LNF KR Histograms - need feedback

[wilson0x4d]

TL;DR? Unfortunately, I am not sure what Main Spark Advance "In Min", "Ex Min", "In Max" or "Ex Max" are, at a technical level. I've arbitrarily utilized "Intake Cam Des Angle" [PID.2170] and "Exhaust Cam Des Angle" [PID.2176] as int/ex vectors, and then I include samples based on less-than or greater-than midpoint (~25 on a scale of 0-50). I believe this to be inappropriate and I am seeking advice on more appropriate PIDs and/or value ranges.

After reviewing some videos and threads on KR calibration I decided to create new histograms which would, in theory, help me to perform a more accurate calibration relative to Min/Max Intake/Exhaust states.

Prescribed method I have seen thus far:

1. Go to Engine->Spark->Advance
2. Make a clipboard copy of "one of" the tables, such as "Intake Min, Exhaust Max"
3. Paste clipboard into all four tables.
4. Copy adjustments from "Total KR" table, such as "CSSOB's KR Histogram"
5. Apply adjustments to all four tables.

However, there are four different tables for a reason, and I suspect I may over/under compensate if I copy/paste the tables over one another, and or copy/paste a single KR vector over multiple tables.

Seeking a more pragmatic approach to adjusting these tables I've set out to create separate histograms to capture KR filtered by the 4 Min/Max states:

LNF-X4D-KR MnMn.hst
LNF-X4D-KR MnMx.hst
LNF-X4D-KR MxMn.hst
LNF-X4D-KR MxMx.hst

The filters for each are as follows:

KR MnMn: [PID.2170] < 25 AND [PID.2176] < 25
KR MnMx: [PID.2170] < 25 AND [PID.2176] > 25
KR MxMn: [PID.2170] > 25 AND [PID.2176] < 25
KR MxMx: [PID.2170] > 25 AND [PID.2176] > 25

I would *reallly* appreciate some feedback on values which may be more correct, or PIDs which are more appropriate for what I'm attempting to do.

Thanks!

[cobaltssoverbooster]

unless you are doing lab testing for a firm or masters program then you are trying to enable features that will rob you of power and consistency.
i currently do not have time to preach on the thought process of an lnf or similar torque demand ecu. I believe you want the safest tune possible by covering all the ranges possible but the lnf is not an ecu with high axis resolution, therefore you cannot tune for enough plots to make this function work properly and deliver max power output on a consistent basis.
hopefully one of the other members can come in and shed light on the situation more in depth.
thank you for asking questions about a high order ecu process. Not a lot of people asking questions about ecu process anymore.

[MikeM173]

Learn the system before you try to redesign what we do. Just a thought. I get you are trying to help but from the little bit I read it seems like you need to slow down.

If I am wrong guys please let me know.

[wilson0x4d]

I noticed a new engine sound at almost exactly 3000rpm, very high pitched (not the BOV, which sounds different and more obvious).. the car has never made the sound before. That's the only reason I began investigating KR, but now I believe the KR is a symptom of a real problem and simply changing spark timing tables may actually be inappropriate.

Before I changed any tables I wanted histograms that matched the tables, both for deriving change, and for quantifying the effects. I have KR I didn't have before and thought perhaps I would need to adjust spark timing to prevent detonation under load.

I'm not looking to redesign what tuners are doing, as "a new guy" I only want to make sure I'm cautious and can verify any changes I make. I am looking for better info since, I repeat, I've chosen "arbitrary vectors" of int des angle and ex des angle to derive InMax/ExMax values from. What I need is a definition/clarification/understanding of the terminology used for the four main spark tables in VCM Editor, for no other reason than to have legitimate/correct histograms.

That said, I found the Chart Display much easier to read for the same data over time. It looks like the only "offensive, and consistent" KR I'm seeing is when I'm cruising @ 2600rpm-2800rpm @ 40%-45% ETC, 100%-111% load. KR gradually grows from 0.2 to 1.5 and sustains for as long as I hold ETC at 40-45%. When I look at my histograms I only see this manifest in the Max/Max condition (again, using my "arbitrary" vectors, which I assume are incorrect until I can get clarification), and primarily along the 2500-2900 rpm range under 98%-110% Load. I can further adjust histogram filters to only tabulate data within these ranges.. VCM Scanner is pretty amazing in that regard.

I've attached chart screenshots to show the KR I'm seeing.

I originally thought I would want to modify the main spark tables, and I'm pretty sure if I *did* then I would (being the hobbyist I am) only edit the In.Max/Ex.Max table, and probably only 4 cells. Unfortunately I think I need to pull back on some of the edits I've made instead, and try to understand what else I can do to eliminate the KR (since I'm such a newb, I've no idea.)

Thanks for the help!

kr001.png

kr002.png

kr004.png

kr005.png

[cobaltssoverbooster]

I noticed a new engine sound at almost exactly 3000rpm, very high pitched (not the BOV, which sounds different and more obvious).. the car has never made the sound before. That's the only reason I began investigating KR, but now I believe the KR is a symptom of a real problem and simply changing spark timing tables may actually be inappropriate.

Before I changed any tables I wanted histograms that matched the tables, both for deriving change, and for quantifying the effects. I have KR I didn't have before and thought perhaps I would need to adjust spark timing to prevent detonation under load.

I'm not looking to redesign what tuners are doing, as "a new guy" I only want to make sure I'm cautious and can verify any changes I make. I am looking for better info since, I repeat, I've chosen "arbitrary vectors" of int des angle and ex des angle to derive InMax/ExMax values from. What I need is a definition/clarification/understanding of the terminology used for the four main spark tables in VCM Editor, for no other reason than to have legitimate/correct histograms.

That said, I found the Chart Display much easier to read for the same data over time. It looks like the only "offensive, and consistent" KR I'm seeing is when I'm cruising @ 2600rpm-2800rpm @ 40%-45% ETC, 100%-111% load. KR gradually grows from 0.2 to 1.5 and sustains for as long as I hold ETC at 40-45%. When I look at my histograms I only see this manifest in the Max/Max condition (again, using my "arbitrary" vectors, which I assume are incorrect until I can get clarification), and primarily along the 2500-2900 rpm range under 98%-110% Load. I can further adjust histogram filters to only tabulate data within these ranges.. VCM Scanner is pretty amazing in that regard.

I've attached chart screenshots to show the KR I'm seeing.

I originally thought I would want to modify the main spark tables, and I'm pretty sure if I *did* then I would (being the hobbyist I am) only edit the In.Max/Ex.Max table, and probably only 4 cells. Unfortunately I think I need to pull back on some of the edits I've made instead, and try to understand what else I can do to eliminate the KR (since I'm such a newb, I've no idea.)

Thanks for the help!

this will not happen if you split the tables up because you will allow the ecu to choose location making it very hard to determine why the ignition advance is running what it is running. if they are all the same it is forced to use that value, therefore if it doesn't run it then it is either an active multiple such as iat/afr/coolant or true kr removal from knock sensor activation. as of now the limits of the four ignition tables are set by the physical extremes and as the cams cross over the middle point between the physical extremes ( ∆ of intake˯max and intake˯min) /2 then you switch between the tables.
that's how it works on a very general basis...i do not recommend tuning with this method as it is hard to control and hereby forewarn you that if you blow up your combination that it will be your fault for tuning in an inconsistent mode of the ecu. no one here will be responsible for problems caused by your tuning so tread carefully in this mode if you choose to go there.

[MikeM173]

Can you tune all 4 tables separate? Yes. Is there any real benefit to it? No.

You need to read a lot more on tuning this engine before you start doing what you are doing. There is definitely plenty of info on spark timing out there that will help you figure all of this out. I don't mind you testing on your own but if you are going to do that you need to be tracking everything and figuring it out almost completely on your own. To ask us to explain something you could have read already just means you haven't done the initial research. Do that first and ask questions based on that then try the tables out and then try to manipulate them differently.

Not trying to be rude. I actually love when people try things on their own. You need to figure out what the understanding of the table is prior to changing stuff or you need to say screw it I'm going to change it and learn what it does and if something bad happens it is my own fault.

[ECaulk]

Assuming you have the boost PID setup correctly you need to stop giving it 50% throttle in 5th gear. You're above 15psi and the command lambda is 1.0 with an advance in the 20s. You're also not logging or at least showing STFT, fuel pressure, coolant temp, I'm sure there is more that I'm missing but I'm not on my computer I use to tune to check it against GMTech's config I use.

[TurboCobalt]

i'll try to help although here's my disclaimer I am also very new, so take my words with a grain of salt.

Main Spark Advance min max are based on intake and exhaust cam phasing.

cam phasing: the cams are actuated by the cam phasers in only 1 direction. They are "parked" during a warm idle, physically, when they are parked the exhaust cam is at maximum advance, and intake cam is at maximum retard, therefore while parked, they have the LEAST overlap. when you actuate the exhaust cam phaser, it retards it, adding overlap. when you actuate the intake cam phaser it ADVANCES (edit: had a typo here) it, also adding overlap. The pids you read indicate desired (and by the way i would pick the measured PID) cam phaser actuation in crank degrees ( i think it was crank, may be cam degrees... now i forgot, somewhat irrelevant to the subject)
In the scanner you observe how many degrees each cam phaser is actuated, that's why they're always positive. in the VCM editor they also have negative values because they are with respect to cam opening/closing and they are in crank degrees which could be negative. From what i gather the high load area of the cam tables is pretty optimized, there is some small drive-ability gain in the cruise region but for me personally i left them alone with just some minor smoothing. If this confuses you i can add detail later, but this is just background. so now we got this far: leave cam phaser tables alone (for now at least).

I will GUESS that the spark tables are based on cam phasing actuated, or duty cycle, so in my GUESS min intake, min exhaust would be no actuation, so parked (warm idle) they could be based off of crank degree retard/advance, i dont know, but you could easily figure it out by making a special tune with test values for cam phasing and spark to figure out which one it uses.

The previous poster was right though, that the computer picks an actual spank by interpolating a spark value, given cam positions.

Here is the flaw in your logic, assuming you don't mess with the cam phasers(my advice is "dont"), they will always do the same thing (assuming engine is warm) so there is no need to make multiple histograms, because given an rpm and load, your cams will always do the same thing; therefore you are always at the same point interpolated between the 4 spark tables.
why try to populate a 3 dimensional space, when it can be simplified to 2 dimensional space.
I'm not as quick to jump on the band wagon of "make all 4 spark tables the same"; they do affect something, cam phasing affects dynamic compression ratio, charge dilution, bla bla bla, so they are meant for SOMETHING, but i understand why simplifying all tables to the same values is a desirable solution, just for that: simplicity, and they are virtually the same from the stock tune as well. i'm sure i'll get crucified for that last sentence on this forum.

long story short, unless you constantly mess with the cam tables, there is no need to make multiple histograms because the cams always do the same thing at a given load and rpm, so there's no need for a 3d space, when it can be done in a 2d space.

lets say you INSIST on doing this, this is a VERY dynamic and fluid environment, cam phasing, don't just split it into 2 (what you did > or < 25), you would need much more points. splitting it into 2 is way too crude for such a fluid system.

your scanner config is very crude as other's have mentioned, learn the difference between "gauge pressure" and "absolute pressure" what most people call "boost" is gauge pressure so 0psi is atmosphere, negative psi is vacuum. you seem to be reading all as absolute pressure, and calling them "boost". Unlike most other engines, you really do need to record STFT for this controller.

hope this helps.

[TurboCobalt]

ECaulk, i dont think he's at 15 psi in those screenshots, i think he's a 4 psi above atmosphere, i think he's just logging absolute pressure. it becomes more evident when looking at his MAP value.

[cobaltssoverbooster]

lambda was .98 at +100% load at 3k rpm running 20* ignition.....of course it knocked. my 100% load column in the pe table would have had that event running .88 lambda. 100% load is a range that commonly starts to see boost especially at 3k rpm...although not much it usually requires more fuel to prevent detonation.
i think you have an issue with commanding a safe value.

Turbocobalt...i dont think you should be crucified. I also believe those tables are there for a reason but for the sake of tuning with reliability i do not believe they should be used as separate tables for the most of the reasons you listed.

Wilson if you dont mind posting a tune file we can thumb through it and normally find the cause of the ignition event just by looking at the file settings.

[ECaulk]

ECaulk, i dont think he's at 15 psi in those screenshots, i think he's a 4 psi above atmosphere, i think he's just logging absolute pressure. it becomes more evident when looking at his MAP value.

Thats why I said if he had it setup correctly, either way he's rolling into boost and commanding a rather high value. As cobaltssoverboosting pointed out it def requires more fuel than 1.00lamdba.

OP for a good link to a config that will have everything you should be looking at for an LNF check this thread out

http://www.hptuners.com/forum/showth...d-Config-Files

GMTech walks you through how to correctly load the config so you'll have the needed data at your fingertips for tuning the LNF

[wilson0x4d]

@TurboCobalt - All good info. A couple terms turned me on to more posts online I hadn't read yet so, thanks!

@CSSOB - I attached my tune (thu4.hpt) and the log file I collected from it (thu4.hpl). I feel like I made modest edits compared to what I've read from others, except maybe in the PE table, which I may flip between stock/tuned and retest once the temp outside goes down. I've been trying to work with low PE lambdas, parroting what I've read elsewhere

@ECaulk - I updated my graph to show more useful data. I've also attached a couple more screenshots, including one that shows sustained KR while cruising ~65mph .. the car sounds like it has a low growl/vibration when this is happening.. I'll definitely go over the GMTech guide this weekend for consistency sake.

kr006.pngkr007.png
kr009.png

@Mike - I read over ZZPs KR page, and a few non-LNF sites for good decriptions of variable cam timing, spark/ignition timing, etc, but was still left with questions I figured had common-knowledge answers. As it stands, I believe the Min/Max values interpolate between 0 and 50 as @TurboCobalt described. I've looked for a forum thread that discusses this, here and elsewhere. I wish someone could (and would) read the PCM disassembly to give a definitive answer of behavior (interpolation or switchover) and bounds (fixed 0-50? delta of min and max as configured in cam tables? etc) -- I understand where you're coming from, I think.. but a small part of me sees a thing and needs to understand it, even if nobody else finds it useful. If nobody helps me to that end I'm not going to be bent, and I'll be the first to edit my post and warn people what NOT to do if it turns out there's an incredibly stupid thing to be done here and I do it.

Thanks again

[MikeM173]

Let's clear one thing up.

Trying to figure something out on your own like you are with just some insight from others is a good thing. Let's change that. It is a great thing!

With that said though, there is some basic stuff you do not know just from reading this post that you need to learn. Go learn some more first is what I am saying then try all you want to develop things on your own.

Lastly, I know exactly how the tables work that you are describing. Do you though. Learn for yourself and see what you come up with. Just learn the very important basics before you jump to this is all. I do not want to see anyone mess their car up by trying to learn something before they new the basics.

[TurboCobalt]

I'll ask the stupid questions:

1) you made sure you have fresh, good fuel, at least 93 octane right? or better question: what kind of fuel and how old is it?
2) you made sure your exhaust doesn't have some heat shield rattling? parts loose, metal charge pipes banging onto something, etc?
3) knock sensors are torqued to spec, not shorted, frayed, damaged wires? (there is also a thread that claims that there may be a benefit to rotating the knock sensors correctly)
4) another stupid question: any codes?
5) just my curiousity, you are able to put down all the torque in first gear at 20 psi? what kind of car / tires / etc are you running? my cobalt would just turn tires into smoke at 20 psi in first.
6) you never specified all the details to the car, solstice/cobalt? complete list of mods, etc.

personal opinion: i know everybody loves to turn off the Cat over temp table if you run catless, i strongly dissagree. Reasons: 1) nobody talks about it, but consider that the turbo is upstream of the cat, therefore hotter, i believe that the COT table is actually more for protecting the turbo from overheating, not necessarily only the cat 2) the COT table i believe is also used partially as a Power Enrichment table. if you are at part throttle, but a high part throttle, such as 50%+, (hmmm, kinda like you are...) the COT table will start adding fuel according to their internal model of chamber temperature (or cat temp model, whichever), enriching the mixture. I agree with the argument that going to lamda of .71 on a catless car may be unnecessary, therefore feel free to manipulate the extreme values of the COT table at your favorite lamda, lets say .85?
Conclusion: I believe that the COT table is used for more than the cat on this vehicle, i suggest you USE IT, even if you change some of those very low lambda values. This one, is gonna be an unpopular concept on this forum.

what you are asking about cam phasing and those main spark tables is know, unfortunately the info is spread all over the place, i did figure it all out at one point, now i forgot the details. it is documented somewhere the physical maximum crank degrees that those cam phasers can actuate, i just dont remember where. the spark values are interpolated between the tables, for sure, they do not snap from one table to another, so that topic is not worth further discussion.

[MikeM173]

I definitely disagree with the COT table. I have definitely tested different stuff with it and did not like the use of adding it in in the general sense.

As for turbo heat... there is ways of measuring that and that should be a part of your tuning process. It is not a direct relation to the turbo but it gives you an idea of what is going on with your motor.

What I just mentioned is probably not something most people know about or even utilize yet it is a huge important factor to understand how it works. These are the basics people need to learn before trying to get away from what the group does. I don't mind people testing tables because I highly encourage it but know what you are doing before you try to test stuff out.

[TurboCobalt]

Come on MikeM173, if we're talking lab testing or OEM development instrumentation equipment, i'll take pressure transducers and thermocouples at turbine inlet and outlet, as well as post cat. while we're at it: pressure transducers and thermocouples at compressor inlet, compressor outlet, intercooler inlet and intercooler outlet. All data collected and synchronized through the appropriate equipment, ETAS, LMS Test.Lab, whichever, pick your poison. Sure would be nice, sure would put to rest all these opinions of where the bottlenecks are, what's worth upgrading and whats not. I don't have a shop, and budget for that kind of instrumentation, I was assuming this conversation here was for the hobbyist that's not putting more money into instrumentation than in the car. If you do have a shop, if you do make a living tuning cars, therefore can spend thousands in instrumentation, more power to you.

[MikeM173]

TurboCobalt,

What the heck are you talking about? None of what I mentioned did I reference anything of needing the such. I put exactly what I know from all of the testing I have done from real world equipment everyone has already on there car. I have done a ton of testing and putting my car through the extremes. This is why I highly encourage people to take ideas and run with them. I only ask that they know the basics of the LNF and some very important safety factors before they do. This is why I made the comments I did.

[ECaulk]

The COT table is not for the turbo, others that have it disabled running high hp builds would of destroyed multiple turbos by now.

OP I took a quick look at your tune, your wgdc is way to high in the 100% desired boost level. Your DAL table is low in the high load areas, which is probably why you have the wastegate slamming shut. Go back through the LNF guide threads and watch Cobaltssoverboosting videos on the initial steps to take. I'd put the timing back to stock and get the maf tables, DALs, and WGDC straighten out then start looking at adding timing.

[wilson0x4d]

It's too hot to do logging today, maybe later tonight. I'll revert any COT changes to stock and give it another go. I maxed it out only to rule out COT temps having anything to do with the KR (it did not help.)

For reference I logged the stock settings and zipped down the freeway and back, the car ran fine (stock001.hpl), nothing like what I see under the previous tune.

As for codes, state of the vehicle, etc: I try to keep everything maintained (fluids, filters, etc) and don't put junk fuel into any car. I suppose I can get in there and verify the sensors are torqued to spec and in good visual condition, clean up the area/etc. I'll make a round and swap oil, clean the air filter and check the plugs this weekend as well.. I won't rule out that something rattles, but I personally don't hear anything over the wind.

Plugs are the colder NGKs, I can swap the OE delco's back in, but somehow doubt that's the issue since the engine is warm, ready, and functions fine under the stock tune.

I reverted the PE table to stock last night and a -significant- amount of KR was removed. I'll review again after my next log with the COT changes reverted to stock, then start making small increments on the PE table to try and see if I can correct the KR (perhaps too lean) since it appears within relatively 'fixed' RPM+LOAD ranges, I can make some minimal edits and compare logs.)

[cobaltssoverbooster]

ill once over the file when i get home but from the few posts i have read you lack the general understanding of the different system operations. as one of my previous posts said...i dont have any time to teach so after i once over i cannot help any further.

-COT Protection is enabled and it gets disabled because it screws with your final afr so bad that also causes KR even on factory ignition tables.
-All of your auxiliary ignition modifiers are still active. if you want to test the base ignition 4 table grouping out that is fine but you first need to default every modifier so your testing only runs off the main table data or else your tests are void.
-Obviously are a super curious dude which is fine but if your going to test the ignition you need to default your cams to run off of a single table per cam. That means run the warm tables in the cold tables and tune them at the same time. If you want to play with hot and cold cams later then you can leave your ignition alone and then adjust cams. My point on this bullet is to default all systems except for the one you are going to test. if you leave a bunch of stuff active you will have shitty results.
-If you want boost you need to lift your dals....running 100% wgdc is bad. ask me how many solenoids i have replaced due to overly compensated tune files. just set your wgdc at 70% and creep your boost up by lifting your dals until you are within a few psi of your overall target. that can leave your dals low to aid in tune stability and help you achieve the linear feel that you were targeting. adjust final boost pressures with the wgdc after this step.
-Be careful with your boost. I would start low and run it up to 22-22.5 psi and stop since your sensors are not capable of reading more than 23. You dont have to run this value, you can run lower but with these sensors when you hit that point you are done.
-Notice your cam factor table...a value of 0 is one table and a value of 1 is the other table. that means the value of .5 allows the computer to use a blend of the two tables. it compares the values in each and then runs a percentage based offset between the two commanded tables. good luck accounting for anything that falls in this range. my advice is leave the cams on one table. its a much more simplified system than the ignition you are testing and without a proper testing facility you wont be able to calibrate the blending function properly.

I really could have gone to town on that tune file but since you only interested in one system, i left the pile of other issues i have out so you can just focus on having clearer results later down the road.
My schedule is slammed and although your posting interests me i unfortunately cannot help you out any more than this. If your still going on this when my schedule slacks up i will come back and follow it.
Good luck mate.