AndyTA

Hello All,

Long story short.
Running a gen 4 ls427 on my 99 fbody. Speed density.. STFT disabled. 102mm intake. Tuning it on my own, have AFR dialed in at 14.9 to 15.1 - with my WOT at 12.7. so far so good. However:

The longer I stay in idle, the leaner my idle slowly becomes. Hence the "lean idle creep" - just gets leaner and leaner, and eventually it'll settle around 17.5 to 18 AFR.

What am I missing? When I first come to a stop after driving for a while, it idles where it should, like 14.8 or so. Just keeps creeping up from there.

I appreciate any input. Is it possible I suffer from vapor lock? (Aluminum intake, and billet fuel rails)

Thanks,
Andy

ChopperDoc

It’s probably due to heatsoak and your “Cylinder Charge Temperature” bias. As the airmass gets hot, the PCM will assume there is less air in the cylinder and inject less fuel, purely based on ECT and IAT temps. Since there is no further corrections from STFT’s, the PCM doesn’t know what happens after the initial g/cyl calc.

Apparently reality is you are getting more air in than it’s accounting for as it gets hot.

2 things you could do are either adjust the table in that area at idle, higher numbers bias ECT, lower biases the IAT...

or,

Actually relocate the IAT sensor to the intake manifold and disable the complex model (bias) completely.

gametech

iTrader: (1)

Excellent advice overall, but do you really like the idea of eliminating the complex model entirely? This could be problematic during cold start transitions or overheating situations, could it not?

ChopperDoc

No not really. The bias is only there to account for the difference in block temp and the intake temp because of the sensor locations. Newer cars have a MAT sensor instead of IAT, which is manifold air temp. It will read the actual temp of the air before it goes into the cylinder, and is more accurate than the bias if you ask me.

On a cold start it’s gonna read whatever the temp is in the intake, which is what the bias is attempting to account for in the first place.

AndyTA

Bad ***... Thanks ChopperDoc.

What do you do anyway? I see you answering literally everyone's tuning questions like a genie being summoned any time someone has a tuning question!

Pretty damn impressive.

I hope you realize how much the board here appreciates you.

Question:

Any negative side effects to turning on STFT in Open Loop?

Thanks,
Andy

ChopperDoc

I'm actually a helicopter mechanic by day, and a dedicated car enthusiast by night, and on weekends, pending written wife approval. I picked up and got into tuning years ago when I did my last build. It was so intriguing and such a challenge, I was hooked instantly. I've been building and working on cars most of my life, and tune friends cars and do mods here are there, mostly for beer and cost (credits), unless they're buying parts from me. I've done lots of tuning experiments on my own car for the sake of learning more. I can't just leave it alone until I fully understand something. I still have plenty to learn of course, as I believe everyone does and no on knows everything, but I seem to have gen III down pretty good lol.

Lately, as an advisor to a foreign military, I have much more time to study this stuff in finite detail (like formulas and so on) and of course answer questions to problems I've encountered over the years. I'm no pro, but I do love knowledge and once I understand something it seems to stick. Hopefully this has helped people out because that's the goal.

The only downside to OL STFT's you'll find is that they tend to mess with idle a lot with larger cams. Large cams (my last upgrade was from the TSP stage 4 to a 244/252 @ .050 by Pat G) tend to actually need less fuel. And since overlap will cause a "false lean" condition at idle, that can cause surging at low RPM. They will work for cruise just fine, and if you want to use them, tune so they are the target over the WB, since the PCM doesn't care what your WB says. To fix the idle surging they will probably cause, just put it in PE mode from 0-1600 RPM or so and set the PE EQ to 1.0. Make the TPS enable 0% and set the map enable under your normal idle map by about 10 kPa. I think the stock setting is like 15 kPa anyway, and if it's still this setting then it's fine. A lot of folks use MAP as the "trigger" for PE because it reflects load better than the TPS %.

ChopperDoc

I'll also say this, lately I've been taking thermodynamics courses just so I can figure out how the F transient fueling works. I understand which way to push the numbers and for what reason, but what if there is a formula that we could use to actually calculate it? That's been my goal lately. Just using scanner and playing with logs night after night as well as formulas and ideas, I know it is possible to figure out what the numbers are actually a calculation for. There is so little information on this available, and I think it's because it's just to complicated to answer in a forum. Chris at HPT has a few general answers on it, but again, not real calculations. I may never find the answer, but it is fun to try lol.

So that give you an idea of how much time I have to myself here. Too much, but if I can produce something tangible from my research, it won't be in vain. Only real problem is I don't have my car here with me to experiment on. I do have a Tahoe though, but it's not mine since it's issued. I might start tuning it just for fun anyway after I go on leave and snag up my portable WB I have in my garage. Not sure if they'll get mad at me, but I can put it back to stock when I'm done lol.

ChopperDoc

Here is what a stock tune would look like with OL idle setup up using PE to disable the trims at idle. I changed the EQ ratio to the standard 1.17 as well, since stock is pig rich at like 1.3. Note the MAP enable as well, stock is indeed 15 kPa.




By the way, I forgot to say I also really appreciate your comment, and I am almost tempted to make that genie part a sig LOL. Hope this is helpful.

JHZ28

iTrader: (3)

i have pretty much he same issue with idle when hot it gets leaner as iat goes up. im running a olsd tune pleae explain how the cylinder charge temp would possibly help this issue

ChopperDoc

The intake is not a sealed space, so as air heats up it spreads out, this includes in the filter area as well. Pressure will remain fairly stable, as you can verify in your scans, the MAP will not change a significant amount. I explain it in detail in the cylinder airmass explained video.

Using the Ideal Gas Law, we know that with changes, something else must change accordingly. So...

PV = nRT

P [pressure] will not change much
V [volume] will not change
n [moles] is airmass, and will indeed change because...
T [temp] is going up, therefore something else must go down, in this case the value of "n" which is the airmass
R [universal gas constant] will not change.

So as temp goes up, airmass goes down, because the intake is not sealed. Molecules can escape back out at the entry point, so with added heat airmass will be reduced, all the while pressure (MAP) will not see much change at all.

The temperature bias is in place to correct for the difference in temp of the incoming air and the block temp. So if it's off from the stock calibration, as in different flow, different parts, ect, it will be off 99% of the time because cams change flow [vacuum or pull], and different intakes, filters, and throttle bodies change the calibration as well. So think if the bias is in place to account for the change in temperature from incoming air vs block temp, don't you think that with a larger cam the amount of time will change as well with reduced vacuum? It absolutely does, because the amount of time the air is exposed to heat before it reaches the cylinder has increased because vacuum is much less than stock, so the air moves slower.

Check out my video on airmass explained and hopefully that helps better your understanding. I cover the bias in there too, and you will see that the math changes the airmass calc completely.

ChopperDoc

Not trying to make this too complicated over that last reply, but another factor is the speed of the air, and I didn't mention what that does to density lol... Though there is nothing to account for this in the tune, Bernoulli's Principle is also factoring in here. That principle deals with pressure... so as the velocity of a gas increases, the pressure decreases. It's how we create lift with a wing or rotor blade. We accelerate the air above it and slow the air down below it, creating lower pressure above the airfoil, thus creating lift. It's no different in the intake when it comes to density and velocity. This principle is actually used in the design of the exhaust ports, and why they are "D" shaped and not round.

So consider if the air is moving slower now, the density of that air has increased. The bias would have been calibrated (stock) assuming a higher flow and pressure delta (difference), and therefore would assume lower pressure in the intake area. Now that vacuum has been reduced, and the flow has been reduced, the amount of air being forced into the space will also have been changed.

So the physical pressure in the intake has increased, the pressure in the intake area has increased, and likely the volume has also increased with different parts in most cases, ALL of which have not been accounted for at all in the tune. An SD tune relies on the math to be right to work in all conditions. When you throw heat at a non sealed area, the density in that area will be reduced, but in this case not enough to be accounted for with heat alone. So the PCM is thinking there is less air than in reality as it get's hot, so it injects less fuel, creating the lean idle issue as you are seeing.

ChopperDoc

The only thing we can really do in such a set of conditions is to lower the denominator in the calculation, in this case the bias when solving for moles. Pushing the bias towards IAT will lower this number, thereby increasing the calculated airmass. It’s the only table we have that can account for this, because we’re dealing with heat, and pressure still doesn’t change enough to be of use to use something like the EQ correction or VE. In my example in the video the cell would not even change on either table because it only moved 4 kpa over a 70 degree change.

The airmass on the other hand came down from .277 to .205, which is 3 cells on the spark table, and this was heat alone. RPM also changed somewhat, because the log was a cold start and warm up, so the MAP was likely affected by this to some degree, but was still only 4 kpa in the log, and actually came down in pressure and not the opposite as would be typically expected here.

So where does that leave us? Back to the calculation and the bias of course because it’s the only thing we can correct in this case. If there was another VE table that used g/cyl and IAT, then we could account for this, but there simply isn’t.

Oh and we’re not even talking about transient fuel either, which also changes with vacuum, heat, and airmass (more air will diffuse more fuel). The PCM isn’t accounting for the increase in port airmass from the increase in pressure due to decrease in flow (Bernoulli again)... Though actual flow from mods will likely be more as we know from having to increase the VE numbers in most cases.

So much for keeping it simple I guess lol.

Try biasing the IAT more by lowering the number in the idle areas and see if your lean numbers come down some. They should.

Just remember when we change parts, we have changed stock calibrations, requiring us to tune for the changes. Understanding the physics that were applied in the first place to make the initial calibration (stock) gives us a taste of what may actually may need to be tuned to achieve OEM accuracy (which honestly isn’t that good in SD as many of us know). When you change a cam, you have changed the valve event which affects vacuum and transient calculations, which in turn affects heat and pressure, and AFR.

Remember in SD you are relying on a lookup table, the VE. This is % of efficiency or the % of moles * molar mass from the calculation. Because it’s math only, and not measured by anything, the parameters we can change are what will affect it. In this case those parameters are solely the %VE, and Temp bias. We can change volume, but this isn’t done to account for a lean condition last I checked lol.

So basically the MAP, volume, and other constants cannot be changed here. From the formula, that leaves us only 2 things left that are tunable: T (bias) and % VE. If the VE is right in most other places, (also assuming transients are right) then where do you think the error is? To go deeper, transient is tunable too, but that’s likely not going to affect it as much as the temp.

ChopperDoc

I can’t test this on mine since my IAT is in the intake manifold, and the fact that I am overseas at the moment. However, later I can take a look at some old logs and work out a bias correction factor. If you guys want to participate in testing it, that would be a fun and useful experiment.

The inputs needed will be

1. IAT
2. ECT
3. g/cyl
4. IPW and IFR (for grins)
5. MAP
6. VE if you can log that, but the numbers from the table will work.
7. WB AFR
8. Commanded fuel
9. RPM
10. Dynamic air

I'll need the volume too btw, from the tune.

I’m probably forgetting something, but if anything maybe I can work out some math in the scanner to give a correction factor, those inputs should do for my ideas here. I figure if we have all the numbers, it’s just a matter of rearranging the formula solving for “bias.”

AndyTA

ChopperDoc,

I will definitely be giving this a shot. I really like either option (changing bias, or relocating IAT) - I think between those two options, it would be far simpler to disable bias all-together, and just move the IAT........hmmmmm. Now the wheels are turning :-)
Oh please do put it as your sig - I would be honored! haha.

I think I'll definitely try changing bias first, if that doesn't work for me, I'll look in to the IAT relocation. I do believe my intake has provisions (bung) for an IAT.

Thanks,
Andy

AndyTA

I plan on taking PTO this week on Friday. If not, I'll see if I can find the time to get all those PIDS in (most are) and attach a short log. No promises!

Thanks again!!!

AndyTA

Has anyone seen ChopperDoc? Hope he's okay! Haven't seen his posts around here lately.

Anyway, I had to post-pone taking action on all of this advice, as I recently took the engine out - it just got done at the machine shop, and I'll be putting it back in soon within the next couple of weeks.

Solving the lean idle creep is a high priority for me. This is not forgotten.

Thanks,
Andy

ChopperDoc

Nah man no one's seen him around lol. Just been busy with some other things lately. I'm still around though

AndyTA

Glad to know it Doc! :-)

So, back to business. Engine is back in the car, and running!
As a reminder, I went with your second option, of relocating the IAT to the Intake Manifold itself. To disable the complex model, I assume I put 0's across the board? Screenshot of what I've done below.




Is this correct?

Secondly, the blue circle of Charge temp vs ECT - do I need to touch that at all?

Thanks for your time Doc,
Andy

ChopperDoc

Just use the drop down and select disable under the complex model. That disables the table.

AndyTA

Oh...silly me. I didn't realize a drop down existed for Gen 3 PCM.
I'll have to take a closer look at it when I get back home. Don't know where I missed that.

Thanks,
Andy