Twin Turbo LS3 Surging and Fueling Problems

[GHuggins]

Hi everyone. I've got an unusual problem that I just can't get a hold on and I know it's something extremely simple I'm just completely over looking. I've already searched on here and other places for answers, but just can't find anything to fix my problem. To begin with the vehicle in question is a 2010 ls3 corvette that has had an ls2 APS twin turbo kit installed onto it. I know this may not be preferred, but this was what was sold to us at the time. The kit came with an already retuned ecm that was supposed to be easy to build off of, only the tuner that did the tuning, well lets just say I'm not sure if he even knew what he was doing and leave it at that... The kit came to us with the LS2 injectors, wrong for the LS3 I know, so the shop it was purchased from sent us the correct shorter 60lb/hr injectors. Ever since day one this thing has surged. It was given to me to me to fix and it just seems that I just keep beating my head against the wall over this thing. Once I get it fixed, I'll fire it up cold and it's right back to where it was before making any attempts at fixing it. Here's what I've gotten it down to or atleast seen happening. I can watch my o2's switching voltage (going rich and lean as they should for the cats) and the engine revs up and down with them. My a/f isn't even where I would like to see it. It too switches from 14.1 to 15.1 with an average of 14.6, which is where I would like to see it, but I don't like it (free ranging) that much. I originally just thought it was the fact that the tune was originally set up for the taller 60lb/hr injectors, so I entered (using Greg's injector data) the shorter 60lb/hr injector data. I even installed a correct map sensor out of the supercharger of a zl1 for the boost (which the original kit did not come with). Initially this cured all of the surging, until I put the car into speed density, then every bit of it came back just as bad as it was before entering the correct injector data and whether it's rich, lean, perfect, nothing matters - it still surges.

The best things I have done for it, which kind of goes against the usual ways of thinking, are to really advance the ignition timing in the idle area, increase my ignition over and underspeed settings, lower my idle to 650 (this was so the turbo's don't free spoil as easily) and finally the single best thing I've done is turn my transient fueling "way" down. In fact if I remember correctly I think I multiplied the mid section by .05 then zeroed out the really lower vacuum numbers. The transient fueling was without a doubt the best thing I have done, but this also made it run richer when first starting the engine as you will also notice in the datalog. Now instead of idling up and down in a 50 to 300 rpm area, it idles in a 25ish rpm area. In fact watching the tach in the car, you can't tell its changing at all, but the tach on the dyno shows it and the exhaust notes changing are the biggest give aways. There has got to be an airflow number that I'm just not aware of being able to adjust on this that will compensate for the added airflow of the turbo's at idle and fix all of my problems thus allowing me to get my questionable fueling back in line. Any thoughts on this? Also let me say ahead of time any and all help is greatly appreciated. I've attached a datalog demonstrating the surging higher in the rpm's and if you bring up a tach display with the 02's and long term fuel trims you should be able to see what I'm talking about at idle. The jeff tune is the original tune from original tuning shop with the closed loop turbo vette being the one I'm currently working off of. If you see anything at all out of place or know what I need to adjust for the added turbo's please don't hesitate to let me know.

I know the tune is rather hacked up with things I've tried with no success. I was going to fix these before I went back to correcting everything right after getting this surge problem sorted out.

To date, I've tried adjusting injector slopes, stoich, MAF and VE tables, O2 settings (for ease of tuning and to try to force it to run closer to a 14.5 af ratio-will adjust switching back to stock once done tuning), open loop EQ's (because once I got it gone hot, it was back cold-this just made things worse), injector timing (did mainly to get rid of fuel igniting in exhaust instead of combustion chamber where it should-a found side effect of extra airflow from the turbos)(also returned to stock just to make sure this wasn't causing side affects-actually got worse at stock), proportional idle, adaptive idle, integral idle, iat's (did this because I noticed it also got much worse with the colder denser air-was 38 degrees fareignheit out), then a couple I'm sure I forgot that you will notice.

Sorry for being such a long winded post, but I figured the more info, hopefully the easier it would be to see.

Thanks,
Greg

[GHuggins]

Just looked back over the datalog. Forgot to mention the lambda sensor wasn't hooked up, so please ignore that part. I've currently got the dyno's a/f sensor in the right bank of the exhaust and am just using it to do my tuning with.

Thanks Again,
Greg

[PST]

Stock engine, just the turbo kit, injectors, and map sensor? I can't look at the tune currently, on my phone.

I'd start from scratch with a stock file. Plug in the injector and map data. Stick it in open loop and VE tune using a wideband and graphing it against commanded afr. Sounds to me like you were trying to get fueling tuned either with fuel trim data or trying to MAF tune using wideband data but still letting fuel trims make corrections. Either put it in open loop and use afrs, or leave it in closed loop and use fuel trims. Otherwise, you'll chase your tail all day.

[GHuggins]

Hi, yes I have found that your better off tuning in closed loop via fuel trims. Takes longer yes, but it allows you to see what the vehicle will be doing once you put it back into closed loop via temp and self timing corrections. It may be a very backwards way of doing things, but this has saved me from having to restart from the beginning in the past. Don't know, may be hurting me on this one.

BUT, here's the other problem... I don't have the stock tune files. Didn't get them before the other tuning shop did "their" tune. I do have a 2012 vette file I could use I guess, but I know the 10's had something unique about them in the fueling department. This really shows up more on the power enrichment side though I believe. GM really leaned out the fueling to get rid of the torque and power dips that these things are known for in the lower rpms. Don't know if anything else was changed for the 10's or not, just for the simple reason I don't have the OE tune file.

Never had a problem quite like this one, but I guess we all have our problem childs.

Oh, and if I didn't say it before. Thank you Greg for all the injector data. Looking forward to when you get more of it out there.

Thanks again

[GHuggins]

All right, taking my stupid hat back off and listening to the wiser. Put it into open loop and what do you know. Even though my closed loop fueling was staying around 14.5 average af, even after tuning in MAF and VE, put it into open loop and now was running at 15.1af ratio. SO starting over. Lets see how it goes this time.

Any other thoughts on the subject other than this?

[ingsteve]

i hope your injectors data are ok.
Anyway 15.1 you are close...

[PST]

As for stock files, have you checked the repository?
Open loop ve tune it, closed loop MAF. I still highly suggest finding a stock file to copy over as a starting point.

Where is the MAF located on this kit?

[GHuggins]

i hope your injectors data are ok.
Anyway 15.1 you are close...

Yes, I was really wondering if my injector data was right or not, but it's the best I've got right now. Really hate it when this so called flowing such and such injector brand throws everything off. After all it seems that not even the majority of injector manufacturers know what their injectors actually do. The 02 switching thing is what's been bugging me along with the engine temps affecting my af values. The temps issue is why I originally tried adjusting the transient and injector timing numbers. Even if your fuel trims are off + or - "lets say even 12 percent" on your long terms, your engine shouldn't be revving up and down with the 02's switching their voltages. I've just about completely dialed my af back into 14.5 and I'm almost afraid to put it back into closed loop because I'm sure my problem will still be there. Really getting tired of getting the fueling right, only to have to redo it due to changing something else with the injectors or so on in the tune. This is why my pre af ratio was out at the 15.1. I had already changed some of the other values in the tune that lead to it being off.

I'll have to check the repository as PST has said. Honestly didn't know there was such a thing on here. As for the MAF, it's a LP large bore "cobalt" MAF located about a foot in front of the TB. I know this will also cause problems later, but for now I just want this surging fixed. Already unplugged and removed MAF from picture to confirm it wasn't my problem.

Thanks again for all the help so far.

[PST]

The repository is not part of the forum. Go to the hptuners website and sign in to access it.
Since you tested without the MAF.. next question. Where are the o2's located in relation to the turbos and wastegates?

[GHuggins]

Precat 02's are maybe 8 to 10 inches downstream of turbos. Haven't had a chance to finish the open loop VE tuning yet and probably won't be able to get back to it until middle of next week since it is a shop car, it doen't take absolute priority. Guess that could be considered a good thing right now. Checked out the repository and factory tune for the 10's. Did notice an airflow idle setting was different now, but I don't believe this was my problem, because I think I had changed it at my attempts to fix the issue. Who knows this VE correction this time may fix it all...

Did notice something else the last time I was dialing it in that was also questionable to say the least. My whole passenger side bank is running leaner than the drivers. Noticed this a little previously in my datalogs, but didn't think it was bad off to cause any problems until this last VE load testing. While doing my high gear loading at low rpms it started bucking all due to my passenger bank spiking at an 18af ratio. Here is the last datalog so you can see what I'm talking about. The drivers bank also went lean in this area, but not as bad as the passengers, so I may redo my af settings in the datalogger since it's in the drivers bank and set things a little richer. It's always easier for the computer to pull fuel for the drivers bank in these areas vs. adding for the passengers, right? Even considered these injectors to be non flow matched and I've got one flowing incorrect on the passenger bank. I know I don't have any vacuum leaks or atleast I shouldn't since the boost would be blowing out somewhere. Guess I could go as far as swapping injectors bank to bank and seeing if my problem follows. The leaner cylinder definately could be the cause of all my problems...

If you click on the 9th histogram and have the table pulled up at the same time you can see what I'm talking about with the bucking. May even need to open the tach display to really see it, though.

Your thoughts?

[MikeOD]

You may be suffering from the "rich after flash" problem...where you're gathering data that falsely appears rich, and causes you to go lean when you think you're done...which causes your fuel trims to add fuel back and throws off the airmass model.

Every time you flash the thing...drive it around for 10-15 minutes (don't be nice to it or it takes longer) before you click record in the scan tool.

I still haven't found a way around it with my LS3...totally drives me nuts.

[GHuggins]

You may be suffering from the "rich after flash" problem...where you're gathering data that falsely appears rich, and causes you to go lean when you think you're done...which causes your fuel trims to add fuel back and throws off the airmass model.

Every time you flash the thing...drive it around for 10-15 minutes (don't be nice to it or it takes longer) before you click record in the scan tool.

I still haven't found a way around it with my LS3...totally drives me nuts.

So this is normal for the ls3's? Huh, I just thought it was because the fueling wasn't quite right yet or my first start up open loop cells (which I'm sure I'm wrong here) but aren't they the first or last cells in the PE table, were not touched yet but yes I started about half way through the tuning process realizing I was fighting myself on this and starting "driving" it on the dyno before I started datalogging just for this reason.

Side note. I had another tuner by the shop today and he tried putting the car into strictly speed density using the 2bar map operating system in the OS section only the car would no longer run like this - it just started then died - which he hadn't seen happen before. Then he tried applying a speed density patch - same start and die problem. He also said I was missing the "live tuning" function where the tune aparently self corrects or atleast gives you the ability to correct the cars tune while it is running. I've got the pro hp setup, but apear to be missing some of it's functions. IS this normal for this car or is something wrong with my HP system. He even showed me a video of it being used, so I didn't think he was making it up or something. (did kinda get the impression that this was only available on the older cars, but it was being used on a newer GM truck) The live tuning would be very nice to have from what I've seen, so how would one get this?

Thanks Again

[PST]

Live tuning is only available on the gen3 pcm's. Just writing a speed density os to the car won't work. You have to have various settings right and the VE table would need to be built. (Use bluecat to create one from your cars oe coefficients.)

The rich after flash issue, for me seems to come and go with various hptuners updates. A quick way around it is to use the vcm controls to reset the fuel trims. Makes them start from zero and rapidly start learning.

I have an 'ok' 2bar LS3 twin turbo file you can sample from, but it's for a 10 camaro, so it will have some different settings. Pm me your email address and I will send it to you.

As for the bank to bank fueling difference.. pull the o2 sensors and swap them side to side. See if the condition follows. My bet is a dying sensor. If not, then suspect injectors.

I still haven't looked at any of your logs as I've been traveling, lol. But make sure LTFT's are off. Just use STFT till its almost done.

[MikeOD]

The rich after flash issue, for me seems to come and go with various hptuners updates. A quick way around it is to use the vcm controls to reset the fuel trims. Makes them start from zero and rapidly start learning.

Doesn't work for me...if I have a tune running +/- 2-3% fuel trims, and I flash the exact same file back in with no changes, it'll wind up with huge 10-15% fuel trims...I can hit reset and they'll quickly learn right back up to 10-15% again...if I drive around for 10 minutes they'll start to learn their own way back down...at that point if I reset them, they'll stay closer to the 2-3% that they were before.

It's been present on every build of HPT (from various beta's of 2.23, through tons of builds of 2.24, and even a few of the 2.25 betas).

[GHuggins]

Live tuning is only available on the gen3 pcm's. Just writing a speed density os to the car won't work. You have to have various settings right and the VE table would need to be built. (Use bluecat to create one from your cars oe coefficients.)

The rich after flash issue, for me seems to come and go with various hptuners updates. A quick way around it is to use the vcm controls to reset the fuel trims. Makes them start from zero and rapidly start learning.

I have an 'ok' 2bar LS3 twin turbo file you can sample from, but it's for a 10 camaro, so it will have some different settings. Pm me your email address and I will send it to you.

As for the bank to bank fueling difference.. pull the o2 sensors and swap them side to side. See if the condition follows. My bet is a dying sensor. If not, then suspect injectors.

I still haven't looked at any of your logs as I've been traveling, lol. But make sure LTFT's are off. Just use STFT till its almost done.

PM sent...

What are your thoughts on using different 2 bar setups in bluecat. LS1, LS2 and so on for the LS3?

[MikeOD]

PM sent...

What are your thoughts on using different 2 bar setups in bluecat. LS1, LS2 and so on for the LS3?

Not sure if you're only asking PST or all of us in general. The 2-bar setup in bluecat works great, but you need to define the table properly and log the same resolution to do it. This will not have a seperate power enrichment and boost enrichment...so if you dial in the VE to accurately match airflow, you're going to command fueling with PE that is going to be too rich out of boost.

If you go to the enhanced OS from HPT, you get a boost enrichment table, which at least allows you to add fuel based on manifold pressure.

The only other way is to use Excel as an intermediate, where when you log data and calculate a VE table, you use Excel to alter that VE to match the AFR you actually want, and also to alter the % error from the logged AFR for what you desire...not necessarily what the ECM is commanding...it works great if you're careful setting things up.

[GHuggins]

Not sure if you're only asking PST or all of us in general. The 2-bar setup in bluecat works great, but you need to define the table properly and log the same resolution to do it. This will not have a seperate power enrichment and boost enrichment...so if you dial in the VE to accurately match airflow, you're going to command fueling with PE that is going to be too rich out of boost.

If you go to the enhanced OS from HPT, you get a boost enrichment table, which at least allows you to add fuel based on manifold pressure.

The only other way is to use Excel as an intermediate, where when you log data and calculate a VE table, you use Excel to alter that VE to match the AFR you actually want, and also to alter the % error from the logged AFR for what you desire...not necessarily what the ECM is commanding...it works great if you're careful setting things up.

I agree with you 100%, although I have been keeping it out of PE. NO NO I know, but I've been staying real easy on it while tuning. I am also using the accel spreadsheet method. Which brings me to another question. Would it be possible to do a half half multiplication setup in HP or Bluecat. OR is it possible to keep you whole histogram showing its numbers instead of all the black cells popping back up. Heck this is the only real reason you have to use accel to paste your numbers in.

The reason I was asking about the different VE tables is because say you use the ls1 2 bar vs the ls3 2 bar. The ls3 2bar really erraticates the table whereas the ls1 only slightly modifies the table. I just noticed it was somewhat easier starting with the ls1 vs the ls3's 2bar conversion for this reason.

I also haven't been able to apply any of the different OS systems to this tune. The motor always just cranks then dies when attempting this. Something about the ls3 from what I believe I was told previously?

[PST]

Sent you a 2bar enhanced OS tune from a 10 camaro e38. Copy the VE table and speed density settings to get you up and running on a 2bar OS.

[PST]

PM sent...

What are your thoughts on using different 2 bar setups in bluecat. LS1, LS2 and so on for the LS3?

Use the one that gets the best resolution, or setup a custom one. The bluecat software just generates a model for you to use, then spits out coefficients for the PCM.

[schpenxel]

Doesn't work for me...if I have a tune running +/- 2-3% fuel trims, and I flash the exact same file back in with no changes, it'll wind up with huge 10-15% fuel trims...I can hit reset and they'll quickly learn right back up to 10-15% again...if I drive around for 10 minutes they'll start to learn their own way back down...at that point if I reset them, they'll stay closer to the 2-3% that they were before.

It's been present on every build of HPT (from various beta's of 2.23, through tons of builds of 2.24, and even a few of the 2.25 betas).

Same here (haven't tried it on the 2.25 betas though yet)