some questions on calc.vet

[estringer]

Ive been doing calc.vet and have come across a few questions...

for reference, this is a stock cubic inch ls1, trickflow 215 cnc heads, vengeance vrx5 camshaft (236/238 .601/,605 113 LSA), vengeance ported FAST90 intake/90mm TB combo, SLP lid, stock 78mm MAF-descreened, kooks 1 7/8 LT headers and 3" Y pipe fitted down to the stock Z28 muffler. I'm also running at roughly 3300 ish elevation.

1. What impact on tuning is the descreened MAF creating?

2. I'm also running a cutout between the stock exhaust and the kooks Y pipe and tuning with it closed. my logic is that the reduction in restriction with it open will cause a leaner condition but will still follow the same basic VE table form, so I may want to set WOT AFR at something slightly rich like 12.5:1 (and obviously double check open cutout AFR values once I have a smooth AFR reading under WOT). am I thinking correctly here?

2. my LTFT have been reset a couple times and thus far the LTFT avg pid logs -1.6 and -2.3. What would cause one bank to have nearly double the LTFT correction percentage? is this normal or should I be looking for a collapsed lifter etc... creating an unequal airflow?

3. My VE map corrections are creating a handful of spikes.. this is an example.

Map before: 21477

Map After : 21478

Is this normal and/or what actions cause these? i.e. blipping the throttle, etc.

4. Ive only primarily corrected/logged light throttle cruising, interstate driving, essentially avoiding PE mode due to an overly rich condition as I have killed one wideband sensor due to a 9:1 initial WOT AFR. this is why the map above shows an incomplete under 70 kpa, above 3000 portion of the map. I did do a few 100% full rpm range corrections with the old sensor before it went kaput and have gotten the AFR to run in the 12's up to about 4800 rpm, where it drops into the 11.5-12.0 range up to 6800.

If I find the maf frequency where it starts to drop into the 11's and decrease the g/s airflow above that point by a percentage or two then blending, am I thinking correctly in terms of how to address this? Joecar, you have answered this on another thread for me.. I wanted to make sure I'm understanding what I'm doing rather than follow without understanding.

Further, Am I correct in thinking also that since the ve table is disabled per the tutorial above 400 rpm that the computer is running entirely by MAF frequency at WOT right now and then also by fuel trims via narrowband o2 in non-pe enriched mode? Meaning I can manually get WOT in the 12.5ish ballpark through manual MAF correction then let selben fine tune it?

Lastly, I'm to the point where the cruising / light throttle portion of the map is doing nothing but creating a moderately cratered surface on the map that I then go smooth out. the area I'm talking about is in the second above photo in the below 65 kpa/ below 3500 range. at this point, how do I know when i'm essentially done tuning that part?

Thank you in advance for any help on these!

[joecar]

Ive been doing calc.vet and have come across a few questions...

for reference, this is a stock cubic inch ls1, trickflow 215 cnc heads, vengeance vrx5 camshaft (236/238 .601/,605 113 LSA), vengeance ported FAST90 intake/90mm TB combo, SLP lid, stock 78mm MAF-descreened, kooks 1 7/8 LT headers and 3" Y pipe fitted down to the stock Z28 muffler. I'm also running at roughly 3300 ish elevation.

Did you start from the stock tune...?



1. What impact on tuning is the descreened MAF creating?

On some vehicles, it makes no difference;
on others it causes the MAF sensor to not see the airflow properly...
the screen is an airflow straightener...
removing the screen allows swirling (turbulent) air to go past the MAF sensor causing it to read incorrectly
( the airflow could be tumbling or bunched up on one side ).



2. I'm also running a cutout between the stock exhaust and the kooks Y pipe and tuning with it closed. my logic is that the reduction in restriction with it open will cause a leaner condition but will still follow the same basic VE table form, so I may want to set WOT AFR at something slightly rich like 12.5:1 (and obviously double check open cutout AFR values once I have a smooth AFR reading under WOT). am I thinking correctly here?
Make 2 tunes and compare.



2. my LTFT have been reset a couple times and thus far the LTFT avg pid logs -1.6 and -2.3. What would cause one bank to have nearly double the LTFT correction percentage? is this normal or should I be looking for a collapsed lifter etc... creating an unequal airflow?
1% or 2% difference between banks is acceptable.
Valvetrain problem (collapsed lifter) will cause a larger split.



3. My VE map corrections are creating a handful of spikes.. this is an example.

Map before: 21477

Map After : 21478

Is this normal and/or what actions cause these? i.e. blipping the throttle, etc.

Is this from VE being corrected from the wideband (of so, is MAF disabled (DTC present)...?), or is this from VE being calculated from MAF...?

Take a look at the 2D view of the corrected VE table, it indicates which cells were edited (i.e. these cells were "hit")...
then bring the un-hit surrounding cells to the same level
BUT: flatten out any tall spikes (positive or negative going spikes);
use only minimal smoothing
( if you smooth something and it comes back, leave it )
drive with smooth/progressive/slow throttle movements
( make sure you applied the transient filter )

Also, are you using LTFT/wideband combination (SELBEN) or wideband only...?




4. Ive only primarily corrected/logged light throttle cruising, interstate driving, essentially avoiding PE mode due to an overly rich condition as I have killed one wideband sensor due to a 9:1 initial WOT AFR. this is why the map above shows an incomplete under 70 kpa, above 3000 portion of the map. I did do a few 100% full rpm range corrections with the old sensor before it went kaput and have gotten the AFR to run in the 12's up to about 4800 rpm, where it drops into the 11.5-12.0 range up to 6800.

When you get this area of the VE corrected (so the BEN is at 1.00 +/-0.02), then by eyeball/hand extrapolate-extend (mimic the trend of) the known hit VE to the un-hit VE... this gives you a better starting point, take a log of a single pull to 4500 rpm at high throttle and look at the BEN.



If I find the maf frequency where it starts to drop into the 11's and decrease the g/s airflow above that point by a percentage or two then blending, am I thinking correctly in terms of how to address this? Joecar, you have answered this on another thread for me.. I wanted to make sure I'm understanding what I'm doing rather than follow without understanding.

Yes, if the MAF is the contributor, and in some frequency range you see wideband indicate richer than commanded, then you would reduce the MAF in that frequency range...

but keep in mind about the airflow straightener, it might cause low speed air to skew the MAF, or high speed air to skew it, or some combination;

the PCM filters/smooths the MAF reading, but the correctness of the resulting MAF reading very much dependsthe MAF sensor seeing the airflow correctly.



Further, Am I correct in thinking also that since the ve table is disabled per the tutorial above 400 rpm that the computer is running entirely by MAF frequency at WOT right now
Yes, correct... disabled VE means that MAF is used exclusively.



and then also by fuel trims via narrowband o2 in non-pe enriched mode? Meaning I can manually get WOT in the 12.5ish ballpark through manual MAF correction then let selben fine tune it?

Yes, you can do that (keeping in mind that SELBEN selects either LTFT or wideband depending on the current operating condition... so you have to drive such that the wideband is selected).



Lastly, I'm to the point where the cruising / light throttle portion of the map is doing nothing but creating a moderately cratered surface on the map that I then go smooth out. the area I'm talking about is in the second above photo in the below 65 kpa/ below 3500 range. at this point, how do I know when i'm essentially done tuning that part?

Drive using minimal throttle movement (i.e. keep throttle pedal as constant as you can, don't move it a lot).
When you smooth, manually bring un-hit cells to same level/trend of hit cells (flatten spikes to this same level).
When you get it right you will see very little change.



Thank you in advance for any help on these!

No worries.

[Supercharged111]

When I smooth out spikes, I do so in a linear manner. My resulting VE map is never as pretty looking as the stock one. How are the cool guys making smooth, pretty looking maps?

[joecar]

They must be hightlighting sections and hitting the smooth button several times.

Generally, long tube headers will cause the VE table to have a few ripples at various rpm's (if you smooth them out, they come back on the next log).

[Supercharged111]

Consistently in the same spot? I remember on a 4 cylinder I had a tough time tuning it because of all the MAP spikes. I ended up putting a MIG welder tip in the vacuum line to the sensor and it only made it less bad. That motor had a header as well.

[estringer]

Thanks Joecar… I’ve replied to your questions below. Im really unsure of how to proceed at this point.. ive added exactly what im doing below in hopes you can point me in the right direction


Did you start from the stock tune...?
I did not start from a stock tune, however I did paste in the stock MAF table from a 98 camaro, then also a stock VE and spark table from a ’02 zo6 I believe. Everything else in the tune I have gone through and compared to a stock tune so that there arent any variables im unaware of. Anything else that was different was my own doing and are things like fan on/off temperatures, cold engine protection, etc.



Is this from VE being corrected from the wideband (of so, is MAF disabled (DTC present)...?), or is this from VE being calculated from MAF...?
I am following the calc.vet tutorial with no deviations.. so I think that would mean the VE is disabled and being calculated from MAF and ltft/wideband feedback?



Take a look at the 2D view of the corrected VE table, it indicates which cells were edited (i.e. these cells were "hit")...
then bring the un-hit surrounding cells to the same level
BUT: flatten out any tall spikes (positive or negative going spikes);
use only minimal smoothing
( if you smooth something and it comes back, leave it )
drive with smooth/progressive/slow throttle movements
( make sure you applied the transient filter )
Okay, ive started to do this now. I was incorrectly working before, attempting to blend the VE table new/old values together. Thus far im one good log in on making adjustments via raising the unhit areas of the VE table to match other raised areas of the table. Im not entirely sure im doing what I should as im not sure what form the VE table should look like for this type of build but im giving it a go.
A thought that I had…. If the VE table is disabled above 400 rpm (thus it has no effect on the engine operation right now)…. Would I not be better off by applying the new values without doing any smoothing at ALL for several logs? Then any spikes may show a proper trend? I could be entirely off my rocker here, this is my first tuning experience of any kind so my learning curve is pretty rough. I don’t know how many logs I should have to take to get a VE table figured. Im just thinking that if the VE table is disabled there is no point in smoothing it until I have enough raw data to show the form of the table.

That said, here is where im unsure of how to proceed! Here is a screenshot of my last log’s cell count. I will also attatch my log at the bottom. 21484 I did about everything I could to populate the entire map, and now I’m unsure of how to really proceed… im operating on the understanding that PE mode is best tuned at only WOT, and not part throttle, so I don’t know how to hit any other areas of the VE table as the values that are left open here I cant hit without entering PE mode. I used braking under throttle, full throttle pulls(pe at WOT only), and anything I could think of to have a complete map of data. I have the PE mode enabled exactly as per the example given in the tutorial. Now I feel like I could go out at do more logs, but I don’t know how I will be gaining anything other that what I have. Am I missing something fundamental in understanding how this works?



Also, are you using LTFT/wideband combination (SELBEN) or wideband only...?
I am using SELBEN as per the tutorial for MAF correction.. paste/multiply with labels. However I am beginning to manually change the MAF values by percentage increase/decrease by referencing MAF frequency to wideband readings taken at WOT… and adjusting the maf frequency where the car begins to run rich as ive made 8-10 logs now (45 mins to 1.5 hours each) and im still running rich at WOT even after the maf has been corrected by SELBEN several times now. Ive only done this when my LTFT AVG has been between 0 to -4 ish as I understand it will add any positive LTFT AVG to PE mode as a safety. Am I doing this wrong?




When you get this area of the VE corrected (so the BEN is at 1.00 +/-0.02), then by eyeball/hand extrapolate-extend (mimic the trend of) the known hit VE to the un-hit VE... this gives you a better starting point, take a log of a single pull to 4500 rpm at high throttle and look at the BEN.
this has me confused. If the VE is disabled, how does the BEN reflect the VE table? Doesn’t it reflect only the MAF correction, then calculates the proper VE table? If I understand you correctly I should watch SELBEN factor on a WOT pull to 4500…. The goal is to have it within .02 of 1.00. but im unsure of what this would be telling me other than that the MAF is where it should be



Yes, if the MAF is the contributor, and in some frequency range you see wideband indicate richer than commanded, then you would reduce the MAF in that frequency range...
I must not understand something fundamental in this process.. I am running with ve disabled above 400 rpm so I am assuming that the MAF can only be the contributor.



but keep in mind about the airflow straightener, it might cause low speed air to skew the MAF, or high speed air to skew it, or some combination;

the PCM filters/smooths the MAF reading, but the correctness of the resulting MAF reading very much dependsthe MAF sensor seeing the airflow correctly.

What do I look for in the logs to determine if this is a problem?

[estringer]

21485

Attatched log

[joecar]

A thought that I had…. If the VE table is disabled above 400 rpm (thus it has no effect on the engine operation right now)…. Would I not be better off by applying the new values without doing any smoothing at ALL for several logs? Then any spikes may show a proper trend? I could be entirely off my rocker here, this is my first tuning experience of any kind so my learning curve is pretty rough. I don’t know how many logs I should have to take to get a VE table figured. Im just thinking that if the VE table is disabled there is no point in smoothing it until I have enough raw data to show the form of the table
Yes, do it like that, take successive logs and watch the VE build itself up.



That said, here is where im unsure of how to proceed! Here is a screenshot of my last log’s cell count. I will also attatch my log at the bottom. https://forum.efilive.com/attachment.php?attachmentid=21484&d=1505249086&thumb=1 (https://forum.efilive.com/attachment.php?attachmentid=21484&d=1505249086) I did about everything I could to populate the entire map, and now I’m unsure of how to really proceed… im operating on the understanding that PE mode is best tuned at only WOT, and not part throttle, so I don’t know how to hit any other areas of the VE table as the values that are left open here I cant hit without entering PE mode. I used braking under throttle, full throttle pulls(pe at WOT only), and anything I could think of to have a complete map of data. I have the PE mode enabled exactly as per the example given in the tutorial. Now I feel like I could go out at do more logs, but I don’t know how I will be gaining anything other that what I have. Am I missing something fundamental in understanding how this works?

Tune part throttle (low-medium load) first, then observe the trend in VE and MAF tables and extrapolate that trend to the upper portions of VE and MAF... and then go and do PE/WOT.

A few more logs (just a few) will fill in the VE gaps (the biggest problem is holding in any one cell for a sufficient number of cell hits, the vehicle just runs away...).

[joecar]

21484

the n button shows the hit counts for each cell;

to see the average value in each cell, click the x-bar button.

[estringer]

21486

[joecar]

...

this has me confused. If the VE is disabled, how does the BEN reflect the VE table? Doesn’t it reflect only the MAF correction, then calculates the proper VE table?
Sorry, I meant MAF.

And afterwards you will want to disable MAF and see how VE performs.



If I understand you correctly I should watch SELBEN factor on a WOT pull to 4500…. The goal is to have it within .02 of 1.00. but im unsure of what this would be telling me other than that the MAF is where it should be
If you get LTFT's close to zero, then at WOT/PE the residual LTFT will not contribute to fueling (it will be zero) so the correction will be from wideband only.



I must not understand something fundamental in this process.. I am running with ve disabled above 400 rpm so I am assuming that the MAF can only be the contributor.
You are correct (I mistyped above, I should have said MAF instead of VE... (( would you believe I was testing you :grin: ))



What do I look for in the logs to determine if this is a problem?
When you look at GM.DYNAIR and SAE.MAF, hopefully they converge.
Same with GM.DYNCYLAIR and CALC.CYLAIR.
SELBEN should be close to 1.
CALC.VET and CALC.VEN should be the same.

Also when done with CALC.VET, disable the MAF, disable CL/LTFT/STFT and use wideband only to see how the VE performs.

[estringer]

Joe,

I am running a 98 ECU/car so I do not have GM.DYNAIR..

Also I do not have GM.DYNCYLAIR selected, Nor do I have CALC.CYLAIR in my logs... all I have is GM.CYLAIR_DMA. is this all a product of having a 98 ECU?

Selben is running +/- .02 almost everywhere which seems good

lastly, watching calc.vet and calc.ven they are typically within .02 of each other at 98% of the time.... on rare occasion they are off by as much as .08-.1 g/kpa

are these acceptable tolerances?

[joecar]

Ah, ok, 1998 (LS1A in EFILive terminology), yes some pids are not available.

If you logged SAE.MAF and SAE.RPM, then CALC.CYLAIR can be added after the log is captured
(while viewing log, goto PIDs tab, click (once or twice) on the column heading Parameter to sort it alphabetically, find CALC.CYLAIR and select it)


Yes, those are quite acceptable.

Questions:
- how is throttle response compared to before...?
- how is driveability...?
- what other improvements do you see...?
- what do you think of the Calc.VET procedure...?

Thanks, I appreciate you taking the time to do Calc.VET and providing me feedback.

[estringer]

Joecar- thanks for all your help on this. I have yet to run speed density to verify the ve table so I cannot comment yet on feedback for calc.vet until I am done with the tuning, but I most certainly will once I do that. It's the least I can do for all your help. Also, in my other thread (sanity check ve table) you requested I post my tune, I would greatly appreciate you looking at it and will do so hopefully tomorrow. I am an on call worker and am on my way out the door as we speak and have barely had the time to eat sleep and the other thing...

[joecar]

No worries, take your time.

[estringer]

Also I ordered a screened Delphi 85mm maf and the bits to go with it on a whim just in case my descreened stock maf was causing a problem...
Is it worth Installing this on a ~500 rwhp car and

I know the maf table will need to be redone however will the ve table drastically change? Or would the new ve table valuea effectively show the gain/loss from a new maf

[joecar]

The screened MAF will not restrict air (not to 500 rwhp).

It may improve driveability.

VE calculation may reflect improvement due to MAF.

We'll see...

[estringer]

okay, so ive finally gotten a break here...



Questions:
- how is throttle response compared to before...?
- how is driveability...?
- what other improvements do you see...?
- what do you think of the Calc.VET procedure...?


throttle response has not seemed to change much. as of this point I have not reenabled the VE table or checked it by failing the MAF. however just by correcting the MAF curve I have found that it SEEMS to run smoother and start up smoother with less idle hunting than before. it may be in my head. also, I am running a spark table that came with my base tune from vengeance that has not produced knock and has more timing down low.

driveability has improved, the car feels smoother all around at this point even with the VE table disabled

AFR at WOT has gone from 11:1 to 12.5-13.0:1 just by correction of the MAF and seems to be pretty consistant.

once I determined that I could watch the table section of efi live scan and determine what cells I'm hitting while driving, I was able to create much better logs

also once I had it in my head clearly that calc.ve creates a CORRECTION FACTOR for the MAF and a REPLACEMENT TABLE for the VE table, I also was able to understand that the correction factor for MAF can be applied only once and if reapplied will skew the MAF table incorrectly whereas the VE table data can be stacked on top of each other from multiple logs until a consistent VE table pattern is showing before I do any smoothing, etc.

I am also unsure of how to tune PE mode... do I do this by modifying the b3618 PE modifier value

i am still fairly uncertain of my work on the VE table... but my concerns cant be substantiated until I reenable the VE table

I will most definitely post a thorough feedback thread once the process is completely done and I can evaluate from a finished product standpoint. I see a lot of posts that ask questions and work through problems but very little in the form of completed tune thoughts.... the people just dissapear

[joecar]

throttle response has not seemed to change much. as of this point I have not reenabled the VE table or checked it by failing the MAF. however just by correcting the MAF curve I have found that it SEEMS to run smoother and start up smoother with less idle hunting than before. it may be in my head. also, I am running a spark table that came with my base tune from vengeance that has not produced knock and has more timing down low.

When you slam throttle open, does soon does engine respond...?



I am also unsure of how to tune PE mode... do I do this by modifying the b3618 PE modifier value
PE and spark are best tuned together on dyno...
you increase spark until you see no appreciable gain in TQ (with no knock) and then you back this off;
you reduce EQR (go leaner) until you see increase in HP (with no knock);
you do then repeat both of those a few times, each time making sure you get no knock.

when dyno time is not avaiable, you set spark close to stock, set PE to safe (NA EQR 1.175, TC/SC EQR 1.28).

And, either way, whenever you drive, you always pay attention to listen for knock (I now do this on any car I drive, tuned or stock).



i am still fairly uncertain of my work on the VE table... but my concerns cant be substantiated until I reenable the VE table

At some point, you will do the opposite of Calc.VET as a sanity check:
you will disable the MAF (a DTC is required to immediately appear), you log, and you compare VE table to BEN, and see if it's close.


I will most definitely post a thorough feedback thread once the process is completely done and I can evaluate from a finished product standpoint. I see a lot of posts that ask questions and work through problems but very little in the form of completed tune thoughts.... the people just dissapear
I know, stuff happens...

and the other thing that makes writing this up difficult incorporating new ideas/methods are found as time goes on (for example, now, with 2 widebands, we have can chose the leanest wideband on-the-fly and use that for the BEN calculation... i.e. this allows MAF and/or VE to be tuned for the leanest bank... I have calc_pids.txt file for this (see attched), and now I have to figure out how to write this into the procedure).