Large Positive STFT after decel

[32vape]

2017 Camaro SS 6 spd, Cam, Headers, Roto-Fab, MSD, Flex Fuel. Anyone have any tricks to eliminate the big jump in STFT after a dfco or cfco ends and the injectors turn back on? I tried turning off the idle LTFT cells, which fixes it on some of the 5th gen stuff, and it still does it. It's causing the idle to drop and flare.

13 seconds into the log is an idle flare from it
36 seconds in is just dropping down and hitting idle no problem without the 27% stft.

[Haans249]

You can’t turn it off. Many have tried and its likely a setting we don't have access to. Other tuners have said that this happens when rear O2’s are disabled so is probably some fueling catalytic safety feature. You’ll just either have to live with it or disable dfco all together. I chose tk just live with it as it only lasts a partial second.

[GapRider]

You can?t turn it off. Many have tried and its likely a setting we don't have access to. Other tuners have said that this happens when rear O2?s are disabled so is probably some fueling catalytic safety feature. You?ll just either have to live with it or disable dfco all together. I chose tk just live with it as it only lasts a partial second.

glad to see your response.
Sounds really similar after my cam install with headers. 34.75% STFT shown on Freeze frames for CELs but otherwise rich cuz my airmodel needs adjusting. No rear O2s for me either. I will try disabling DFCO.

Since you say it only lasts a partial second, I'm guessing no CEL? I am probably not understanding something, I'm just learning

[TriPinTaZ]

This can be caused by the Virtual Torque model being off and not seeing enough delta change in torque to kick out of DFCO properly. Not saying this is the issue but I have seen this.

[GapRider]

This can be caused by the Virtual Torque model being off and not seeing enough delta change in torque to kick out of DFCO properly. Not saying this is the issue but I have seen this.

Man, great to know, thank you!
I'm pretty sure both my Virtual Torque and Virtual Volumetric Efficiency models are off (both are just calculated from stock based on diff in dyno numbers for stock and with cam/headers - I haven't adjusted them based on EQ error yet).
But I have played with the DFCO torque delta stuff before so thank you, I'm going to look at that again.

[GapRider]

Well I've changed all DFCO, TQ acc/dec stuff back to the tune I ran when I just had headers but still get the CEL for P015B,D (freeze frames always show 34.75 STFT).
I finally got the situation on a log (no CEL yet).
I want to change something with torque model but honestly don't know how to tell what I should do based on the logs I have.

I've added a bunch of torques to log channels (not on the attached log) - which torques should I be looking at?

This is what i saw on the log about 6:38
TMA High STFT.png

C1v9A3e-drive around WOT.hpl
C1v9A3e-increased off idle spark.hpt

[TriPinTaZ]

Definitely some issues with that tune. It looks like some of it was someone guessing at things where they don't know what the implications/effects are.

[GapRider]

Definitely some issues with that tune. It looks like some of it was someone guessing at things where they don't know what the implications/effects are.

That someone is me . I?m learning and really appreciate the insight. The tune is based heavily on a post that compiled changes made for cam and headers, really helpful. I?ve done other things trial and error because I just don?t know, some of that stuff could still be in there.
For VVE and VT, I calculated the difference from stock based on dyno curves showing stock and with cam/hdrs. I haven?t tweaked either outside of that.
All that said, I?m thinking it would help me if I knew what torques I should log to look for differences between commanded and reported?
Any nudge in the right direction is really appreciated

[GHuggins]

You need to dial your VE in along with your torque and density tables. You also had cat test still running even though you don't have cats, so possibly some of the problem?

This is one of the ones I'm working on dialing in the VE and density right now on along with your tune with the cat test off - do a write entire so it'll take. The log posted is a cammed with headers and cat delete along with cai modded truck.

[GapRider]

You need to dial your VE in along with your torque and density tables. You also had cat test still running even though you don't have cats, so possibly some of the problem?

This is one of the ones I'm working on dialing in the VE and density right now on along with your tune with the cat test off - do a write entire so it'll take. The log posted is a cammed with headers and cat delete along with cai modded truck.

thanks so much GHuggins, I'm going to try it and I've got some reading to do now before I post back. Again, thanks so much.

BTW, I could open the tune but not the log (please update your software) which I'm scared to do because I've seen stuff that makes me worry I'll lose access to some parameters

[GHuggins]

This is a screenshot of the most negative trim right after 0 pedal was achieved for dfco. I'm still tuning in his VE and density table. Even though it's not technically correct I'm working both models at the same time to get the most smooth airflow models and torque readouts that I can.

[GapRider]

This is a screenshot of the most negative trim right after 0 pedal was achieved for dfco. I'm still tuning in his VE and density table. Even though it's not technically correct I'm working both models at the same time to get the most smooth airflow models and torque readouts that I can.

man there's so much there but it helps me to see it. I still don't know what you mean when you say 'density' , I just know VVE (which is speed density I thought) and VT (torque).
I did read in another post that the guy says he does similar as you say where both models are adjusted simultaneously - he cared more about the result it seemed.

[GapRider]

BTW, I had a hard time doing a full write because the car kept going to sleep before the tune had fully written. It finally worked when I hit the 'start' button on the car as soon as it went to sleep and the tune finished writing successfully. Is there something else you recommend to accomplish a full write?

[GHuggins]

It shouldn't go to sleep as long as you have it in full power on mode, but who knows with the way they change things. Density is fuel density. Not the way you're supposed to do it, but it seems to be working out really well on the ones I've done it on so far. I do the base dial in with the engine idling and then fine tune and shape it on the drive logs. Very smooth air models on top of keeping fueling mostly the same. The log I posted is only about halfway done. Hoping to have it balanced out and move on to dialing in the different cam positions here shortly, but this one is being a pita - doesn't want to balance out like I want it to

[TriPinTaZ]

That someone is me . I?m learning and really appreciate the insight. The tune is based heavily on a post that compiled changes made for cam and headers, really helpful. I?ve done other things trial and error because I just don?t know, some of that stuff could still be in there.
For VVE and VT, I calculated the difference from stock based on dyno curves showing stock and with cam/hdrs. I haven?t tweaked either outside of that.
All that said, I?m thinking it would help me if I knew what torques I should log to look for differences between commanded and reported?
Any nudge in the right direction is really appreciated

I think your Virtual Torque changes is the reason you're not going in and out of DFCO properly and the STFT's are acting up. I've run into this before. Try putting your torque model back to stock as a test. When you change VVE you change the torque as it is used in the calculations. To be honest, I have had better luck not messing with the Torque Model a ton and only change it where necessary rather than changing it right away JUST because the car has a cam. Often you jut need to lower it a bit at lower RPM's and can often leave it alone up top if you do the VVE model correctly. Then if the throttle is closing or the ECU is pulling timing, then you may need to alter the torque model at WOT. You can also increase it a little bit and reshape it for 0 degrees of movement if you want more accurate torque values. When you mess with the torque model and the off throttle airmass rows (0.10-0.20 area)get changed too much, it will mess with DFCO.

Anyhow, that is where I would look first. There are a few things that can affect this so my answer may not be the solution. But trial and error to find what changes you've made that are causing this is a forward step in the learning process. We've all been through it so don't give up.

As for Torque PID's. Email me and I can send you a config file that has what's important to log.

[GHuggins]

^^^ It can mess with throttle and spark both as well. This is why I like to balance out the models to one another and shape them a particular way now.

[TriPinTaZ]

^^^ It can mess with throttle and spark both as well. This is why I like to balance out the models to one another and shape them a particular way now.

Absolutely! We have the luxury of experience and working the torque models into our logic. I think when newer tuners start messing with Gen V stuff they mistakenly change Virtual Torque because they think they should rather than changing to to work with the rest of the changes.

[GHuggins]

I still shape it for the changes including idle areas - in fact I'll adjust it as I go to keep things in line - I just try to wait until driving reports start coming in for off idle areas and mostly after the VE and everything else is pretty well dialed in. I've studied stock dual overhead cam models to see what having a lot of overlap around idle and so on does to them and after talking quite a bit with one of the guys who originally set it up for GM and knowing what parts of the torque models do what I started doing the map side primarily different as it's what's primarily affected by mechanical changes - I just do minor tweaks to the MAF side. I do leave a lot of torque settings mostly stock elsewhere in the tune, but picky customers and wanting things like AC engagement while shifting in and out of gear to go unnoticed has forced me to relearn how to shape things.

Now, I don't make "global" changes to the tables as a whole, but instead focus on key tables then let the calculate coefficients function make the tables work to one another. I can tell you the 40 spark and -10 spark tables are pretty much tied to one another and the positive torque tables just need to keep rising. The OS seems to always revert to the table with the highest torque value off idle, so if that's 20 degrees on a manual car for instance, you're going to wind up with it bouncing back and forth from the 30's to 20's degrees of timing on take offs - feels like bucking - you have to mold it to go to the highest possible so it'll run off of the main timing table. DD plays into this too. If it's too low, it'll cut timing at lower throttle inputs and then cut throttle with higher. Again another example of things needing to work together.

I will also say you can REALLY screw the pooch with what appears to be "slight" torque model adjustments such as the "off idle" ones. You can wind up with run away conditions if it's not high enough or if you take too much out in the wrong area. Tuning in the air models on their own won't correct the torque models and when a big cam is involved it can really require you to rework things, especially if it's idling around 70 or so kpa map now vs the 30 it was at stock. Manual cars and take off issues really benefit from torque model shaping when big cams are in the picture.

BUT it's like you said, you really need to tweak most of the off idle stuff after everything else is mostly tuned in and then touch up other areas again after the fact. As of late I've started using the density, VE and torque models in unison with one another to get fueling, torque and control as good as I can. Not politically correct, but it has REALLY worked good on the last 4 I've done it on. I'm getting axle torque commands in double digits, 0 pedal torque to stay -10 to +10 (mostly 0 to +10) with no rpm dips or anything like that at idle with auto trucks and cars. Pretty happy with it. Those reading are with MAF failed or MAF only too. Key that each model agree with one another.

You just have to watch out with the torque changes on their own. You can cause as many problems as you correct It all needs to work together.

AND then that's like these cat test - cat test and light off routines like these run in the OE factory vehicles. They shift rich after decel to light the cats back off or I've even had one person recently tell me that her's would surge on a very rare occasion after a decel where the cats were running self test routines. A lot of OS routines and things to have to work with on top of everything else...

[2023_ZL1-AUTO]

So, related to DFCO and tuning it off completely. Is it as simple as changing the [Disable ECT Hyst] from the stock 41? F value to something like 450? F ??

dfco.JPG

[GHuggins]

Your rumble problem as posted in your other thread isn't caused by it going into dfco, but instead caused by it taking too long to go into dfco. Leaving dfco off and having the timing up will work around it some, but it won't fix it. In the ZL1's there is a "rumble" setting from the factory where it delays dfco. Problem though - not in hp tuners. I have someone change stuff like this for me when they're able. Otherwise you can completely disable it and do like you're doing. It'll just use more fuel and still pop along a little.