Big off idle stumble, whatt first? IFR? calc.vet? Something else I'm missing?

[Supercharged111]

I had a post in the COS forum, but not sure if it goes there or here. If I'm wrong, please advise. I just swapped a marine intake and Whipple supercharger onto my 1998 GMC with 5.7 Vortec. The intake allowed me to run proper injectors, Bosch 0280155811 at 3.5 bar. Stock they flow 36# at 43.5psi, so using this calculator

http://users.erols.com/srweiss/calcpchg.htm

and some googling I arrive at 38.888 lb/hr and an IFR of 4.8988. I plugged that in (or something close to it) and the truck fired right up. Trims are off which I wasn't expecting at low load plus I have this NASTY off idle stumble (all other transients are fine) that's worst when the engine is cold. I don't want to go changing a bunch of crap in the wrong order. The truck seems to run better if I fudge the IFR down I believe in the 3.55 range but I assume that's just covering up the real problem. So should I have faith in my IFR math and just go do a calc.vet or is there more than meets the eye when it comes to nailing down a new IFR? Or is there some super simple idle setting I'm totally missing here? My only thought is the addition of some real estate between the engine and the MAF with the supercharger being in the mix now. Physically its location has remained the same as stock.

21217
21218

[joecar]

Get VE/MAF correct.

Make sure injector data correct.

I'll look closer later tonite.

~ posted by phone ~

[statesman]

I'll look closer later tonite.

There's a lot to look at in this tune file.:laugh:

[statesman]

I have this NASTY off idle stumble (all other transients are fine) that's worst when the engine is cold.

Post a log.

[Supercharged111]

Here's a log.

21221

I guess my biggest fear is committing to the mathematically correct IFR, doing a calc.vet, and then realizing that for some reason the IFR is wrong and having to start over again. Sometimes I have KR, sometimes I don't. I really need to get the fuel right on this thing so it's more consistent. I thought I'd pulled enough timing, but I suspect now that since the bypass valve closes before atmospheric pressure is achieved, the IATs are higher at a lower MAP than I'd anticipated. I don't really want to monkey with that until the fueling is correct. I was about to do a calc.vet yesterday, but decided to wait on what I heard in here first. So can I trust that 4.8988 IFR and should do the calc.vet with it and report back? The PIDs that I have currently selected to log are the calc.vet PIDs, but I don't think any of the logs have enough data to try and correct anything. Plus I'm running the wrong IFR to fudge more fuel into it before trims now in a ghetto effort to help the off idle stumble. It's there with a warm engine too, just less bad.

I'll add that I didn't do the quasi-closed loop operation because the truck is only making 3.XX# of boost which gets me up to 100 kpa on a cool night. I think I'm going to do it though to add a little knock resistance to the range right where the bypass valve closes but I may or may not be exceeding atmospheric.

[Supercharged111]

There's a lot to look at in this tune file.:laugh:

Could you please elaborate instead of being a douche about it?

[joecar]

Could you please elaborate instead of being a douche about it?I think he's referring to me looking at it...

[Supercharged111]

It sounds like he's implying that I have a lot of things wrong. If that's the case, why not lay it out instead of making you do it all? If not, I gotbmy panties in a twist over nothing.

[statesman]

Could you please elaborate instead of being a douche about it?

It's Aussie humor, and it looks like some people just don't get it.

[statesman]

It sounds like he's implying that I have a lot of things wrong. If that's the case, why not lay it out instead of making you do it all? If not, I gotbmy panties in a twist over nothing.

If I knew what was wrong, I wouldn't need you to post a log.

Nice panties. :moon1:

[joecar]

It's Aussie humor, and it looks like some people just don't get it.you might have to explain it mate...

[joecar]

It sounds like he's implying that I have a lot of things wrong. If that's the case, why not lay it out instead of making you do it all? If not, I gotbmy panties in a twist over nothing.I know he's not having a dig at you, trust me :)


remember, we are different (or everyone else is different), we down under view things differently, my wife (born in Anaheim) always reminds me :)



21227

[joecar]

Compared to a stock file there appear to be many changes (unless I have the wrong stock file to compare against)...

did this file start life as a F-car file or as truck file (looks like truck file)...?

[statesman]

you might have to explain it mate...

Everyone knows you've got a busy life, but you still come in here and try to help people. Sometimes it takes a few days before you can look over someone's tune file... "There's a lot to look at in this tune file" is Aussie humor for "you might be waiting a while before Joe gets back to you".

[Supercharged111]

Compared to a stock file there appear to be many changes (unless I have the wrong stock file to compare against)...

did this file start life as a F-car file or as truck file (looks like truck file)...?

Started as a van I do believe, then converted to COS5, then got a 4L80 added in, and most recently added the 2 bar setup. The only thing preventing me from reverting my IFR and doing a calc.vet is this assumption that I'll adversely affect transient fueling. Since I'm not aware of any transient parameters that I can change, I ASSume the only way is to fudge it in the IFR. Kinda like an older 5.0 Ford I was playing with. Didn't have tuning software, but did have an adjustable FPR and that was the only thing that fixed the off idle stumble in that car (which had better flowing heads and intake and exhaust than stock).


Everyone knows you've got a busy life, but you still come in here and try to help people. Sometimes it takes a few days before you can look over someone's tune file... "There's a lot to look at in this tune file" is Aussie humor for "you might be waiting a while before Joe gets back to you".

I know he's a busy guy so I don't want him needlessly looking too much at a tune that I think I know what is wrong with. On the other hand, the only thing I have to lose is gas, but I gain experience. I should probably point out that driveability is good enough for me to DD the thing. When warm I see the wideband peg full lean momentarily but it doesn't seem to stumble and fall on its face so badly then, only when cold. I'm kinda curious if there isn't just a bad handoff from the idle stuff to the VE. Either way that requires VE to be correct. How can I know if my calculated IFR just isn't jiving with the tune?

[statesman]

There's something wrong with your setup. At 3000rpm WOT, you're only getting 96kPa manifold pressure... even at a high elevation, you'd expect more than that with a boosted engine.

What MAP sensor are you using? What elevation are you at?

[statesman]

I'm kinda curious if there isn't just a bad handoff from the idle stuff to the VE. Either way that requires VE to be correct. How can I know if my calculated IFR just isn't jiving with the tune?

The VE is very forgiving towards bad injector data... you can use any junk injector data and then bend the VE table to work with it. The MAF is not so forgiving... use bad injector data and you'll never get the MAF to work right.

[joecar]

lol, that's what I thought :)

I look at everything, I learn alot from it (and I don't mind), and I'm not going to let lack of time deter me :)

statesman is right, I have so many things going all at once (family health, job, 2nd job, upgrading my Trans Am, EFILive, keeping up at LS1tech, sleep) but I can't give up any of them.

[joecar]

+1 what statesman said about VE and MAF tables... and the place where you see it show up the most/worst is at idle.

[Supercharged111]

I'm at 6000' so a typical day is 77-79kpa. I believe that puts me in the ~3.5# boost range on something that claims about 5# at sea level on a stock engine. Mine has a freer flowing intake manifold and exhaust setup than stock. Cam is very close to stock, just has a smidge more duration and lift but better airflow nonetheless. I'm not sure how significant of an impact the elevation should have on a mildly boosted engine. Here's the MAP that I used, ACDelco 213-1631 from a Supercharged Cobalt I do believe.

https://www.amazon.com/gp/product/B000C9JPQO/ref=oh_aui_detailpage_o02_s00?ie=UTF8&psc=1

I should also mention the only injector data that I changed was IFR, everything else I left as it was.

[joecar]

You can test that MAP using a vacuum hand pump (that has a vac gauge) and by rigging up some shop air fittings (probably have to drill a hole) if you have an air pressure gauge.

[Supercharged111]

Well with key on engine off it registers high 70s for MAP so I think I got that portion right. My concern is what was said about IFR vs VE/MAF. If I have injector data wrong, VE won't care so much but MAF will. Now I'll admit it needed a calc.vet before the intake and blower because the 411 engine portion from the van has different MAF data due to its different air intake tube/box setup. It was enough that I needed to fudge PE fueling but the trims elsewhere were minimal as I recall with no stumblewhich is why I was lazy and left it as is for so long. It seems to me the truck ought to run about the same as before at idle and low RPM/low load. Should my next step be to put that 4.8988 IFR back in and do a calc.vet? Or is that stumble an indication that I've missed something else? I just want to make some forward progress here and report back if I'm still having trouble.

[joecar]

I would try doing a Calc.VET with the injector data that is more correct.

[statesman]

What's the history of this engine? Did it have the stumble before installing the supercharger? Did it have the stumble before switching to a COS. Can you post the original OS tune file you were using before switching to COS.

[Supercharged111]

It's never had the stumble. The 411 had COS5 before it was ever ran on my truck and in fact it also had the 4L80E segment swapped in before I swapped it in. Truck ran like a dream that way. The most recent changes (intake manifold , blower, 2 bar) are when the stumble started and I did all 3 things at the same time. Unless I screwed something up with the 2 bar, I don't think that that's to blame here as I indicated earlier where the MAP scaling looks to be correct. The intake manifold is damn near identical to the factory piece but it lacks the spider in the path of the airflow through the plenum which NA guys report makes for more pull starting at 3k. Throttle body has a little more real estate between it and the MAF now with the blower. Bypass valve is working properly. Sorry I'm not home to post anything else but I am thinking of unplugging the MAF and going for a spin to see if the stumble goes away. That would indicate that the MAF is extremely unhappy with the most recent changes, am I right?

[statesman]

Sorry I'm not home to post anything else but I am thinking of unplugging the MAF and going for a spin to see if the stumble goes away. That would indicate that the MAF is extremely unhappy with the most recent changes, am I right?

If the IAT sensor is incorporated into the MAF, then it's not a good idea to unplug the MAF sensor.

Just set the MAF to fail in the tune and test it that way.

[Supercharged111]

IAT is separate and has been relocated downstream of the blower to read hot, boosted air. It's going to make the trans shift like crap without disabling those transmission parameters, but I probably won't even get off base before plugging it back in and deciding if it helped.

[Supercharged111]

Embarassingly enough I have no gas money to burn on a calc.vet right now, but one observation I made is that unplugging the battery to weld on the truck seemed to erase the IAC trims or something because when I restarted the truck it would barely idle, stumbling around under 500 RPM for a bit. Go for a spin, come back, and everything's fine. Not sure if that's an indicator I should concern myself with or not.

[statesman]

Sounds like you need to do some idle tuning.

[Supercharged111]

Yeah I wonder why though? I don't think the intake or blower affected idle VE.

[statesman]

The blower is a load on the engine.

[Supercharged111]

Probably not much with the bypass valve open, but you do raise a good point. Also realized I never raised idle timing with high octane timing and there's a huge gap. This may help the engine idle with less effort since it's now an issue for me. I found that idle writeup for cammed cars. Even though mine's not cammed to the point of lope, there may be some good nuggets within for me. I'll go slap that new IFR in there for now. Not sure if I should bump that idle timing table though? I realize I want it somewhat in the middle as the PCM uses timing tweaks as a way to control idle speed in addition to the IACV, correct?

[Supercharged111]

I just glanced at the VET values in my logs and they approach 2.0 on the cells with higher counts. I had to have fvcked something up on that 2 bar swap. Trims aren't maxxed out though, but a VET and SELBEN value of 2 just seem wrong don't they?

EDIT: Perhaps I should first apply that pesky filter.

[Supercharged111]

So I'm glossing over the tutorial again and noticed that, for 2 and 3 bar applications, the MAF must be disabled. Doh! Thought it could be left in there and that it had to since spark uses the MAF. So without the MAF then the PCM just uses MAP and IAT to determine air density and airflow?

[joecar]

COS uses spark (HO/LO adaptive) regardless of MAF or SD.

Start with your objectives, which one of these is your intent:
1. to correct MAF using WB...?
2. to correct VE using WB...?
3. to correct MAF using WB and LTFT...?
4. to correct VE using WB and LTFT...?

Note that 1 and 2 require disabling CL and trims, and disabling the other air source.

Note that 3 and 4 require disabling the other air source.

In all cases you correct the air source and (if you wish) you calculate the other air source (... noting that CALC.MAFT requires the OS to support the pid GM.VETABLE_DMA)... note the distinction between "correcting" and "calculating".

If you pm me your phone # I call walk you thru it tomorrow sometime.

~ posted by phone ~

[Supercharged111]

Thanks Joe, I don't think I want to waste your time until I've actually tried and failed. I have a full tank now and fully intend to try this this weekend.

[Supercharged111]

W02LAM1 is pegged at 1.701 whether it's plugged in or not. I won't get anywhere until I figure this one out. What a fuckin waste of gas. Going through the tutorial again now. This explains the VET values of 2+.

[Supercharged111]

Alright, seems I'm a retard on multiple accounts. I made questionable progress on the VE map but feel pretty good about the MAF. I didn't really get enough data to do VE because I didn't want to take another 45 minute hike only to have screwed something else up again. I started the day by flashing what I believe to be accurate IFR data and then proceeded to log for the calc.vet. I had trouble smoothing the map and eventually said screw it and just flashed what I had. I stopped short on the MAF so I took the highest cell I had and pasted that data into the remaining MAF cells. I noticed that the values pegged before the MAF frequency did. This is all unchartered territory for me. On the way back the logs indicated that the MAF data I changed based on logs seemed to be closer, and the stuff I guessed on just needed a little more. The thing I didn't realize about VE all this time is that it's your new value, not a correction factor like AutoVE or the MAF data in calc.vet. Derr. Anyway, Log 9 is the log I used to make my 1st change and Log 10 shows the result of that change. Tune V5 is with the IFR change and V6 has the changes I made. Because I filtered out data below 800 RPM my stumble isn't gone. I really think if I were to change the VE below 800 that I could fix that damn stumble. Do you guys have any tips for VE smoothing, guesstimation, or any other trickery to make the truck do what you want? Are my VE correction values way out in left field? Does it look like I'm bending this stuff all around an IFR that is mathematically correct but otherwise wrong?

21258
21259
21260
21261

[Supercharged111]

Sat down today with a cup of coffee and took a guess and smoothed things out and fudged to the best of my ability. Now to see if it helped. I didn't touch the MAF, it's going to peg by 11000 by the looks of things though. Is that weird? Do I have my IFR wrong somehow? I suppose I can go measure actual fuel pressure to see, but I'm not sure how accurate it'll be at altitude. I guess just add 3 to whatever I see to have my sea level value.

21262

[Supercharged111]

Alright that tune did NOTHING to fix my stumble. I guess I'lm search for an idle VE tuning tutorial. I'm running out of ideas here.

[joecar]

I'm looking at your recent files (in the last day)...

( what was wrong with the wideband pegging at 1.7, was the sensor not working...? )

[Supercharged111]

It had been so long since I plugged the serial WB into the Flashscan that it was still set for my Techedge wideband. That's now my floater WB and the truck has an AEM permanently installed.

How helpful would this be for a guy who doesn't have a big cam?

https://forum.efilive.com/showthread.php?4661-Idle-Tuning-Helpful-Info-Inside

[joecar]

That would be every helpful, regardless of cam.

[Supercharged111]

Alright won't be able to chase the idle for a bit as a friend is borrowing the truck, but I feel like I got the VE and MAF mostly squared away. I will be going with the semi open loop and probably re-implement the factory PE timing adder now that boost is able to build in closed loop. When I ease down into boost and then into PE the truck seems to take off when it gets the fuel, so I think it would like those changes. For now I've simply pulled 2 degrees of timing off the whole table as I'd probably gone too far NA anyway in search of torque that just wasn't there. I ended up basing my final tune off of 2 separate logs, 1 that had little if any PE operation and another with like 5 WOT pulls through 1st. It matched another log that had a bunch of WOT pulls and both matched the log that had little/no PE operation. I was hesitant to apply the values I observed in the 95 and 100 kpa range because they were dramatically different than what was in there. Much more so than any of the other cells I'd already fixed. Same with the MAF, it pegs way before its max frequency now. I took a leap of faith and went with it and the truck is now within a couple tenths of delivering commanded WOT AFR. It may trim itself out, who knows. I would like to smooth it better if possible. All I did for boost VE was observe the uptick across the 60-90 kpa range and extrapolate that with a small extra margin in the boost table. Since I'm commanding 11.3:1 in PE mode I have some wiggle room in either direction as the truck is headed from 6000' down to 3000'. I also added in some boost timing retard since I won't be there with the truck. If it all goes to shit, I gave him instructions on how to get the bypass valve to open under boost and eliminate it altogether, but I don't think it'll come to that. I think my boost VE guess is way closer to right than what my normal VE started today as. I'm feeling more confidence in the process, so all I really need right now is validation in what I've done so far. From there, I just want to be sure that I haven't tuned around an erroneous IFR and fix the damn stumble and I can repeat the process on my dually. And my Z06 since I just picked up a Kenne Bell for it yesterday. :cucumber:

21267
21266
21265

Log 12 is one that I pored over to fix the VE and MAF. Log 13 validated those changes and then I tweaked boost VE before handing the truck over. I'm just spitballing the 400 RPM values here as I'm filtering out everything below 800 per the calc.vet instructions.

[joecar]

If your MAF is maxing out (512 g/s is the limit) then you may have to scale (there are various ways to do this).

[joecar]

...

21267
21266
21265

Log 12 is one that I pored over to fix the VE and MAF. Log 13 validated those changes and then I tweaked boost VE before handing the truck over. I'm just spitballing the 400 RPM values here as I'm filtering out everything below 800 per the calc.vet instructions.I'll look thru those today.

[Supercharged111]

If your MAF is maxing out (512 g/s is the limit) then you may have to scale (there are various ways to do this).

I did scale with calc.vet, that's what caused it to max out before the max frequency. Is that an indication that IFR is off?

[joecar]

If your IFR is correct (matches actual injectors), and you hit the MAF 512 g/s limit, you then have to scale MAF, VE and IFR by some factor...

e.g. multiply each by 0.8 for example...

and then you have to look at all the tables that reference g/cyl (or simply g) and shift them down by the same factor

e.g. following on with 0.8 example, the HO/LO spark tables, you shift the 1.0 g column's contents to the 0.8 g column, 0.80 g col -> 0.64 g col...

you will have to do some interpolation

( yes it is tedious )

( yes, it is easier to scale by 0.5 instead of 0.8 )

[joecar]

BTW:

terminology

scale = multiply the MAF, VE and IFR by the same factor (all together).

correct = multiply either MAF or VE (not both) by a correction factor.

calculate = calculate (using mathematical equation) either MAF or VE (not both) from the other.

[joecar]

( note that if you're not hitting the 512 g/s MAF limit, as in non-LS1 ECM's, there are alternate ways to scale, e.g. IFR and STOICH )

[statesman]

If your IFR is correct (matches actual injectors), and you hit the MAF 512 g/s limit, you then have to scale MAF, VE and IFR by some factor...

e.g. multiply each by 0.8 for example...

and then you have to look at all the tables that reference g/cyl (or simply g) and shift them down by the same factor

e.g. following on with 0.8 example, the HO/LO spark tables, you shift the 1.0 g column's contents to the 0.8 g column, 0.80 g col -> 0.64 g col...

you will have to do some interpolation

( yes it is tedious )

( yes, it is easier to scale by 0.5 instead of 0.8 )

It might also be worth pointing out that scaling the fuel and airflow affects calculated torque, which affects transmission shifts... so if you have to scale, then it's best to scale by the least amount possible (scaling by 0.8 is better than scaling by 0.5).

[Supercharged111]

Is any of this an indication of bad IFR? I really expected the truck to run normal pre-calc.vet. But then again I called these things 35 pounders and it still had that stumble. I'm confused as to the point of tuning the MAF if, when you do, you just end up scaling it back and undoing what you did in the first place? Would an SD tune make my life easier? I saw the COS tutorial stated that you had to but I have a friend who didn't. Not sure the state of his tune at the same time.

[statesman]

I'm fairly sure that the injectors you are using are the same injectors used in the ASA tune. Get a copy of the ASA tune and use the injector data from that tune file.

[Supercharged111]

Man I forgot all about that!

https://forum.efilive.com/showthread.php?20205-corvette-ASA-2000-tune-strange-spark-v-high

Well I can put that one to rest finally. I'm not 100% following why MAF scaling is required? IFR notwithstanding. It was scaled over farther before I started messing with it, so now I'm kind of lost on the point of messing with it. Unless the start and end values remain the same and it's everything between them that gets tweaked.

[joecar]

With the LS1 calibrations, there is an internal limit on the MAF at 512 g/s, so the MAF flatlines at 512 g/s.

This only matters if you're running MAF.

If you run SD then DYNAIR may flatline at 512 g/s, but this has no effect on VE g*K/kPa (i.e. VE is calculated ok, and DYNCYLAIR g/cyl is calculated from VE and is ok).


Post another log file, let's see if you hit it (with boost you should).

[joecar]

It might also be worth pointing out that scaling the fuel and airflow affects calculated torque, which affects transmission shifts... so if you have to scale, then it's best to scale by the least amount possible (scaling by 0.8 is better than scaling by 0.5).Yes, a very good point.


( not only should any table referencing g/s or g be shifted down/left, but so should any table referencing anything that could be calculated from g/s or g ).

[Supercharged111]

With the LS1 calibrations, there is an internal limit on the MAF at 512 g/s, so the MAF flatlines at 512 g/s.

This only matters if you're running MAF.

If you run SD then DYNAIR may flatline at 512 g/s, but this has no effect on VE g*K/kPa (i.e. VE is calculated ok, and DYNCYLAIR g/cyl is calculated from VE and is ok).


Post another log file, let's see if you hit it (with boost you should).

Well my truck is only making like 3# on top of what we call atmospheric, so it only got up to 9000. Truck basically makes sea level power up here now, but now that the fueling is so much better I'll have another look at timing and order a smaller blower pulley. I imagine I'd still have to go to sea level to get it to peg the stock MAF. What do I lose by going SD? Anything important? Does the truck become more sensitive to weather/altitude changes? Does keeping it make tuning weird? Like with the handoff from the hi timing table to the boost timing table? I know I just corrected the MAF and calculated the VE. If it used erroneous MAF values to calculate the VE then something has to be off even though delivered fuel seems to be tracking with commanded. Maybe we should have that talk now, I know a little bit more of what I didn't know before. :rolleyes:

[Supercharged111]

It appears my IFR is damn close if not bang on. I may just slap an LS1 regulator in so I can plug the 0 kpa value from the LSA tune in across the board on mine. I have 39#/hr in my tune which is at 3.5 bar and the ASA tune seems to base its numbers on 40# per hour at what I ASSume is 4 bar. My math says 42#/hr at the LS1 4 bar so I may still be off a little. Easy enough to make it the way GM intended.

[joecar]

MAP referenced FPR does provide benefits in that GenIII table B4001 only handles MAP in the range 20-100 kPa (i.e. vacuum in the range 0-80 kPa).

Same regarding the table B3701.

[Supercharged111]

So for boost does it just use the 0 kpa value? In my case that works nicely.

[joecar]

So for boost does it just use the 0 kpa value? In my case that works nicely.Yes, for the range of MAP (boost and non-boost) it uses the 0 kPa values...

but you have to edit B4001 and B3701 as follows:
- that all cells (B4001) are set to the 0 kPa cell value;
- all rows (B3701) are set to the 0 kPa row.

The reason for this is mathematically explained as follows:
absolute rail pressure = base_pressure + MAP
absolute manifold pressure = MAP
pressure differential across injector = base_pressure (i.e. subtract those 2) = constant.

[Supercharged111]

Yeah, I noticed it was that way in my tune, i.e. the difference between a referenced FPR and not. Having said that I realized, derr, an LS1 regulator won't help me if it's not referenced. I think then that the smartest thing to do here, in an effort to make sure IFR is as accurate as possible, is to measure my fuel pressure with the vacuum line disconnected and calculate IFR based on the known pressure and flow on the ASA tune. I think because of my elevation I'd need to base it on the 20kpa cell. Or better yet see if a 454 regulator will fit, that was 4 bar and used normal injectors.

[joecar]

A GenIII truck regulator would work (you might need fuel rails to match), you would have to add a return line back to the tank (where's the tank on your truck...?).

[Supercharged111]

My truck has a return line. It's a GenI engine, but existed when the GenIII stuff was coming online. The late GenVI big block was manifold referenced and likely fits my rails. The truck originally had this spider garbage inside the intake manifold.

http://prosilversystems.com/images/Engine/oldintake/100_1351.jpg

But along with the blower, I swapped on this marine intake instead so I could feed a boosted motor. Same shape but freer flowing plenum with no crap in the way.

http://bertok.us/pics/l31marine/6/cpi-wired.jpg

Those pics are notional, since Photobucket got greedy none of my pics will work anymore. I didn't keep that giant connector, I chopped it off and peeled back the harness so I could solder in all 8 injector plugs as cleanly as possible with as much of the factory wiring as possible. 6 were long enough to reach the pigtails I had on hand and I just made 2 from the just enough scrounge I had laying around. Same with the IAT, it used to be up by the MAF but it was long enough to reroute and pop in at the elbow where the throttle body was as delivered from GM. I actually have the original IAT still installed to plug the hole and used a 2nd to measure boosted IAT. It's a very clean and subtle installation and I like that, I just wish the blower was bigger. I think if I change the truck over to hydroboost that the 2.3L compressor will bolt right into place, it's just that the vacuum booster would interfere with the throttle elbow as it sits now. I'll probably just pulley it down and live with it since I have so many other projects to deal with, it's still got more torque than it ever did NA and doesn't seem to fall off up top. Likely because the compressor is a twin screw and not a roots.

[joecar]

In the second pic I see that the regulator is MAP referenced.

Which of those 2 lines on the regulator is the return line...?


Yes, Photobucket has recently gone stupid.

[Supercharged111]

I've got to wonder if this is being caused by the change in trajectory of the injectors? They don't have as straight of a shot at the back of the intake valve. Could this be cylinder wall wetting or something else I've yet to deal with? I say this because, without a permanently installed wideband to stare at, you'd never know it leaned on tip in when fully warmed up.

[Supercharged111]

This is another thing kicking my ass. For some reason the VE map picks up again after torque peak which should occur around 3200. I assume the blower is causing this phenomenon and I also assume my VE does not rise again. Something I'm missing?


21341
21340

I consider the map a work in progress. It's still a bit rough and I haven't yet ventured beyond linear blending.

[statesman]

It doesn't look like you're using the injector data from the ASA tune yet.

[Supercharged111]

No, that's going to force me onto another calc.vet and I want to get the next one right. My VE map shouldn't look like it does now after the next round. WOT AFRs are a full point richer than they should be.

[Supercharged111]

New pulley took me from 99kpa to 109kpa. LAME! This blower is way undersized. Anyway I copied over the ASA injector data save for bank assignments and flow rate which is likely cooked into my existing tune. I think I need to take a whack at RAFIG and RAFPN. I noticed when my battery died the other day that the truck really struggled to idle at first. It must be using airflow trims to correct itself, because after it warmed up it was fine. Does this sound logical? I logged grams/cyl and that is peaking around 4400 RPM. I forget if that it airflow per revs or per time. The IATs are higher than I thought they'd be, even cruising. That must be why the top end feels so weak, it's around 160-200 under boost which costs me 5 degrees of timing. I'm only seeing trace amounts of knock at WOT, but also only getting around 22 degrees of timing. NA was closer to 33 degrees and I've only picked up 30kpa worth of air!

[Supercharged111]

I did a couple RAFPN logs and noticed that the datalogs are indicating a consistent 2+ gm/sec less than the stock tables. The curve is consistent with stock too, nothing erratic. Vacuum leak?

[joecar]

I've been away for a few days, I'm catching up on the forum.

[Supercharged111]

Well I hosed down the top end with carb cleaner with no effect to the idle so a vacuum leak seems unlikely.

[statesman]

I did a couple RAFPN logs and noticed that the datalogs are indicating a consistent 2+ gm/sec less than the stock tables. The curve is consistent with stock too, nothing erratic. Vacuum leak?

Post your latest tune file.

[joecar]

+1 post latest current files.

( are you running SD, if so make sure you see MAF DTC )

[Supercharged111]

I think I finally hit the jackpot randomly digging through the cax forum.

https://forum.efilive.com/showthread.php?9135-Expanded-Support-for-Transient-Fuel

https://forum.efilive.com/showthread.php?9127-Tuning-the-transient-parameters

Here Howard explains EXACTLY what I am going through to a T! I guess I'm just a little late in the game. I do remember reading years ago about the different spray patterns affecting cold running.


I'm pretty sure that Howard has played with a few CAX files. One of them being the transient fueling tables. The marine intake retrofits will experience a lean stumble when applying throttle, it's worse when the engine is cold or if you have a supercharger. I really havent heard of an intake swap that doesn't have this issue. James B. ,who uses Tunercats OBD2 worked his issue using VE tuning among other things. It's been years since 2005, but I think James even posted screenshots of his VE tuning and it was considerably differnt than stock.

the marine intake has significant injector aiming differences than the stock L30/31 does. So when the injector fires, more of the pattern is deposited on the walls of the intake runner. There are a few different terms used for this subject. There is "wall wetting", "transient fueling", "evaporation rates" etc etc etc. I would love to see a full list of comments on both a blackbox PCM and a 411 PCM carrying a L30/31 calibration. I have a full L31 calibration in hex and the code has 300,000 lines. This pales in comparison to the newer PCM/ECM's which have millions of lines of code. The L31 code is not fully commented,. but the more important stuff is there. I remember when Boosted Z71 had one of his GM Powertrain friends tune his truck. Boosted fired up his copy of TC2 and the engineer took a peak and commented how few of the blackbox PCM parameters were supported com,pared to how many are actually contained within the code and are available to GM Powertrain calibration engineers. Remember that this was with TC2 and TC2 has the most Vortec 4.3/305/350/454 parameters supported when compared to HPTuners/EFILive. IIRC GM was using ETAS/INCA at the time for their PCM calibration and PCM flashing. I have "heard" that this stuff is worth 10's of thousands of dollars.

[Supercharged111]

Turns out it was in there all along. B3406 but I'm not seeing the results I want and my table is more than DOUBLED in some areas. This is ridiculous. It was so ineffective that I maxxed the table and flooded the engine and had to revert. I need to figure out how to smartly change this table instead of just doing global bumps. Here's stock and increased Impact Factor. I can still watch the wideband peg full lean on a warmed up engine on tip in.

21358
21359

Next is my stock Desired Airflow table and a log from a "cold" start (it's summer). Filtering didn't do anything to the 80 degree cell. Notice how much lower it is than stock. Truck idles a little rough for some reason so I figure this needs a look. It does fire instantly, but as previously mentioned when it forgets its idle trims it really struggles until it warms up and is allowed to learn.

21360
21361

And finally here's the current tune with increased Impact Factor and RAFPN log.

21362
21363

[joecar]

I think I finally hit the jackpot randomly digging through the cax forum.

https://forum.efilive.com/showthread.php?9135-Expanded-Support-for-Transient-Fuel

https://forum.efilive.com/showthread.php?9127-Tuning-the-transient-parameters

Here Howard explains EXACTLY what I am going through to a T! I guess I'm just a little late in the game. I do remember reading years ago about the different spray patterns affecting cold running.Good find :cheers:

that was a long time ago, I had forgotten about those threads.


Note: some of the cax TF parameters are duplicates of existing OEM TF parameters, so just pay attention to the tables you look at.

[Supercharged111]

Yeah, I finished reading the thread and saw that, in a sibsequent EFILive release, thise cax transients were unlocked and included in the software. The desktop had the right software so I quickly just copied the 12212156 and 0202005 calz files to my laptop to unlock the fuel dynamics. I believe wall wetting fixes lean tip in and evap fixes fat tip out. Seems that it requires significant changes to make any kind of difference though, I need to do more digging so I can approach it motw methodically. What about capturing logs and filtering to show only transients? I.E. only when TPS is changing so many percent per second? The thing with this intake swap is that its injector trajectory sucks compared to factory. Maybe I can just rip off some LS1 data.

Nope, LS1 stuff was the same as my stock values pretty much.

[Supercharged111]

Gonna try B3426 next. Finally found something that describes them better than the software does. It's the bits in red that I hope to be my smoking gun.


Further to that....


B3406: Impact Factor

Is a fuel correction "adder", so it increases the amount of fuel injected. It is calibrated via ECT and MAP to allow for different corrections based on temp and load. This table tends to have greater influence on smaller throttle movements rather than slamming the throttle open.

B3401: Evaporation Time

Tells the PCM how long it takes for the fuel left on the wall to make it into the intake charge. Again, this is a "correction" factor, not an actual figure for the evaporation time. Higher numbers here will extend the time that B3406 is used for as it indicates a longer evaporation time correction is needed. again, it tends to have more influence on smaller throttle movements.

B3426: Stomp Compensation

This tells the PCM to inject additional fuel when jumping on the throttle. Think of it as a pump jet in a carby based engine. As would be expected, bigger numbers = more fuel.

B3428: Stomp reduction

This tells the PCM the rate at which the "stomp compensation" is ramped out. Higher numbers result in a raster ramp out.


So for a lean tip in, I would look at....


Increasing B3426 to give a slightly rich spike at tip in spike
Adjust B3428 to blend the rich spike into the normal commanded fuel. Fir the spike goes, rich-lean-commanded, increase the factor to stretch out the rich spike. If it stays rich too long, reduce the factor to lean it out quicker.


For the rich spike at throttle close....

I wouldn't worry as much, but it can cause issues if you are running CL and the STFT's pull lots of fuel just prior to the throttle being snapped open. You could try reducing the values in B3406 and B3401 for the cells effected. This would normally be in the very low MAP cells, but may need some correction higher up to "catch" the rich spike earlier on.

If you really want to try it out, you can also mess with the Injector timing to mess with dynamics. The basic rule here is...


the earlier the injector fires, the more time the fuel has to evaporate


So an early injector event results in more of the fuel being used for the intended induction event. ( IE. less fuel left on the wall )

I've done some testing with firing the injector just after the inlet valve closes. Doing this you can pretty much zero out the impact tables, but you may also miss time the throttle stop fueling as the injector has fired before the PCM signals it needs stomp fuel.

I've also fired the injector as the inlet valve opens, it makes the stomp fueling pretty good, but needs lots of work on the impact tables to get things working.

Also remember that at full song, the injectors will be open for longer than the inlet valve is open......

Simon.

[Supercharged111]

I reverted back to a tune with a more sane 30% global Impact increase and tried a 20% global Stomp increase. I only took a quick jaunt, but it seems like Stomp may be there and Impact needs more. A lot more. Initial tip in is lean and as soon as you reach for more it lurches, but the motor hadn't cooled off that much so it was tough to tell. A trip to work tomorrow should be more telling. I just wonder if I'll need to fudge those 100kpa cells for boost? I can only get 109kpa here but I'm sure at sea level it will be less pitiful.

[statesman]

Set all values in {B4003} and {B4004} to 0.9 or you're gonna get a rich decel.

Set all values in {B4105} to 450 and...

Put these values into {B0304}
0.659912
0.705078
0.750000
0.800049
0.850098
0.895019
0.949951
0.995117
1.000000

Are you using a stock throttle body?

[Supercharged111]

I do have a rich decel, it's just that the lack of drivability from lean tip in was way more annoying. I'll plug those values in when I get home. I do have the stock 75mm TB still.

[Supercharged111]

Finally got to this, I'd been wondering about minimum pulsewidth after reading some other threads. What was the 2nd axis on that O2 switch point map and why wasn't it already set to 450? Weird. What's with the baro stuff?

[statesman]

What was the 2nd axis on that O2 switch point map and why wasn't it already set to 450?

The "CL Mode" axis is defined by table {B4107}. Why that table wasn't already set to 450 is a bit beyond my psychic abilities.


What's with the baro stuff?

I'm not real sure what this table actually does... but the values I've given you are the values I always see in stock tunes, so it must be set like that for a reason. It's an offset multiplier so maybe the MAP sensor performs a bit differently at different altitudes and this table corrects for those differences.

[joecar]

Supercharged111, I'm wondering why your file has so many tables different than stock (I have beaten myself silly trying to tune some files that were too far from stock).

[Supercharged111]

The stuff that Statesman had me change is stuff I haven't ever touched before. I've only disabled TM/TR, played with shift/TCC points, played with emissions (EGR/cats/rear O2 delete) stuff, timing, and now VE/MAF for the most part.

[statesman]

I did a couple RAFPN logs and noticed that the datalogs are indicating a consistent 2+ gm/sec less than the stock tables. The curve is consistent with stock too, nothing erratic. Vacuum leak?

Put these values into {B4403}...
0
0
0
0
0
0
0
1
9
18
26
33
40
48
56
64
72
80
88
96
104
112
119
126
133
137
143
149
155
159
164
167
170
173
175
178
180
183
185
187
190
194
199
205
210
215
220
227
232
238
245
252
260
272
289
300
305
310
310
310
310


Then do another RAFIG/RAFPN. Update your desired airflow table {B4307} and post the new tune file.

[statesman]

Supercharged111, I'm wondering why your file has so many tables different than stock (I have beaten myself silly trying to tune some files that were too far from stock).

Starting with a stock file is a lot easier... but then you never learn anything. :laugh:

[Supercharged111]

Starting with a stock file is a lot easier... but then you never learn anything. :laugh:

I was under the impression I had started as such. I'm beginning to wonder. . .

[joecar]

What year/model/vehicle do you believe your stock tune came from...?

[Supercharged111]

01-02 Chevy Express Van.

[Supercharged111]

And because I'm not struggling enough already, I went and snagged a newer style 4 wire MAF so I don't peg at sea level. I am curious though, since this one has integrated IAT, I can still wire it without said IAT correct? My IAT is downstream of the blower where it belongs.

[statesman]

And because I'm not struggling enough already, I went and snagged a newer style 4 wire MAF so I don't peg at sea level.

The new MAF you 'snagged' won't make any difference... your OS has limits which you can't get around.... 512g/s airflow and 12,000hz raw frequency. You'll still peg exactly the same with your new MAF (that's if you even get there with your low boost). Save yourself some headaches and keep the original MAF in place. The only real way to get around pegging the MAF is to use a larger MAF tube if you hit the frequency limit... it will lower the frequency response and prevent you hitting that limit. Scaling the airflow and fueling will stop you hitting the airflow limit.

[Supercharged111]

I was under the impression that this larger MAF with its larger tube would have that effect right off the bat and then I could rescale until it matched? Then again if I rescale the stock MAF like I was supposed to it may give me the headroom I need. With it pegging artificially soon now I'm getting fairly high MAF readings and it's pulling a lot of timing. Truth be told with the high IATs the truck doesn't seem to want much, if any, more.

[statesman]

Have you actually pulled 512g/s through the MAF yet or is this just a projected problem? Your MAF curve has been beat up pretty bad in the low and mid ranges frequencies but the higher frequencies look very smooth... which makes me think you're just curve fitting those higher values.

[Supercharged111]

Yep, the upper end of what I logged was requiring me to copy/paste with multiply in the .2 range, so I just applied that to all the cells I couldn't hit. Ya know, I have yet to do a calc.vet after actually plugging those fictitious values in. I probably ought to rescale first, then try to clean it all up. I'll admit I SUCK at making things smooth when I try to plug in the new data and blend it. At this point, I feel good about delineating the tip in stumble from any VE/MAF tuning.

[statesman]

I probably ought to rescale first, then try to clean it all up.

Yes, well considering you're using ASA data (which is meant for 4bar fuel pressure) and you're running 3.5bar fuel pressure... you probably should rescale.

[Supercharged111]

Don't forget that I left the IFR geared for 3.5 bar and bank assignments for the GenI. I picked up an aluminum LS motor yesterday so maybe I can steal its 4 bar regulator and toss it on the truck if it fits.

[statesman]

You really like doing things the hard way... don't you?

[Supercharged111]

It sure looks that way. Fixing the fuel to this extent is new to me. There's a lot more to it than the standalone I used to work with and a lot more depends on getting it right. Once I finally get it, I'll repeat this process by 411 swapping my dually. Hopefully after that ordeal I'll become more efficient.

[joecar]

So you're running rail pressure at 3.5 bar...?

with an un-referenced regulator (I don't remember)...?

[Supercharged111]

3.5 bar vacuum referenced. If there's a 4 bar at home then I'll be set.

[dian]

how can op be pegging the maf at 100 kpa? or 109 or even 135?

[Supercharged111]

If I had an engine that breathed more it would be quite feasible. I'm not pegging the MAF right now, and if I remember to rescale tonight I probably won't even at sea level. I just don't see this thing making much, if any, north of 400hp at sea level. NA it'd be in the 300 range so add whatever the "boost" gives me to that.

[statesman]

If you're staying with the 3.5bar regulator, then put 4.52 into your IFR table and put the stock values back into your MAF table... you should be fairly close with that.

[Supercharged111]

So I did a little digging and noticed something interesting: quite a few GM tunes have the MAF pegging early. The Express van did not, but the Camaro did with the early MAF. That makes me think that maybe I did get the MAF halfway right since everything else seems to be right, at least that's what I tell myself.

5.7 Vortec truck with black box PCM.

22620

5.7 Vortec van with 411 PCM and same MAF sensor.

22621

LS1B Camaro, same PCM same MAF.

22622

Unless GM fudged the numbers? Corvette MAF table looks very similar to the Camaro's. As far as the off idle stumble goes, I have some transient tables I got from Alouicious that he ran way back in the day on his Whipple'd 411'd truck so hopefully I can put that one to bed. Too many projects. . .

[Supercharged111]

Alright! That transient fueling table felt like it really did the trick! It's exponentially different from stock, even that far off of what I concocted by globally doubling stock I don't remember how many times over. The only way I could really figure to logically approach this was to log AFR on a VE chart and filter by TPS changing at a certain rate per second but never got around to it. I experimentally maxxed out the table and the truck straight up wouldn't run, so that told me its effects were more far reaching that I currently comprehend. Here's stock.

22623

Here's what I hacked together.

22624

And here's what Al intelligently put together.

22625

Just need to drive it a bit and validate it. On another note, the platform change seems to have allowed my truck to shift all gears in 4lo, but they shift way late/high. Not too sure what to make of that. Here's my tune, any takers?

22626

[statesman]

Here's my tune, any takers?

Any takers for what?

[Supercharged111]

Why the truck shifts so high out of 4lo. Starting to think the platform change didn't do the trick. Transient is night and day better though.

[statesman]

Transmissions really aren't my thing. Hopefully someone else will be able to jump in and help you with it.

[ScarabEpic22]

I looked at your tune, looks like in your part throttle that Normal+Cruise are set the same and Performance+Hot are set the same. The low end is much higher than Normal+Cruise, try lowering these? Have you compared your settings to a stock tune with 4LO, I have a feeling that in 4LO the difference in transfer case ratio may be throwing your shift points off. 4HI is 1:1, 4LO has a multiplier (of what depends on transfer case).

Log it and see what the shift status is when you're in 4LO, plus see when the PCM commands the shift.

[Supercharged111]

As far as I know, 4lo doesn't use any of those tables. There's a wire coming from the transfer case that tells the PCM it's in 4lo and then the PCM references that 4lo ratio to correct the shifts. I did copy a few of the normal tables to cruise for TCC and unlocking, forgot those were bugging me. So back in the day, the trick to to getting 4lo to work in the 411 was to start with a 4x4 S10 tune, then segment swap your engine and trans in, then make that into a COS. I believe this would require me to un-COS and re-COS and since I never use 4lo I've ignored it all these years. It still bugs me as an incomplete aspect of the 411 swap and I really thought the platform selection would be the simple version of what I just described as it was known to be buried in that segment somewhere. Maybe I'll try calling it a GMT800 and see what happens.

[ScarabEpic22]

Ah got it, that makes (some) sense. GM making it hard, sigh. S10 base file as it has the transfer case ratio + 4LO setup. You'd have to start with that as a base, seg swap whatever's needed, then upgrade to a COS.

[Supercharged111]

Aaaaaand I think I just figured out why RAFPN/RAFIG wouldn't produce sane values for me when I did it before: I thought the output was a new value, not a corrective value that I add/subtract to desired airflow. After a reflash the truck does not want to idle until it gets up to temp and has time to learn. Though MAF/VE may be a smidge off since correcting IFR I think doing a quick RAFPN/RAFIG will get me a lot closer to correct now even after I redo those, which will happen after I deal with whatever is causing my O2s to go bonkers.

Nevermind, I lied. Checked old RAFIG stuff and it wanted to cut around 1.5-2 g of airflow. For the sake of science I globally added 1 g/s to the entire desired airflow table and now over 2 different ECT ranges it's trimming 3.5 g/s. Not at all the right way to do it, but good to see if cuts more airflow when I demand more. Just doesn't make sense that it claims to be trimming air when, at times, it struggles to stay idling at all. Mostly after a start, but only occasionally. I suspect it's a mechanical issue and it's been beating my ass for a long time now.

[statesman]

which will happen after I deal with whatever is causing my O2s to go bonkers.

Are you logging your O2 sensors?... do they appear to be switching rich/lean properly?

[Supercharged111]

Are you logging your O2 sensors?... do they appear to be switching rich/lean properly?

The right side, no. It was pretty well locked in at .457 for a good chunk of the time. The left side was reading lean and moved around more than the right side, but indicated a constant lean situation with only around 3.9% positive LTFT. The left sensor was new in September. Since the right one just took a shit, I've decided to figure out this damn misfire that's been absolutely plaguing me since I put the marine cam in. It was creeping on me before the cam, but after was 1000x worse and really impeded on driveability. Then as I drove it more, it got better. Last September the truck sat for about a month while I did a big round of maintenance on it and it ran like absolute garbage for the first 2 weeks after that. It really seems like I may have more than 1 mechanical fault on my hands which makes it very difficult to pinpoint. I had an extra cam sensor laying around so I threw that in tonight with a new rotor as that wasn't looking so hot and will drive a couple days to see if that had any impact whatsoever. I doubt it since I was just driving around with no cam sensor but I have to cross it off my list because all that remains that I can figure is that, a bad crank sensor, wiring, or a wallered out upper bushing in the distributor housing itself. Need to remove, disassemble, and inspect entire distributor to verify that. One thing I noticed that struck me as very odd was that it was idling around 16.5:1, so I disconnected the MAF. It then idled around 12.0:1. I kind of doubt the two disagree with each other that much, so I squirted the MAF with cleaner, put it back in, same thing. Going to dig up my spare MAF tomorrow and do another search for vacuum leaks. My last search for vacuum leaks was fruitless. Here's the log with cam sensor plugged in from a couple nights ago, you can see just how whack the LTFTs are. They were pretty happy way back when I did the calc.vet.

22634

The only code is for bank 2 not switching enough, is that enough to prevent the fuel trims from doing their thing?

The more I think, the more I overthink. How are fuel trims applied? Globally or bank per bank? The engine measures the latter, so to me the latter makes more sense but it's not something I've ever thought about before.

[statesman]

I'll try to look over your log in the next day or two... but in the meantime, check your O2 heater circuit fuse and also check for any damage in the O2 sensor wiring. If the wiring is damaged, it usually happens in those hard to reach places... so check all the wiring thoroughly.

[joecar]

If NBO2 sensor wiring is soldered or spliced or heatshrink sleeved within 6 to 8 inches of the O2 sensor, the the O2 sensor may not be able to read free which it needs as reference.

[Supercharged111]

If NBO2 sensor wiring is soldered or spliced or heatshrink sleeved within 6 to 8 inches of the O2 sensor, the the O2 sensor may not be able to read free which it needs as reference.

Well. . . I did make my own O2 extension years ago. Weather just turned to crap. I think what I'll do first is inspect wiring. If I find nothing, or if I repair what I find to no avail I'll swap left and right O2s to see if the behaviour swaps sides too. I do have a manufactured O2 extension that I bought so I should try that too. Last September when I did the left O2 the right side was still happy.

[Supercharged111]

O2 problems were fixed a couple of months ago. Ended up being a ground lug spot weld failed for the heater circuit. I reattached it and the sensor has been working great since then, even seems to have fixed the switching latency. For the first time in eons the truck doesn't throw any codes, even the ones that don't throw a MIL. Back in September I did something I've wanted to do for a long damn time: I changed over to hydroboost. This is how the truck should have been stock. It wasn't without some 2nd order effects though as the blower inlet severely interfered with the HB accumulator and brake master cylinder. I ground enough off of the blower inlet to eliminate static interference, but I could still hear it when hitting bumps, taking corners, or hitting the brake pedal in reverse down my driveway so I removed the blower and had the inlet modified by a local shop. While I was at it, I also opened up the blower outlet as there was an orifice cast into it so I hogged it out.

https://photos.app.goo.gl/ZpRDvCCzqsrTjfBi8
https://photos.app.goo.gl/wqd4k2sHGLdsvf388

I don't recall what the airflow was when I first put the blower on and my old logs seem to lack timing advance of all things so I don't have a before dyncylair picture. I know NA and pre-CALC.VET the max dyncylair was .64. My MAP from the initial post-modded blower datalog last week showed almost no boost (~95kpa with ~79 being atmospheric), so I removed the blower and tweaked the bypass valve preload. After that, I had to remove it again to further clearance the inlet and zinged through a weld with the hammer and grinder so back to the shop it went. I got it back a few days ago and got it all back together this time with excellent clearance and was showing boost similar to what I saw pre-porting (109 kpa pre porting and 107 kpa then) so I was pretty disappointed there. Tonight all I did was globally add 2 degrees of timing to the high map (and copy/pasted the low map to match) yet somehow mustered 110kpa. The difference that really caught my attention was in airflow. I showed a consistent 1.00 g/cyl right after porting, but tonight am showing a consistent 1.12 which is right at the top of the table for timing. Strange that this table tops out before pegging the MAF, this will definitely cause issues at lower elevations. Were I to skew VE/MAF/IFR I'm guessing I could buy myself some headroom in B5913. The thing that makes my brain hurt is the handoff between it and A0010 since one is MAF referenced and the other MAP. Below is a log from the other night. My big question from all of this is how do you explain the increase in dyncylair with MAP staying pretty much the same?

23125

[Supercharged111]

As for the transient fueling, here's the latest. I was running that B3406 table from Alouicious and his COS3 Whippled truck from 100 years ago and it was WAY better, but with 1 tiny problem. The truck would want to stall after firing on a restart. It did the same thing when I tried maxxing out the table when incremental changes were proving ineffective. The only way I can figure to smartly fix this table is to plot AFR against ECT and MAP, then filter by MAP changing more than 5kpa/sec or something along those lines? Does anyone have a smarter way to address this?

[Supercharged111]

So I tried this last night and the map wasn't as effective as I'd hoped. Looking at the data plotted in 2D, however, was helpful. Another epiphany I had was seeing how the impact factor affected starts. It seemed that it had a much more pronounced effect on starts than tip in fueling. Duh. TP isn't changing on a start, but MAP is, so if changing impact is making starts worse then this must not be the right place to try fixing my tip in issue. As expected, the results of trying to get it right via impact table alone was not the right way to go as the truck would not run this morning. Similar to how I maxxed out the table a couple years ago and the truck wouldn't run, only not as extreme. I trimmed impact back almost to where it was, and started bumping the stomp table. The title is very misleading. On a cold winter morning the truck works well, but then as it starts to warm up just a little it gets bad quick, then once warmed up it's tolerable. The reason for this appears to be that the table is flat up to 20 degrees C, then drops off very quickly after that. For now, I'm slowing this taper with 10% bumps. It's going to take a few cycles to see if this is the right way to do it. To ballpark it, I really don't need to log because I can watch the wideband flick lean any time I poke the gas. That and I forgot to plug in the wideband on my last log. :doh2: I decided logging BEN as ECT vs MAP was a better idea for the colder temps. Nos sure how I'll make a 2D table to capture stomp once I'm closer, I figure I'll just make a single column but what to make it is the question.

[Supercharged111]

I went and horsed around with Stomp and wasn't really seeing any effect. I blew well past the EFILive limits for that table and the truck still never got a rich tip in. I'm going to put it back where it was mear maxxed out and leave it there. I was about at my wit's end, then decided I should try to boost the Impact table in the high MAP areas. I realized the lean spike I was going after was happening at 70+ kpa for the most part, and the Impact table is the only one that breaks it down that way. Again, in parts of the table I blew well past EFILive's limits, but was well within the GM limits. I finally saw an observable improvement, though it wasn't extreme. On the last drive, it looked close, but AFRs were jumping around quite a bit. Since I left the 20, 30, and 40 kpa columns alone, the truck was still happy to restart.

I'm thinking about revisiting IFR and doing another calc.vet to have a good starting point. I also made some changes that I thought I'd made years ago recommended by statesman. I changed the O2 switch point to 450 mv and the min duty cycle to .9, so I need to do a log before thinking about changing the evap time for a rich lift condition, though that's much less of a priority than this other stuff. I thought I'd been running semi-open loop all this time and it turns out I wasn't, I'd only disabled idle STFTs but closed loop was free to come on. That explains why it never seemed to work. But these jumpy AFRs really have my attention right now and I think that's where I need to focus. Right or wrong, tip in is much improved while retaining a truck that's happy to restart. I've also gotten the truck in a better mechanical state than it's been in years. Since it doesn't feel like a piece of crap anymore, I'd like to finish addressing this tune that's fallen to the wayside for far too long. After the calc.vet I need to type up some notes so I can modify the IAT VE multiplier. I've seen significant AFR swings WRT IAT. Every time I reread the COS tutorial I remember how half baked this thing really is.

So after all that rambling, I have 3 big questions: 1, what happens when you exceed an EFILive limit but stay within the manufacturer's limits? 2, my values do not seem sane compared to stock. Is that unusual on a modified vehicle? The injector trajectory is significantly different that stock. That, plus the baby blower I have on there really screw with the tip in fueling. 3, is that tip in spike transient related or should I be looking elsewhere seeing how far I've gone to get where I am with the transient fueling? Below is the current tune and the latest log, both subject to change depending on how motivated I am.

23132
23133

18 tunes and 19 logs since last night. For a lot of them I knew they were junk the instant I cracked the gas and it took longer to get back to a parking spot than it did to flash a new attempt.

To be clear, I did a restart at 130 degrees and the truck drove great up to operating temp of 180. It's the wideband jumping around that drives me nuts at this point. Plus I have to see what it does from a 30ish degree cold start tomorrow.

[statesman]

I've also gotten the truck in a better mechanical state than it's been in years.

That's not good enough. Being better than crap doesn't mean it's good.... and you can't tune an engine until it's good. Smoke test your intake, put in a fresh set of spark plugs, check your spark wires and replace if not in great condition, check your air filter and replace if needed, replace your fuel filter and check your fuel rail pressure (if you can). Then come back and tell me you know for sure that the engine is mechanically good.



18 tunes and 19 logs since last night. For a lot of them I knew they were junk the instant I cracked the gas and it took longer to get back to a parking spot than it did to flash a new attempt.

And what exactly did you learn from that? Spitballing tunes is fine as long as you're keeping track of the changes you make and what effect each change has on how the engine performs. A lot of people learn to tune that way, but you have to try to learn from the changes and not just make random changes which are discarded and forgotten about if they don't work.

[Supercharged111]

A lot of the improvements aren't tune related, but my O2 sensors seem to both be working and I discovered a couple vacuum leaks, one was the evap vent valve of all things. At this point, I don't think there are any issues being tuned around. I sprayed carb cleaner at the intake manifold and no change. Filters are relatively new, but I should check fuel pressure. This morning's drive revealed that the values I felt weren't sane aren't, those were the cold values. The factory has a large disparity between cold and warm impact factor and this trial and error tweaking of mine seems to imply the engine wants less of a disparity. For instance, if it gets cold enough, tip in works fine. As the engine warms up, response gets worse, then fully warmed up there is only a slight hesitation. The 80 C values in there now have excellent response but the data shows a lean condition at tip in. Not sure what to take from that. I've kept track of my changes from stock, to a globally bumped map, to the concoction I have now that's getting scrapped. I've searched and searched for ways to methodically alter these tables and found nothing. I even tried one of my own and feel it was worthless. I can see a lean dip on the 2D graph over time, but the way the AFRs jump around I feel I need to address that first. If you have a way to methodically alter these tables I'm all ears.

[statesman]

Are you using the latest version of EFILive?

[Supercharged111]

Ha! Yes I am! I finally got off my ass on that one.

[statesman]

Post the stock tune you've been using as your reference.

[Supercharged111]

This is what I pulled off the PCM when I first got it. I bought this 411 from Justin at Blackbear 100 years ago and had him flash an 01-02 Chevy Express tune (12212156 OS) on there with COS5 plopped on top of that. Bear in mind this would have had a GenI 5.7 like my truck and not an LS based engine. I wonder if that's why you noticed so many things that seemed wrong.

23134

[statesman]

Get yourself a new set of spark plugs... yours will most likely be fouled from running rich all the time. Check your fuel rail pressure. I'll make some changes to your tune file and I'll try to get that posted here for you by tomorrow. Don't fit the new spark plugs until I give you an updated tune file to run.

[Supercharged111]

Thanks! I'm going to check fuel pressure at the parts store to have that as well.

[Supercharged111]

Alright, fuel pressure was 53psi with key on engine off. I flashed the converter in 4WD to prevent wheelspin and it looked like the gauge went up around 57psi, I'll reverify in the daylight tomorrow when I verify that fuel pressure holds to redline, something I should have done as soon as I put the blower on. Now that 53psi KOEO also occurs around 78kpa, so does that mean at sea level this would be 56psi? Just wondering what to put into the fuel injector calculator to reverify my IFR numbers. Also attached is the latest tune. This impact map seemed to be the best blend of starting, cold, and warm running. Not perfect, but a step in the right direction.

23135

I started with a stock map that was bumped about 32%, dumped in the values I came up with at 80C (which I still may recurve), and did a linear fill from -40 to 80, then 80-140. Rudimentary, but more in line with what the motor wants. And I'm not exceeding any limits this time.

[statesman]

Alright... I've made some changes. It's a rough cut, so don't expect big things from it. Install your new spark plugs and do a log which includes the cold start. Then log a drive. Then when you end the log and shut down the engine, let it stand for about 10mins before restarting... then do a new log which includes the hot restart. Then post your logs and let me know how the changes felt.

[Supercharged111]

Alright... I've made some changes. It's a rough cut, so don't expect big things from it. Install your new spark plugs and do a log which includes the cold start. Then log a drive. Then when you end the log and shut down the engine, let it stand for about 10mins before restarting... then do a new log which includes the hot restart. Then post your logs and let me know how the changes felt.

Do you need any more/less PIDs than what I have in the log above?

[Supercharged111]

I was able to verify in the daylight today that my fuel pressure holds all the way to a WOT shift so it's good to know for sure I'm not tuning around that. I figured the pump was enough especially with me up at altitude. I should recheck at a lower altitude too. Below are the 2 logs, one of the cold start and drive and the other from the hot restart. For some reason, the only way to get EFILive to see the external WB, I have to go into the handheld every time I hook it up and select display O2. You'll notice in the hot restart there was no AFR so I shut the truck off, got the wideband displaying, then did another restart. I also changed cylinder volume back to 713 from 708 as it is a GenI 350, not a GenIII LS1. It's misleading me putting the thread in here years ago, but the truck does run off of a GenIII PCM.


23137

23138

[statesman]

Okay, here's the next rough cut. This time I've set it to run MAF only. The stumble may be slightly worse on this calibration, but this is just to get a baseline so suck it up. And yeah, I gave you back the cc's I took off you in the last update. I've reduced your IFR, so you will be a little bit richer... but that's okay as you're a little bit lean now. I also changed your PE entry conditions to something a little bit safer than what you had. I will need you to log specific PIDs this time...

Wideband
Narrowbands
STFT (both banks)
LTFT (both banks)
GM.VOLTS
MAP
MAF airflow
MAF frequency
RPM
ECT
IAT
TP
VSS
Commanded AFR

Don't log anything extra, just what I asked for.

I'll need a log with about 12,000 frames of driving on warm engine. Try to get a good mix of driving.... 12,000 frames just crusing down the highway isn't very helpful... and remember to make sure your wideband engages.

This calibration has DFCO disabled, so if you want DFCO then re-enable it after you do the log.

[Supercharged111]

Actually the impact table changes made a night and day difference in the truck. That still doesn't fix the IFR/MAF/VE (or idle airflow) that we both know was off so I'm holding off on any more of that until after this is done. Here's the log, it came in around 18,000 frames. I had to go north of 12,000 because of how much of the drive was downhill. Here you're either going up one or down one. The truck actually drove fine MAF only, so I'll just leave that on for now. Thanks for all your help and collective patience with me. Going back through this thread is painful even for me at times.

23140

[statesman]

You're dangerously lean in boost. I'm not going to help you damage your engine. If you want my continued help, then you'll have to fit a boost referenced fuel pressure regulator and make a new log.

[Supercharged111]

That is a vacuum/boost referenced regulator. It idles at 48, has 53 when you turn the key on and the pump primes, and holds 57 under boost. I figured it was lean because it was MAF only and that that pointed to a MAF calibration issue.

[Supercharged111]

OK something's not right. The couple times I jumped on it on the way into work the PE AFRs were in the 10s which is what I would expect to see in this cold air. Since I brought the laptop in with me, I'll snag a quick log on my way to my neighbor's shop after work.

[statesman]

That is a vacuum/boost referenced regulator. It idles at 48, has 53 when you turn the key on and the pump primes, and holds 57 under boost. I figured it was lean because it was MAF only and that that pointed to a MAF calibration issue.

That's weird. When I looked at your log and I saw you going lean in boost I thought your fuel pressure regulator is not referencing boost. I'll take a closer look at your log now and see if I can figure out what's going on.

[Supercharged111]

It gets better. My AFRs under boost seem to vary from one mile to the next. I'll jump into last night's log too and see if I can spot a pattern. I will say that idle and cruise AFRs seem more consistent MAF only. I did learn some stuff about the transient last night and have to wonder if that corrupted some of the data I got last time I tried fixing VE and MAF. Among the other stuff I'd gotten wrong.

[statesman]

Okay, I've had a proper look over your log and I've determined that your wideband is toast.

[Supercharged111]

No shit? Narrowbands look good? Just need a new WB sensor? Truth be told it's 10 years old.

[statesman]

No shit!.... when you're in boost and your wideband is reporting in the 13's and once in the 15's, your narrowbands are always in the 900's. Looks like you're not running lean in boost after all.

[Supercharged111]

Ugh, I was wracking my brain because the MAF was reporting consistently but the AFRs disagreed and I didn't know who to trust. I'll let you know when I have a new sensor on the truck. Hopefully it only takes a few days to show.

[statesman]

I can't use the last log... you'll have to run another log for me when you've got the new WB sensor. 18,000 frames was good. I need minimum 12,000 but anything over that is alright.

[Supercharged111]

I can easily redo the 18k route with a new sensor. How were the transients, gear changes, all that? I didn't choose a route where I could really load it hard WOT in 3rd gear, rather something that better resembled my my daily routine but pulled the shifter back to 2nd and 3rd a lot to better populate the mid range cells.

[statesman]

The more MAF cells you can hit, the better. Try to get one low gear pull to at least 5000rpm if you can. Load isn't important when you're at high rpm, it's the airflow that's important... and at 5000rpm with the throttle open you get lots of airflow so you hit those higher MAF cells. I only really need one of those, so the rest of the log can be your normal driving... just try to get a couple of harder decels in there as well.

[Supercharged111]

Can do. I'm having trouble figuring out how to set up a 2D map to monitor cell counts in a way that mimics the MAF calibration map in the tune. How do I get it to read in a single column? It's plenty easy to use the tool to build the 85 rows.

[statesman]

I haven't used the scantool maps in quite a while, but I just set one up for old times sake (and to remind myself how to do it). You just have to select "Open Map" and then choose any of the ones labelled "{B5001} MAF Calibration" and it sets it all up for you.

[Supercharged111]

Well the sensor didn't show up today like I had hoped. Monday it is, I can't wait to have this buttoned up. It's already way better than before, start and restarts and tip in are solid and will only require tweaking after the MAF and VE get to where they need to be.

[Supercharged111]

So the sensor I screwed in back in 2009 refuses to come out in 2020. Not even vise grips and a cheater pipe would touch it after 3 rounds with the propane torch. Hopefully I can score a new O2 bung tomorrow and if so I'll just shove the new sensor in there and go for a log. Typically the vise grips and pipe with a torch are the end all to a stuck O2. I've never been boondoggled so hard by a damned O2.

[statesman]

So the sensor I screwed in back in 2009 refuses to come out in 2020. Not even vise grips and a cheater pipe would touch it after 3 rounds with the propane torch.

I said it in August 2017...


You really like doing things the hard way... don't you?

And I'll say it again now....

You really like doing things the hard way... don't you? :laugh:

Try spraying on some penetrating oil... and then wait at least 24hrs for it to 'penetrate'. Then crack it cold, not hot.

[Supercharged111]

Unfortunately I have to drive this to work every day so 24 hours won't happen soon, but I'll try a couple overnighters leading up to the weekend. The current sensor is in an excellent location and I'd prefer to keep it there, but zapping in a new bung may end up being easier than fighting with that sensor if none of the tricks end up working. I've never had great luck with penetrating oil, the torch always seems to be the thing that saves my ass. I may be able to borrow a real torch this weekend so may have that at my disposal, I have a solid axle that needs a little negative camber and made a deal on some acetylene torch usage but that agreement was initially at my neighbor's shop.

[Supercharged111]

Shortly after my last post, I encountered some aging/injury issues that precluded the use of my right hand/arm to the extent necessary to muscle this O2 sensor out. I just got the sensor swapped out last night and, MAF only, I'm pig rich at WOT so will get a log up here real soon along with my own hack at the MAF to get an idea of any deltas between where I think things should be and where they really ought to be. With the transient fueling where it is now even MAF only the truck's driveability is better than it has been in years.

[statesman]

Don't beat up your MAF curve... just get a log for me using the last tune file I gave you.

[Supercharged111]

I have no intent of flashing whatever I come up with, just want to compare it to what you come up with.

[Supercharged111]

Here's what I got tonight, after 13,000 it's basically just coasting home downhill. I got bored at a red light and killed it there. Let me know if this flies, I'm going to read over autoMAF again as I'm guessing that's what we're doing now?

23209

[statesman]

I got bored at a red light and killed it there.

That's not a good look. If you couldn't give a rats about getting me a decent log, then why should I give a rats about helping you?



Let me know if this flies

NO, it doesn't fly. It's a useless log... which is not surprising considering your attitude towards tuning.

Take a time out and think about if you really want my help. If you decide that you do want my help, then find some motivation to provide me a decent log... and then do another log, but this time make sure that you're actually logging the wideband.

If you decide that you really couldn't give a rats, then that's alright... you've already got a better tune than what you had before.



I'm going to read over autoMAF again as I'm guessing that's what we're doing now?

Wrong!... if we were only doing an autoMAF, I'd just tell you to do it yourself.

[Supercharged111]

?? The first 13,000 frames are what you asked me for. Transients, a couple WOT pulls, and a couple decel events. I'm lost. I killed it at a red light as opposed to waiting until I got home because it wasn't worth keeping it going at that point.

[statesman]

I'm lost.

13,000 frames of the wideband reading ZERO.... what am I supposed to do with that?

[Supercharged111]

Aww fuck. I wasn't even looking at that, just the gauge. Every time I plug the thing in I have to go into the handheld and view the wideband reading there to get it to show on the software, forgot all about that. And I couldn't find the autoMAF tutorial in the tutorial section either, not that it matters.

[statesman]

Have a look at the last log you posted. When you go into decel, you get a huge split in your fuel trims. Something is wrong with one (or more) of your injectors.

[Supercharged111]

Well crap. Truck runs fine, so if an injector is to blame I'm not too sure how to pinpoint it. They were cleaned and flow tested before I installed them back in 2017 and I do have a couple of junkyard fresh uncleaned spares on hand to troubleshoot with. I know I can shut down individual injectors as the engine is running, but is there a cylinder balance test I can do too? I know the EECIV Fords were capable of this. I imagine I'd just be wasting my time with any more logs in the meantime. I can see where it repeatedly and intermittently leans out the left bank on decel.

[statesman]

Since they were cleaned and flow tested, I'd suggest you do a little test to see if it one of the injectors going bad. Swap your injectors bank to bank and see if the lean condition follows the injectors. If the left bank continues to be lean, then you'll need to look for air leaks... both in the intake and the exhaust.

[Supercharged111]

I'm thinking I'll swap O2 sensors first and make sure I'm not chasing my tail at the rate this thing has eaten them. I blame the tune for that though. Once narrowed to a bank the million dollar question is how to narrow it down from there.

[joecar]

Do you have a way of measuring temperature on each of the exhaust header tubes...?

If one cylinder is leaner, it will produce a different temperature (thermal imaging cameras have come down in price, I'm thinking of getting one).

( this may have been covered: are you sure there are no intake port leaks...? )

[Supercharged111]

I sprayed carb cleaner all around the intake a while ago and couldn't find any leaks that way. I don't know how to check if it's sucking from the crankcase. I have a Raytek Raynger ST2 that I got from my dad. Could try that at an idle tomorrow all warmed up. I just pointed it at my space heater and it read north of 400 degrees F, so hopefully it has the range for exhaust heat. My POS Harbor Freight handheld did not.

[Supercharged111]

I did measure EGTs when I got home from work before I shut the motor down. I had to do it twice because they dropped a bit over the course of checking them, but stabilized at the values below.

380s 380s
480s 480s
480s 480s
380s 480s

I redid the cold ones and a few warm ones a couple of times to make sure there wasn't a variable like engine fan affecting things, but those values were very consistent. I swapped O2s last night and will take a log tonight and will be watching decel. I'm expecting that to remain the same. Next will be swapping injectors bank to bank and noting where they were and where they're going to see if the injectors are to blame.

[Supercharged111]

Well the problem definitely stayed on the left bank, I went for a quick log that ended up around 8000 frames, this time with WBO2 reading. Check frame 3700 for my first of the 2 big decels I performed, seems to be a 5+ second delay before things go haywire just like in the last log.

23216

So I think the next logical step is to swap injectors bank to bank like you said, but track where they go and put the injectors from the cold cylinders into corresponding warm cylinders on the opposite bank to make sure there isn't something that follows there too. It seems to do it on all decel events and not just the big ones, more like a matter of throttle off time that seems to be driving it as it happens at 1500 RPM lifts too. I think I'll unplug the PCM connectors, shoot all the injectors, and jiggle the harness around to make sure there isn't something wrong in there as well.

[Highlander]

that VE Table looks very very weird.... kind of like if the columns are rows and rows columns....
you have your dynamics B0120 set to 4000 so the ECM is using both... wrong maf (too much fuel) coming off from a stop and you will have the stumble... is there a log of the stumble?

(im on the first page) looking at the tune posted in the first page still.... Your PE enable B3616 is set to 70%. at a lot lower pedal you will get into boost, might be part of the cause for the stumble
Injector data in general just looks wrong....

If this has all been addressed, i apologize.

[Supercharged111]

Stumble is gone now, just trying to clean up the rest of the tune. Truck actually runs the best it has since the blower went on. Stuff is obviously still wrong though, my injector data must have been off and then that screwed me on my calc.vet. Among other things. I should also point out, especially to Statesman, that my wideband sensor is on the driver's bank.

[Supercharged111]

I need to upload my logs when I get home. I tried taking a different route, but far as I can see my results were inconclusive. While driving, I disabled 1, 3, 5, and 7 one at a time to see if the jagged logs went away. They did for a few seconds, then they came back. I tried it on 2, 4, 6, 8 and got the same result.

[statesman]

You need to do the bank to bank injector swap. Until you do that, you won't even know for sure if it's definitely an injector issue you're chasing.

[Supercharged111]

I did already, my most recent logs are post sensor swap. Kid's on the laptop now.

[statesman]

So I think the next logical step is to swap injectors bank to bank like you said, but track where they go and put the injectors from the cold cylinders into corresponding warm cylinders on the opposite bank to make sure there isn't something that follows there too.

Did you do that?

[Supercharged111]

Nope, took the logs and left it at that for the time being. I wanted to make sure there wasn't something there I'm not seeing before moving the injectors around. Of course I managed to forget to post those yesterday once the laptop opened back up. I've been a bit distracted the last few days getting a new rear main seal into the dually and all the "since I'm in there" stuff that goes with that. Headed straight to the dually today after work, I made myself a reminder to post the logs tonight when I do get home.

[Supercharged111]

OK, here are a couple of logs I grabbed. This one is while driving and I shut off 1, 3, 5, 7 one at a time. On this one 1, 3, 5 appear to react similarly and disabling #7 is where it gets jagged. This was the cold hole when I lasered the header. Turns out my memory failed me and the front 2 were much less significantly cooler than I initially reported. I disabled 1 around frame 1188, 3 around 1479, 5 around 2125, and 7 around 2663. It looks like the WB is tracking with the NB on the left bank, but the logs are a bit deceiving with the way the NB spikes are so much larger than the WB.

23225

Next is one I took idling in the O'Reilly's parking lot and it's the right bank.

23226

I took this for 2 reasons: 1 I saw the right bank go nuts when I disabled a cylinder on the left bank and 2, I noticed it going weird on its own. The truck needs to be driven a good 20 minutes before this starts showing up. On the left bank, because TP and load are constant, I can see that I disable cylinder 2 from frames 126-343, cyl 4 from 505-753, cyl 6 from 885-1020, and cyl 8 from 1253-1463. I'm curious to see if anything jumps out at you from these or if I'm misinterpreting something that's right in front of me.

[joecar]

Which cylinder had the least effect (or lack of) when you disabled it...?

[Supercharged111]

Which cylinder had the least effect (or lack of) when you disabled it...?

From the driver's seat they all felt the same, so I was hoping it'd shake out in the logs but it doesn't appear to have so I'll need to swap injectors and remeasure EGT I think. Blower has to come off for that, so yay.