IAT temps changing my VE

[Dale]

I have added some new parts to my engine, and this part of the season the air temps can change up to 40degs in a day. I noticed doing data logs during different times of the day are throwing my VE all over the place. I see no place to change this in the "stock" tune. After searching on here I have found:

1. "lock" the IAT into one temp (is there a way in the pcm?).
2. Upgrade to cos5 and use A0014
3. If option 1 is not available, install a resistor inplace of my IAT while doing data logs.
4. Log during the same time every day, which doesnt help really cause even from day to day, you get about 10deg difference.


Now, also, if I go with option 1 or 3, then my fuel is going to be way off if I drive during really hot times, or really cold times correct? Making cos5 really my best choice? Any other thoughts?

Why/how would GM not have some option in the stock tunes to not adjust for this better then it does? :bad:

[mr.prick]

atmospheric conditions are going to have an effect no matter what you do.

[redhardsupra]

Dale, there's _a lot_ on the topic on the forums, just search for temperature bias, or temperature modeling, i've done a ton of work on it, it's all on my site, take a look at it, try not to get scared ;)

[Dale]

I have searched on here, but so much stuff. Plus I was just searching on IAT. I'll take a look at your site and the key words you gave me.

[macca_779]

A0014 is a huge help in this regard. I use it religiously.

[WeathermanShawn]

Dale, are you just logging VE values, or is it your intention to stay in OLSD?

Not trying to start a controversy here, but running in closed loop with feedback from the MAF will address many of the air density challenges you are facing.

Not discounting that there are ways to do it with a COS upgrade and many many people successfully run OLSD.

Just adding that I am pretty much successfully tuned and am running Closed-Loop. Did the VE tuning, calibrated MAF, re enabled closed loop, and car is running great (every day).

Living in Colorado where terrain and atmospheric conditions change very rapidly, I have learned to work within the parameters of closed-loop tuning.

Just a thought. You said you wanted any other ideas.

Don't know if this is an option for you, but in any case best of luck.

..WeathermanShawn..

[Dale]

I'm still learning this stuff, but I am infact running SD, no maf as the propper one wont fit in my air intake setup. I use to have one in it, and got killer mpg, but kept getting maf codes so I took it out.

I keep getting OL and CL confused. I have no clue what it is I am running. :doh2:


The car runs fine. But when my STFT swings -12 on the way to work, and +6 on the way home... I am sure something is wrong and I need to work on it. But the only thing that has changed is the air temp. I'm very surprised the tune doesnt adjust better for air temps, or have a chart that you can tweak.

[redhardsupra]

what are your IAT's on the way to work and back? if they're that consistent, i could probably eyeball a better bias table for ya...

[Dale]

The readings while moving are accurate with the weather channel so I know the sensor works.

The morning on the day I noticed, it was 72degs.
Drive home it was around 95-98deg.

Now the setup I have, when setting at a light it gets radiator heat BAD. Which I am thinking of ways to fix this.

If we didnt have 10days of rain, I would do a morning and afternoon log and post them with my tune. But I refuse to drive this car in the rain for safety of many people.

:edit: Shawn, are you saying if I had a maf sensor in the car it would adjust better? Maybe I need to rethink my complete intake setup and have it so a maf will work.

[WeathermanShawn]

Dale:

Understand. I just started EFILive about 2 months ago. A hundred hours later and a lot of help from this board, I think I finally have it down.

As RedHardSupra eluded to, biases exist in the VE computation concerning temperature. Just look at the denominator in the VE computation (charge temperature). It is a function of IAT and ECT, and how much each influence the charge temperature is dependent on the 'airflow factor' or how fast you are going.

I guess I would suggest going to one of the newer custom systems that allow more accurate commanding of the VE fuel computation as temperature varies.

As you stand now, there is a small error or bias. If you do a SD VE tune in summer when the air is warmer and less dense, at worst in winter your tune will be a little richer (air is cooler and more dense).

My advice is to not let it overwhelm you. As far as open loop or closed loop..I always just try to remember it like this:

Open-Loop: No narrow-band O2 feedback.
Closed-Loop: Narrow-band feedback.

When you set closed-loop enable temperature to 254F, it will never enable, thus you will always remain in open-loop..fuel is commanded by your 'commanded fuel when in open-loop'.

Hope all is this helps. Never any absolutes, but you can successfully tune these cars. And when you do you will be very happy.

Good luck..

..WeathermanShawn..

[Dale]

You sure learn faster then I. Ive been at this about a year, 200+ tunes. And its still not right.

Atleast I got my speedo right!

[Dale]

Finally got some dry weather. Here are yesterdays morning and afternoon logs, along with my tune. AM, it was 73deg outside. On the way home it was around 88-89deg.

I did a log this AM, will do one in the afternoon as well. I am finding I am getting heat soak from my motor even while driving :bad: This is getting moved up on the list of things to do.

[redhardsupra]

Dale, I'll look at it in the evening, I got no EFILive (or Windows for that matter) at work.

[Dale]

Thats fine, no major rush. I'm going to edit my tune once more, then start working on trying to fix my heat soak issue.

Trying to come up with ideas and its not going well.

[redhardsupra]

Dale, you don't have enough PIDs in there to make this useful unfortunately. Bias/filter stuff is based off airflow, and you're not logging airflow or airmass, or airflow frequency, or anything else that I could derive the airflow with. IAC, cracker, follower, all this stuff doesn't help me any.
so here's a list of what you need:
speed
IAT
ECT
DYNCYLAIR_DMA
DYNAIRTEMP_DMA or CHRGTEMP_DMA (whichever works)
MAFflow and MAFfreq (if you're using MAF)
VETABLE_DMA (if you're using SD)
AFR (commanded)
AFR (wideband, if you dont have any STFT's will do)
IBPW's

and it's best if you log it all in metric, makes math much easier.
sorry for the bad news, should've told you this earlier

[WeathermanShawn]

Dale:

While you are waiting for a full-blown test using multiple Pids and a lot of intricate time-consuming parameters and numerous variables, here is a interesting alternative suggestion.

While you may or may not have access to a Roadrunner PCM, under the EFILive forum 'Roadrunner', is an interesting tutorial concerning VE smoothing using Strims as a method of refining the VE curve.

While it is perhaps more specific to vehicles utilizing closed-loop function, the technique may still be of help to you.

Since Strims vary due to environmental changes, the technique described allows you to utilize those changes directly back into the VE curve. In effect it is addressing some of your IAT bias concerns you are trying to address.

I am sure on a purely scientific engineering standpoint that someone will dismiss the technique or tell you why it will not work, but I would just simply suggest you take a look at it.

Tuning is more fun if something works, and unless you have a Doctorate degree in math or engineering, might be more rewarding to to something that might may satisfy your needs.

I opted for a Roadrunner, and did more to nail my VE curve in 2 hours than the previous 2 months. And it didn't require numerous pids or formula's.

Don't get me wrong, knowing math and physics are great, but the point of science is not to make everything harder, its to make to work for you.

Maybe someone else can enlighten us as to the applicability of the above technique.

Again, best of luck.

Respectfully

..WeathermanShawn..

[Dale]

Supra, thanks for looking at it even though you couldnt do anything. As of right now the laptop broke and I have to fix that. I'm also going to make it get fresher air, and smooth out my intake. So it will be a bit before I am able to log again.

[redhardsupra]

While you may or may not have access to a Roadrunner PCM, under the EFILive forum 'Roadrunner', is an interesting tutorial concerning VE smoothing using Strims as a method of refining the VE curve.

RR does not give you any new functionality, it merely automates and speeds up the (incorrect) process. Thus, the results you'll get eventually will be the same, with RR you'll just get there quicker.



Since Strims vary due to environmental changes...

you got it in reverse, STrims are merely a correction to make up for what we cannot account for properly through models. the problem is the proper attribution: since STrims are the final 'catch-all' correction, you have no way of knowing if you're making up for dirty MAF sensor, imprecise VE table, clogged up filters or injectors, bad calibration of...anything. So I wouldn't be in such a rush to attribute STrims to anything in particular, that's not what they're for.



I am sure on a purely scientific engineering standpoint that someone will dismiss the technique or tell you why it will not work, but I would just simply suggest you take a look at it.

Tuning is more fun if something works, and unless you have a Doctorate degree in math or engineering, might be more rewarding to to something that might may satisfy your needs.

That's a slippery slope you're taking there. If you dismiss the scientific process, what are you left with--the ouija board?



I opted for a Roadrunner, and did more to nail my VE curve in 2 hours than the previous 2 months. And it didn't require numerous pids or formula's.

And you still get the same wrong result, but you got there quicker! progress!



Don't get me wrong, knowing math and physics are great, but the point of science is not to make everything harder, its to make to work for you.

and the fact that you dont understand something should motivate you to observe and discover, and not run toward the comfort or acceptable misery.

Shawn, for a guy that has asked some very insightful questions, you truly sadden me with taking the hit off SSpdDemon's "git-r-dun" crackpipe. There's a lot of ways you can get close enough to the limit of observable errors doing completely wrong things--if you've studied any science you probably know about the schemes people have come up with to explain the heliocentric sun system, or the 'negative' mass of air. Just because some hack works for you right now does not mean it will work for everyone else at any other time. Only proper science can provide you with the models universal enough to hold true regardless of conditions. Everything else is merely a hack, in the most pejorative meaning possible. If you want to understand this stuff in more details, I'd be delighted to explain it to you, I've sent you my contact info before.

[redhardsupra]

Supra, thanks for looking at it even though you couldnt do anything. As of right now the laptop broke and I have to fix that. I'm also going to make it get fresher air, and smooth out my intake. So it will be a bit before I am able to log again.
no problem, I got plenty of other research going on. however, when you get your setup back up and running again, i'm very interested in seeing the data, as i've never heard of these effects so consistent and so pronounced.

[Dale]

no problem, I got plenty of other research going on. however, when you get your setup back up and running again, i'm very interested in seeing the data, as i've never heard of these effects so consistent and so pronounced.


Will do. I think my air intake setup which is not whats normally found on an OBD2 car is part of it. Then the air here changes so much during the summer. Its quite normal to have 68-72deg in the morning...Then hit 95+ during the day. Hopefully adding some ducts infront of the air intake will help get rid of some of the radiator heatsoak.

[WeathermanShawn]

Marcin:

First off, there is no absolutely no doubt that you have mathematical and engineering insights and gifts that far exceed the norm.

But, as with many highly gifted people, one can fall into a trap of rigidness in their values and easily misinterpret someones else views.

You need to be a little less rigid.

The point of any forum is to exchange views. Views can be based on facts, and yes sometimes opinions.

The Roadrunner does allow functionality and speed. Thats a fact. As a Meteorologist I realize I could write pages explaining how a cloud forms, but I think most people want to hear the condensed version.

While your statement that an incorrect process can just be done more rapidly, but so will a correct process be done very efficiently.

Concerning Strims, you are correct. There are many variables. My point is that environmental changes will affect Strims, and undoubtedly effect VE. That was the original posters (Dale's) concern. I wasn't excluding the many other contributing factors.

Would never dismiss the scientific approach. Thats what I do. I forecast the weather which arguably has far more variables than EFIlive utilizes. I have just learned you need to bring knowledge down to an understandable form.

Marcin, better to teach someone than overwhelm with intricate formulas and mounds of paperwork.

Concerning the Roadrunner and VE. Respectfully, you need to change your attitude. My reference to the Roadrunner and VE computation was using the published EFILive tutorial method. What is wrong with faster? My reference to the Strim method of VE smoothing was simply to point out another persons method that had success. That person seems to be a highly respected tuner, and his explanations make more sense to me and obviously to many others.

Marcin, I have not abandoned science. You are probably a genius, but could you be a little less intolerant of others? Your conclusions are too stubborn and absolute. You are mistakenly jumping to conclusions that are not really in the spirit of friendly information sharing.

Remember, this is a forum.

On a side-note I think the comment about SSpdDmon and a 'crack pipe' is a little out of line. I really respect you are passionate about this topic, but in all frankness if you could be more succinct in your explanations more people would understand you.

I am sure the find people who developed EFILive understand that insulting people or making their software to hard for an ordinary user to use, got it.

Marcin, we are all just trying to tune. I love science too. But, maybe this board is not the place to show off, but to help us all learn.

Thank you for letting me express my view.

..WeathermanShawn..

[WHYTRYZ06]

cos5 and a0014 are a life saver....