general thoughts on tune/log?

[1 Bad Cav]

Hey everyone, first post here. I would like input from anyone willing and that has time to review my tune and log. I am sure I have done somethings wrong but hope to be able to learn from them. I am a complete novice when it comes to tuning. Tuning this car has been my entire learning experience.

I have a 2003 Cavalier 2.2 5spd that I have recently supercharged with an old cobalt ss kit. Additionally I have a 2.9" pulley, Heat ex changer, Seimens 60# injectors, dual pass and option b reservoir kit.

I know there are some obvious issues but have had limited time to do as thorough of tuning as I would like to. I can't figure out exactly yet why I get some real rich areas and some real lean areas other than the VE tables aren't what they should be. I feel the trend is way off in the low rpm tables. Again any input and thoughts on corrections would be appreciated.

Thanks in Advance.

Tune1.hpt
Log1.hpl

[eronq9]

I would cut back on timing till you get that fueling stabalized, you had quite a bit of spark knock, with boost that can cause damage quickly.

Do you have an error histogram for VE using your wideband, where you can force the comp into open loop and dial in your airflow? That will be the quickest way to get accurate fueling so your fuel trims arent so off.

[1 Bad Cav]

Thank you for your reply. I recently put the timing where it is now, which is actually the stock timing. Interestingly I had similar spark knock with timing decreased a couple degrees as well, which like you said must be due to inaccurate fueling? Which is one of the issues I am attempting to smooth out. I will pull a few degrees around where I get the knock to be safe tho.

The first graph in my scanner file, error vs tps, is the histogram for ve. I have been playing around with filters so I have quite a few histogram setups in my log.

The only problem I have with tuning this setup in open loop is its constantly commanding rich, whether in PE or not. Never commands stoich. The closest it will come is 14.0:1

[Dr. Nopps]

Cavaliers are VERY different. Only place in the tune you can actually command AFR is in closed loop PE. eronq9 is a more advanced tuner who is used to tuning much newer (normal) and far more complicated (modern) cars with literally hundreds more tables than the old dinosaur Alpha-N / Speed Density Cavaliers. You should be able to set up filters for scanning to ignore fuel trim changes/data that will happen in closed loop or get a wideband that can simulate narrow band o2 sensor inputs and prevent the trims from doing much of anything through the process. You might want to use far less timing than was in the stock tune's maximum areas untill most of this is ironed out, especially if the stock tune was naturally aspirated.

[eronq9]

I see I did look at the tune but didn't look at every table, that's a bummer. You could at the least use the fuel trims as a way of dialing in the fueling on the VE tables.

Then use your wideband exclusively to tune VE. Even if open loop Afr isn't constant you can still find the error between your wideband and the commanded AFR. But It won't be as accurate because the big injectors aren't as linear as stock, especially in lower pulse areas. So fuel trim % may not actually correlate 1:1 with actually percentage of fuel removed or added.

As far as timing, you need to recognize the fact that the stock timing table was made assuming 100kPa was the highest pressure. However since you're boosted and your sensors literally aee maxed, your log is returning a boost value of 0psi at WOT which is not correct.

You can get around this by only modifying the rightmost timing cells to just cover WOT, or anything above 100kPa, however you will have no resolution in your spark table so you will need to retune when you change pulleys and such. Go ahead and command significantly less timing while you tune fueling then once you get all the fueling dialed in you can come back and work on the timing table.

[1 Bad Cav]

Sorry it took so long to respond to this

Only place in the tune you can actually command AFR is in closed loop PE.

understood. When running with closed loop off the ecm will command 14.00 or lower during normal cruise.

You should be able to set up filters for scanning to ignore fuel trim changes/data that will happen in closed loop or get a wideband that can simulate narrow band o2 sensor inputs and prevent the trims from doing much of anything through the process.

I will have to do some research on how to set up these filters. Is this a better idea when the ecm commands rich all the time when in open loop than tuning in closed loop?

You might want to use far less timing than was in the stock tune's maximum areas untill most of this is ironed out, especially if the stock tune was naturally aspirated.

Completely agree. The stock tune was NA

[1 Bad Cav]

I see I did look at the tune but didn't look at every table, that's a bummer. You could at the least use the fuel trims as a way of dialing in the fueling on the VE tables.

Then use your wideband exclusively to tune VE. Even if open loop Afr isn't constant you can still find the error between your wideband and the commanded AFR. But It won't be as accurate because the big injectors aren't as linear as stock, especially in lower pulse areas. So fuel trim % may not actually correlate 1:1 with actually percentage of fuel removed or added.

Makes complete sense. If I take the commanded and divide by the actual could I use that number to then multiply by % to correct the areas used in closed loop? As they do not correlate 1:1 would you then use multiply by % 1/2?

As far as timing, you need to recognize the fact that the stock timing table was made assuming 100kPa was the highest pressure. However since you're boosted and your sensors literally aee maxed, your log is returning a boost value of 0psi at WOT which is not correct.

I agree. I understand esp with knock, timing needs to be dialed back. I am not 100% sure of what your referring to when you say my boost value is 0psi? Just meaning that it is only reading atmospheric pressure at WOT?

You can get around this by only modifying the rightmost timing cells to just cover WOT, or anything above 100kPa, however you will have no resolution in your spark table so you will need to retune when you change pulleys and such. Go ahead and command significantly less timing while you tune fueling then once you get all the fueling dialed in you can come back and work on the timing table.

understood, I will reduce timing until I see no more knock and tune from there. I do not plan on doing any more upgrades to this thing so future tuning will be minimal.

Thanks for your time and responses gentleman. They are greatly appreciated.

[Dr. Nopps]

Sorry I've been too busy to remember to check in every day myself actually. Looks like we're going to give you the basic crash course on Cavalier tuning. What's meant by "boost of 0psi" is that 100kpa Max (1 BAR) is the maximum value shown in any MAP based table (fueling or spark) in your current tune, however if you view a tune for a factory equiped turbo car you'll see those tables go farther usually 200kpa or so. There are a couple ways to "solve" this. You can convert your tune and setup to a "Fake 2 BAR Tune" where you replace the stock 1 BAR MAP sensor with a 2 BAR sensor, then multiply your injector constant and VE fueling tables by 2 and multiply your spark tables by 0.5 and there's a serious chance I've got those three backwards so research "Fake 2 BAR tune" for more info on that before you try it. A slightly easier approach would be to install the MAP sensor from the GM supercharged Cavalier, with the supercharged Cavalier tune from around 2002-2004 which may still be in the repository. I do not know if you can simply Write Entire this tune onto your PCM or if you have to get a PCM that's more model/year/application specific. That tune is essentially a factory GM version of a Fake 2 BAR Tune (I think it's a 2 BAR) but with all the tables correctly re-labled so you would'nt have to multiply & divide your spark & fueling with that one, just properly do the conversion math for the upgraded injectors. And sorry to eronq9 but you will not be able to tune off of %error of commanded AFR in open loop on this car because running in open loop, (Alpha-N) there is no commanded AFR to calculate error off of like there is when running in (Speed Density) closed loop. Open loop Cavalier commanded AFR is in constant fluctuation depending mainly on RPM vs. TPS vs. Air Charge Temperature. The filters mentioned earlier are little personal custom equasions that you add in when creating what used to be called histograms and I think are now called math parameters, which are basicly the setup of your scan. The filters force the scan to ignore certain data, like leave out all results below 1500 RPM, or how to ignore errors greater or less than a specified factor for example. (there are many, many more people have thought of) And I personally believe almost anything is better than running too rich all the time as long as we're not talking about trading one evil for another, lol. Also for clarity the only way I could command AFR was: in Closed Loop, in PE Mode, and only after at least 1% throttle regardless of whatever qualifier I entered in my tables to enable it (0%). My laptop is fried (motherboard) so I unfortunately can't post a copy of that 2BAR GM tune that I actually have saved on my hard drive, and I won't be able to look at your tune or scans otherwise I definately would, sorry. But at least I took the time to type all this out on a bloody iPod and will continue to check back from time to time to see where you're at with this. EDIT: PS It's a day or so later now I also forgot to mention if there are certain check engine light conditions present or missfiring these cars may just outright refuse to go into closed loop as well. It's indeed a very weird little car to be tuning when compared to most anything else on this forum, but it's (MAF' less) true speed density from OEM "sometimes", and as mentioned can be "faked" for boost if properly equipped, and have fewer tables to learn than most other operating systems so if you're willing to work with it all they're pretty cool and do have potential. I still plan on building a setup with this older stuff regardless of what newer stuff is available. It's neat just to see how other stuff even if drasticly different, or rare, was employed. There have got to be actually tons of early fuel injection setups, mabey even far later if you consider motorcycles that probably use this model or something like it.

[1 Bad Cav]

sorry again for the late response. I really appreciate the information and your time. I believe it's a 2.5 bar sensor which I do have currently. I am actually using the IAT signal from it in place of my factory IAT sensor. Interesting you talk about the fake tune due to the fact I have been looking into this more lately. I am definitely willing to try the fake tune out as it could be fun. I do believe the GM factory S/C tune is locked and cannot be read nor modified at least for the ecotec. Once I get time to tune I will report results. With the holidays and family and work and you know...Life, sometimes tuning takes a lesser priority lol. I'm not sure if this was the best platform to start with and if it will help or hurt when attempting to tune other platforms in the future... Thanks again I truly appreciate it.