Idle Cells Vary too [uggghhh] much!

[2002_z28_six_speed]

How do you guys prevent the idle cells from producing such a volatile AFR?

I have adjusted them again and again. But, when sitting the AFR goes around like crazy from pig rich 12s and 11s to extremely lean. This is just stupid for it to be perfect 14.7 one second and 15.6 after 20 seconds.

I am sure you guys are going to say oh well it is your IAT table, A0014. I have been watching the IAT and I see it climb to 130 F when sitting and that seems to be when the 15.6 AFR shows up. WHY??? When the IAT gets hotter why does it do this? Shouldn't it be the other way around? Plus, I have adjusted the table to no avail.

MY ECT table is programmed pretty flat and the ETCs are holding fairly well. I doubt that is the issue. If I had some stock 02s I would just plug them in and be done with it.

Maybe the way I am running the AutoVE is just flawed all together. But, that doesn't explain how I so easily nail down the cells where my vehicle is moving down the road...:bash:

[SSpdDmon]

How do you guys prevent the idle cells from producing such a volatile AFR?

I have adjusted them again and again. But, when sitting the AFR goes around like crazy from pig rich 12s and 11s to extremely lean. This is just stupid for it to be perfect 14.7 one second and 15.6 after 20 seconds.

I am sure you guys are going to say oh well it is your IAT table, A0014. I have been watching the IAT and I see it climb to 130 F when sitting and that seems to be when the 15.6 AFR shows up. WHY??? When the IAT gets hotter why does it do this? Shouldn't it be the other way around? Plus, I have adjusted the table to no avail.

MY ECT table is programmed pretty flat and the ETCs are holding fairly well. I doubt that is the issue. If I had some stock 02s I would just plug them in and be done with it.

Maybe the way I am running the AutoVE is just flawed all together. But, that doesn't explain how I so easily nail down the cells where my vehicle is moving down the road...:bash:
Welcome to my most recent fight. :) Under IAT's of 70*F, I idle ~14.2:1. IAT's around 80~90*F, AFR is spot on. IAT's over 100*F and it starts to lean out. I've messed with the B4901 Charge Temp. Blending table quite a bit. But, I can't seem to find the right numbers. :mad:

[2002_z28_six_speed]

Interesting. No matter your consultant: calibrator or the people that make the algorithms... they will all tell you any concievable model will be flawed. The question is what is the answer to this! Fun yet frusterating.

I just don't get why less dense air requires more fuel. Hopefully, it isn't a flaw unique to this COS which I just noticed. :-S

Maybe disable {B0109} Blend IAT with ECT to see if it makes a difference and then mess with the blending factor. If I know that ECT is not a variable during summer temps I should be ok messing with IAT.

I definetly know that at idle the IAT is weighted to make little difference. Without the MAF model to balance the SD model --- perhaps the MAF was mainly responsible for determining the temp because it knows the grams/sec. Temp is kinda factored in. SD can't know temp without the sensor which is currently being rendered useless due to the factory blending bias!

[redhardsupra]

bias might be only one part of the story. remember that on cold startup your fuel is cold as well, doesnt atomize as well, so you need more fuel to get the same amount of vapor. that's why OLFA dictates more fuel on cold motor conditions. if you dont believe me, on a fullly warmed up engine, go fill up your tank with cold fuel, and your AFR's are gonna go lean, as out of the same amount of fuel you'll get less ignitable vapor. few minutes of driving around, fuel in the tank warms up, and the AFR goes back to normal.

and yes, you are right, bias is calibratable with MAF. the problem is, how do you know that your MAF is properly calibrated?

[5.7ute]

I just don't get why less dense air requires more fuel. Hopefully, it isn't a flaw unique to this COS which I just noticed. :-S


It would be the IAT sensor giving a hotter than actual reading through heatsoak.

[2002_z28_six_speed]

bias might be only one part of the story. remember that on cold startup your fuel is cold as well, doesnt atomize as well, so you need more fuel to get the same amount of vapor. that's why OLFA dictates more fuel on cold motor conditions. if you dont believe me, on a fullly warmed up engine, go fill up your tank with cold fuel, and your AFR's are gonna go lean, as out of the same amount of fuel you'll get less ignitable vapor. few minutes of driving around, fuel in the tank warms up, and the AFR goes back to normal.

and yes, you are right, bias is calibratable with MAF. the problem is, how do you know that your MAF is properly calibrated?
I should be less careless when I post! Sorry, I am talking about warm restarts. My complaint mainly with idling while the car is getting warmer and warmer.

I like that cold fuel idea. I knew of that principle. I am going to watch my AFR next time I fill'er-up just for kicks!

Also, I am not using the MAF because my nitrous plate won't allow it unless I find a new plumbing solution. Which I am pulling slight vacuum past 4K RPM (90-95 kPa)



It would be the IAT sensor giving a hotter than actual reading through heatsoak.

I have an infrared thermometer gun. "Gunning" my plastic intake lid might confirm this if the "IAT" matches what my intake lid comes out as! I will have to check that out tommorow. I am aware that some of you have relocated the sensor. Perhaps this is the answer. Or just program the sensor for the false data.

I have some ideas now. Just need to test them.

[2002_z28_six_speed]

It would be the IAT sensor giving a hotter than actual reading through heatsoak.


You are so right as I found out today. I don't know why I thought the model was so perfect.

I blanked out A0014 and the AFR varied so little [as compared to before] as it should due to air temp while I was warming the car up to this proclaimed 57C!! [Blanking out the table got rid of all those multipliers throwing it so far off]

The IAT sensor reading was dead on with the temp of the intake lid as measured by my gun. It is not funny. :D

I opened the hood to induce Newtons Law of Cooling. It didn't take much time for the IATs to drop! The intake lid temperature and what the IAT said never strayed more than 3 degrees from each other. Most of the time 1 degree or dead on! At least the IAT sensor is accurate!!

I am left with either relocating the sensor or blanking out the A0014 as natural temps won't skew the AFR that much as compared to corrections applied at the wrong time. Blanking out the A0014 still means the car will go lean some while sitting. And other errors when the IATs arent the same as for when I tuned it...

I know this is old news to you, guys. New to me since running OLSD. [] Anywho, thanx. :cheers: Time for a Konig Ludwig Wiesse.

[mr.prick]

are you in closed loop?
because this happens to me in CL, with high IATs and only at idle.
and when it does it can hit 16.5 AFR :eek:
there is table {B3633} and {B3632}.
maybe someone here knows if they come into play in regards to this "phenomena", and if so how to adjust either one. :nixweiss:
also i notice for me 96*F IAT is when it starts to get funky.

[2002_z28_six_speed]

are you in closed loop?
because this happens to me in CL, with high IATs and only at idle.
and when it does it can hit 16.5 AFR :eek:
there is table {B3633} and {B3632}.
maybe someone here knows if they come into play in regards to this "phenomena", and if so how to adjust either one. :nixweiss:
also i notice for me 96*F IAT is when it starts to get funky.

Open Loop SD.

I just wish there was a way to log in a true 3D or 4D enviroment. Sure 2D maps are easy to read but looking at a more complex map helps you out in the long run.

Man! If I could run a RPM vs MAP vs IAT logging map with AFR as the result then I could see if air temperature effects the AFR more or less at different speeds. That could help me decide a more proper blending table.

Tons of other possibilites, also.

I am still getting bugged. I zeroed out the A0014 and the AFR still gets leaner when IATs go up. Just not to outrageously high values. I relocated the sensor and I am not seeing near high 50C temps. I guess the best thing to do is find out what the IAT will be when I decide to tune going down the road. Zero that area out. Tune only when IAT is at that temp. Then mess with getting it to work at other IATs.

To be honest an OS on the factory computer that read a WB would be even cooler. It could look at the trend to modify a semi-permenant map and other than that only modify the table when things get too hairy. Heh. I have a copy of Matlab....

[redhardsupra]

You're asking all kinds of good questions, and i've been working on this for months, so here's a quick example of just how much effect can wrong Bias settings have.

http://static.flickr.com/1315/1327844894_aca7256c33_d.jpg (http://flickr.com/photos/92934077@N00/1327844894)
the top table displays the various airmass figures given settings will yield.
the lower table displays the resulting AFR comparatively to the 'middle of the road' (.45 bias, 55C difference between ECT and IAT, IAT of 25C, GMVE of 1.3, MAP of 50kPa) values. this way you can compare how much it's gonna change if you got it wrong, whether it's your BIAS or your IAT that's got wrong data.

for example: if your settings show that at these conditions you have bias at .6 and ECT-IAT is 80C, and want to know the real bias, then you know you're idling at 15.7, but you want it to idle at 14.6, then you just move left in the 80C row until you find the 14.6. it happens to be at 0.3, so that's the bias value you should use.

of course this is not going to work if your GMVE values are wrong either, so you gotta get these values some other method, but that's a whole different issue. setting new bias is going to change airflow, which is going to alter GMVE anyway.

if you wanna play with different values, i can put up the full spreadsheet, for right now it's still unorganized, once i'll clean it up i'll post it on my site
lemme know if this helps you understanding some of these idle issues of yours. I find that most people dont realize how big of a difference can bad temperature accounting make. notice that there's a spread of 13.8-17.0 AFR for 80C ECT-IAT at bias values between 0.05 and .85. I knew it's going to be big, but not that big. no wonder doing idle in GTO's sucks if your IAT is being fed bogus information

[redhardsupra]

so is that useful to you?

[5.7ute]

Have you tried playing with B4902 Charge temp filter? If you set the low G/cyl area low you should be able to hold the blending factor from changing with the incorrect IAT's.

[redhardsupra]

the way i understand charge temp filter vs charge temp bias is the filter is the function, and the bias is the rate (derivative) of it. so if you got one, you can calculate the other. but it's nothing verified, that's just me oogling numbers so i hope someone has a more concrete idea.

[5.7ute]

B4901 The PCM calculates the charge temperature (in degrees Kelvin) using the following formula
273.15+IAT+((ECT-IAT)*factor) where factor is obtained from this calibration.

B4902 Determines the rate of change of the calculated charge temperature in response to changes in the IAT or the ECT.
Values closer to 1 will cause faster changes, values closer to 0 will cause slower changes.

The way I read it B4901 is the factor used in the aircharge calculation & B4902 dictates how sensitive to changing temps the calculation will be. I will do some logs hopefully this weekend & see if I can confirm or deny this. Since I have an aluminium OTRCAI heatsoak is a real issue.

[redhardsupra]

B4902 is an unknown to me. i've seen some people getting a repeated and expected effect from intended changes, but when/why/how/how much sort of relationships are still 'muddy'

i've done shitloads of work on this bias issue:
http://ls1tech.com/forums/showthread.php?t=710290
and
http://www.hptuners.com/forum/showthread.php?t=11527
are the most exhaustive discussions if you havent seem them yet.

while i understand temp blending much better than before, I dont have a solid calibration procedure that would yield a clearly superior bias table. I can get it, but it's going very sensitive to short pulse fueling, which only reminds of of the fact that we run all but stock injectors with a proper set of characteristics. we simply dont have enough data to deduce it, because we dont have a clear function, and too many unknowns.

[joecar]

Hmmmm.... interesting, all roads lead back to injector characteristics....

[SSpdDmon]

Hmmmm.... interesting, all roads lead back to injector characteristics....
So true.

[2002_z28_six_speed]

so is that useful to you?

Haven't had time to mess with bias yet. :(

I went back and reran my whole VE with very very tight constraints on the IAT and ECT. I also changed my fan temps to help keep ECT more of a constant. I only take data from the same IATs and ECTs Now, I know. :bash: Looks like this has helped alot.

I plan on keeping the running conditions as 0 on the tables. Now, I need to wait for the conditions to change from the idealistic ones I tuned under and see what the trends in AFR are like. I wrote down all the temps to keep track. That is all I can think of ATM.


Maybe I can vary to fan temps to get enough data points on ECT's effect.
IATs I will just have to wait for conditions to change to guess.

Learned alot since deciding to pitch the MAF and O2 sensors completely. I have no L or S trims to help...

[2002_z28_six_speed]

The following graph makes me think IAT has nothing to do with my problem.

Maybe it is the dynamics of the engine. You wouldn't think it to be that dynamic, though. It was explained to me that engines work better than hot. Sure, I know that. But to go from perfect AFR to lean in a couple of degrees? If this was true I wish the IAT table had better resolution.

Would someone with more experience give me a reality check.

[log ava on request]
[map of IAT vs ECT during idle]
http://forum.efilive.com/attachment.php?attachmentid=2096&stc=1&d=1189557398
[map of IAT vs ECT during idle]

[5.7ute]

Can you post the log & tune file for that graph. I would think there is some other variable coming into play.

[2002_z28_six_speed]

hmmm... I think I am going try the whole process of letting the car idle until it does this again with some different PIDs logged.

I have seen that sometimes it appears switching the AC on and off effects the AFR for a short period but starts to go lean again after a couple more seconds.


Right now I am speculating the following things:

Fans don't have the right compensation, not likely
Something in the direct airflow table, i have lm4 modded like ls1 with z06 cam, possibily
It is just the dynamics of the engine, I know engines make more power when hot maybe it pumps better when hot -> use more fuel
160 degree stat means idling temps and going down the road temps /= and require a lot of compensation for extreme differences
Seal or thermal issue

ram air hood could make the moving non moving issue more extreme

on the road
IAT 75-78F-84
ECT 185-189F-196

while sitting in the car port
IAT 86-92F-97
194-201F-205


{EDIT!}

Just watched the log again and the battery voltage dipped about the time the AFR went nuts.

This means something which is a large electrical load could be the problem or it could be because the battery voltage went to 11.3V and it affected the injectors dispite their voltage offset table.

[joecar]

13.3V->11.3V is a serious drop...! Can you account for this...?

[redhardsupra]

voltage drops fuel pressure, which results in lowering the IFR, so it should go lean (computer thinks it will get more gas than it will)

[2002_z28_six_speed]

voltage drops fuel pressure, which results in lowering the IFR, so it should go lean (computer thinks it will get more gas than it will)

Does {B3701} help or is it just one of those things our cars don't actually use? Thanx.

Just got back from under the car and found out the nut on the back of the alternator was very loose. The issue has been addressed.

I still wonder why the issue only showed up at idle. Voltage appeared normal while driving. Hmm.

Will update.

{edit}

Now the car runs 14.7 for awhile and then just runs very very slightly lean. Cycles between 14.7 and about 15.2.
That is probably all I can ask for....

Thank you for input regardless of if it helped or not. :P
It was appriciated.