KR, when it was really just returning the timing a lot slower. I've since reverted to the stock timing table and removed some timing when IATs are really high like recommended above - we'll see how that works out.

I can also say without a doubt that the Overkill high octane table is generally much higher advance than stock (except for VERY low airmass cells). Some cells in the high octane table are up to 20 degrees higher than the stock tune.

[Road]

In closed loop the O2 sensors are not going to let it run lean or rich. Anytime you see KR the ECM is adjusting towards the low table. Make your average timing equal to the high table and you make best power.

[jtrosky]

So I had some great success today just by shaving a degree or two off of the high octane timing table where I consistently saw KR. So far, I'm very happy with the results. I also removed a few degrees of timing from the entire low octane table so that the ECM has some "wiggle" room to adjust if needed (due to bad gas, etc). Previously, there was only like 1 or 2 degrees of wiggle room. Now there is about 4.

I do have another question though - as you go across the timing table for a certain airmass (as RPMs go up), is it "ok" for timing to go down for a certain RPM range and then go back up? For example, is it ok for timing to be 25@3200, 26@3300, 24@3400, 26@3500, 25@3600, etc... Wasn't sure if that small "valley' in there was OK or if the table should follow a natural "curve"... I have a 200rpm range where I get KR that I'd like to remove a few degrees of timing, but it would "interrupt" the "natural" curve of the rest of the table...

Thank you all very much for your help - you've been extremely helpful! I feel much better now that I'm getting a lot less KR - especially in the 1200 RPM area at very light throttle (when crusing) - it would consistently give me KR at 1200-1300 RPM at very light throttle and it was really bothering me. Pretty much gone now.... I also set up my IAT tabales so that certain ranges get pulled a little at high IAT temps (between using the main IAT table and the IAT multiplier table).

Thanks again.

[jtrosky]

By the way, what is the purpose of the VCP Spark modifier table? Since it uses the same criteria (airmass and RPM) and the multiplier table is all 1.00, wouldn't it have the same effect as modifying the spark tables themselves directly? Trying to understand what purpose the VCP Spark modifier table servers if it just adds to the main space tables based on the same parameters (airmass and RPM)....

[507]

It doesn't just add timing , you are able to use it to subtract timing. Your timing can be perfect on a 75 degree day zero humidity. Take a 98 Degree day with 98 percent humidity, with super high time just to say. knock knock. So If the knock is because of hot weather used to spark modifier tool to lower timing. if knock occurs because your motor is running too hot for your timing you can use the engine coolant temp spark modifier to lower timing so you don't have to go on your a high octane and low octane map just to modify every time is present

[jtrosky]

I think that you misunderstood my question. I'm not talking about modifier tables in general - I completely understand how the IAT or ECT modifier tables can add/remove timing based on temps (and I'm actually using the IAT modifier tables successfully now), but what I'm asking about is the VCP Spark modifier table specifically. Supposedly, it's a "Variable Cam" spark modifier table - but the axis of the table are the same as the axis in the high octane table - so if the VCP multiplier table is all 1.0's, then wouldn't modifying the VCP Spark table be the same as just modifying the main spark tables directly (since it's based on the same parameters, aimass and RPM)?

I'm trying to understand why I would use VCP (Variable CAM) Spark table instead of just modifiyng the main spark tables directly - at least when the VCP multiplier table is all 1.0's...

Thank you!