ECT - IAT Bias table

[LazMan]

Yeah, I could do that too ... save me time. My tune is good-enough for open loop, so that should work ok.

Will do deanm11 ... tx for the good tip. I'll post the tune(s) and scans.

[redhardsupra]

yea open loop is the only way to gather data when you wanna attribute causes to results, otherwise you dont know if it's something you did or the computer just trying to outsmart you.
another tip: gather data only after the full car is warmed up, like 15min of normal driving, this way all the fluids are up to stable temps, so things like temp of fuel doesnt affect fueling.

[LazMan]

yea open loop is the only way to gather data when you wanna attribute causes to results, otherwise you dont know if it's something you did or the computer just trying to outsmart you.
another tip: gather data only after the full car is warmed up, like 15min of normal driving, this way all the fluids are up to stable temps, so things like temp of fuel doesnt affect fueling.

I intended to gather the data after driving home after work. (30-45 minute drive) ... Flash, start the engine, let it idle for 5 minutes, drive around the block once, start logging, drive around the block twice, come home, reflash and repeat.

I'll still do all 4 anyways, 2 open loop and 2 closed loop, with each set of Bias and Filters (stock GTO LS2 and stock 2006 LG4).

BTW... did you notice the 2007 LG4 has some fancy FANCY Bias values that are speed-influenced? ... now THAT is a lot to deal with!

[LazMan]

While my laptop charges ... here are the tune from my drive home today, the scan file, and scan data.

These will let you see that my tune, despite my not having a wideband anymore is pretty good.

I was able to peek at the laptop while driving and at just before the 23rd minute, I got a few throttle ins / throttle outs and I don't see any appreciable lean / rich swing.

Anyways, take a look. I'm charging my laptop now and hope to be able to go for another drive later tonight. If not, tomorrow.

-Laz

[redhardsupra]

BTW... did you notice the 2007 LG4 has some fancy FANCY Bias values that are speed-influenced? ... now THAT is a lot to deal with!

it's not hard at all if you approach it mathematically.
and without a wideband it's gonna be only informally informative. plus narrowbands take up a lot of PID bandwidth so the sampling frequency gets lousier, gotta keep the PIDs down to the minimum. to make such logs useful to me, i'd want:
GM volumetric efficiency,
manifold temp
Injector flow rate,
dynamic cylinder air,
pulse widths (one is ok if they're consistent across banks)
MAP,
RPM,
IAT
ECT,
spark,
retard,
throttle position.

if you're using MAF, we'd also need MAF frequency and MAF airflow.

[LazMan]

... I wasnt able to go out tonight... but I prepped 4 tunes.

All are the tune I posted previously, but they have DFCO turned off and also the following changes:

1. Stock LS2 GTO BIAS and FILTER tables
2. Stock LS2 2005 Corvette BIAS and FILTER tables
3. Stock LG4 2006 MonteCarlo SS BIAS and FILTER tables
4. Stock LG4 2007 MonteCarlo SS BIAS and FILTER tables.

When I say "stock" ... note my file is NOT stock. I mean the BIAS and FILTER tables will be stock as indicated above.

Yeah, I hate NarrowBands, but thats all I've got right now. Im glad my tune is still pretty close, at least part-throttle.

I'll strip everything out of my scan file except for the things you are interested in. Injector pulse widths are indeed consistent across both banks since I'm in open-loop, so I'll cut the 2nd one out. I'll set it up tonight, run it tomorrow, then post.

...oh ...and yes I'm using ONLY MAF ... no SD (even though it too is somewhat dialed in, but not enough. I turned it off)

[LazMan]

Almost an hour of set-up, stabilizing temperatures, waiting for sunset, loading tunes, and driving the exact same path ... and I forgot Dyn Cyl Air.

...and I'm adding Vehicle Speed so I can log the 2007 LG4 one. I'll collect more data tomorrow, but for now, here are the 2 runs I did today. I'm also thinking of removing my 2nd O2 sensor ... 1 crappy sensor is enough (?)

-Laz

[redhardsupra]

logging with a wrong scanning setup and failing hardware yields no useful results. wait for wb and fix your hardware in the meantime, until then, it's not worth the gas at these prices.

[LazMan]

Thank you for the straightforward assessment ... I tend to agree. It was worth it though. I had forgotten how much more data you get when you log less things! ...and I'm inspired. I got the software setup just right, and saved the file, and am ordering a new WB asap.

I'm very curious about this temperature modeling yielding usable tuning software.

So, happy Thanksgiving, and more thoughts and experiments (hehe) when all is ready.

[Phil99vette]

I took my IAT apart and changed the sensor inside. Car seemed to change IAT faster and my SD issues went away.
Phil

[redhardsupra]

I took my IAT apart and changed the sensor inside. Car seemed to change IAT faster and my SD issues went away.
Phil

you mean you put in the omega sensor or something else?

[LS6FD]

Am I correct in assuming this bias table is still used when running MAF as well?

[redhardsupra]

Am I correct in assuming this bias table is still used when running MAF as well?

we don't know for sure (only people who look at the ECU code can answer that for sure <HINT HINT to chris@hpt>) but i'd assume so, since there's two MAF fields, raw and SAE, so i got $10 on there's gotta be a TEMP adjustment. whether the TEMP gets calculated the same way as in SD or it's something else, again, we don't know (and again, Chris to the rescue!)

[LS6FD]

I've been thinking about this, but it doesn't totally make sense to me.

This table is basically a set adjustment for the IAT measured based upon ECT.

It seems to me that this adjustment would be the same regardless of if the IAT is 20 or 90 degrees (probably takes a percent number from the ECT and adds it to the IAT?) since it's a simple bias, and if the PCM is over/undercompensating for IAT changes, modifying this table wouldn't help, as it's just going to change the bias across the board for hot and cold IATemps.

I would think there to be another table based upon calculated MAT you'd need to adjust if you're having over/undercompensation problems with temperature changes?

[redhardsupra]

so you wanna compensate for the compensation? why not compensate once but well ?

[LS6FD]

so you wanna compensate for the compensation? why not compensate once but well ?

My point was that the table described doesn't seem to compensate for changes in the IAT itself. All it does is make an adjustment to the IAT based upon ECT irregardless of how cold/hot the IAT is. How would that help you if your tune goes lean when it gets cold outside?

If the air temp goes from 90 degrees to 20 degrees, and PCM over/undercompensates (for example you go lean when it gets colder, so the PCM is undercompensating), this table just doesn't seem to be the right one to be adjusting to compensate for that at all.

I'm probably wrong, but that's just my understanding.

[redhardsupra]

i'm not sure i understand you fully, but i'll give it a try:
the reason why the computer does not adjust for IAT (and IAT only) is because the IAT sensor is not in the intake, thus the temperature it measures is not the temperature we care about. the temp in the intake is something in between the two sources of temp: the engine temp (ETC) and the incoming air (IAT).
the whole biasing mechanism we're discussing here is exactly the mechanism that determines how in between between the ECT and IAT the real MAT is, how to arrive what it should be, and whether to adjust what we got, or calculate it from scratch.

and to answer the later portion, if the IAT temp goes from 90 to 20, that's the temp of the incoming air, not the temp in the intake. the temp air in the intake is going to be in between, depending on how much of the cold air is coming in, and how fast it's coming in. if you want more gory details, i highly recommend reading my paper on it: http://redhardsupra.blogspot.com/200...-modeling.html

[LS6FD]

I understand exactly what you're saying redhardsupra.

Basically ALL that this bias table seems to be doing to me though is just estimating the MAT at different airflow rates to give the ACTUAL air temperature as it enters the cylinders. I would think there to be another process/calibration to actually add/remove fuel based upon the calcuated MAT.

Say by whatever bias you have set it's adding 5 degrees temp to the IAT charge at a certain airflow and ECT.

So for 20 degrees and 90 degrees, all else static, it will estimate 25 and 95 degrees MAT at that given airflow and ECT (just an assumption). That's ALL that would come out as a result of the bias table calculations.

Next I would think the computer would take the calculated MAT and perform other calculations based upon that number to adjust the fuel. This is the step I would think you'd need to adjust if the computer is over/undercompensating for changes in air temperature. ----Some additional table that adjusts fuel based upon calculated MAT.

Note that these are all assumptions and I have no proof of what exactly the PCM is doing, but that is how it makes sense to me.

Basically I see the process as: IAT along with airflow and ECT is measured -> IAT is adjusted based upon set bias and ECT, resulting in calculated MAT -> Another process takes calculated MAT and adds/removes fuel accordingly

Hypothesis: Adjusting Bias table will only change the calculated MAT. Even if the MAT is 100% perfect, if whatever process using that number isn't calibrated right, you're still not helping the over/undercompensation issue for different air temps

[redhardsupra]

either you're not reading what i'm saying, or i'm not seeing why do you want to adjust IAT...

MAT is an ESTIMATOR. we dont really know what the temp inside of the intake is, it's our estimate. what is right is what causes the least AFR swings as the temps change.

let's not muddy up this thread anymore, if you'd like more clarification let's take it offline

[LS6FD]

as stated by bluecat:

In the past GM not only had a bias table for calculating the temp, but also a table that said how to change the fuel for each given temp. These tables don’t show up in the ls1 and later tunes. I’ve always assumed its because the table no longer exists, and the estimated air temp is just used for air density in a hard coded mathematical calculation as opposed to being looked up in a table. Anyway, don’t be too concerned on how temp directly effects the fueling, just understand cold = more, hot = less.

What I bolded is the table I was talking about that seems to be the one that would affect over/undercompensation of drastic temperature changes.

Given that information I just don't see how adjusting the bias table will help if your PCM is over/undercompensating for temperature changes. By changing the bias table ALL you're doing is just attempting to provide the correct estimated temperature to the table that actually changes the fuel for the given temp (The table that bluecat said is no longer shown.) Is this not correct??