Idle issues and lean surging on decel at highway spees

[95GMC]

Having a couple issues on my tune which are kind of frustrating and I can't seem to fix. First issue is that in closed loop idle the engine is running lean and surging. I can see the factory narrow band o2 reading extremely low which agrees with my wideband which has read up to 16.5 or so when the worst surging happens, however, the integrator and block learn are both indicating rich? Doesn't make much sense to me.


I've tried more fuel and less fuel, neither helped. Tried more timing and less timing, neither helped. Is there something I can do to get the integrator and block learn to agree with the actual mixture? I also seem to be having this same issue in the low kpa range when coasting at highway speeds, the faster I'm going the worse the lean/lean surge gets. I thought maybe exhaust leak introducing extra air so I used the powersmoker smoke machine at work (goes to around 30 psi) and there are no leaks. Just to be 100% sure I ran it in open loop and had the same issue.


The rest of the tuning has went pretty decent and I'm fairly happy with everything else, if I can get these 2 things squared away I'll be ready to do some WOT tuning. I attached a log as well as the bin, xdf, and adx I'm using below and am running a 7427 ecm with $0D mask. I would really appreciate some advice on this, I've been pulling my hair out on this one.


One last thing I will note is for whatever reason my AEM wideband displays a slightly different reading than what tunerpro reads when using the linear egr trick to display afr with the supplied formula, so if you don't use my adx file the afr reading displayed will likely be off. Still not really sure why that is but with my adx they match up fine so I haven't invested much time into looking into it. Idle issue is near end of log, several decel issues in log but good one at 215 seconds run time, thanks again for any help/pointers.

[tayto]

what has been done to the engine? cam, etc. sounds like you are running lean if it's surging. WOT tuning is usually the first thing to tackle once it's ”safe” mixture wise

[95GMC]

It's a 355 running 9.5:1 compression with a stock lt4 cam. Has a factory 5.7 tbi unit bored out to 50mm and horns shaved off with new factory spec 5.7 injectors with fuel pressure bumped up to 18psi and 30 psi big block fuel pump in the tank. Edelbrock performer intake with summit vortec heads and shorty headers, I think that's pretty much it. I definitely need to get to the WOT tuning for sure

[tayto]

How does it run in open Loop? You can try disabling closed loop until you get ve map sorted. On near stock stuff not really needed but you are modified.

[tayto]

Here's a good thread to read: https://www.thirdgen.org/forums/tbi/398174-disabling-async.html

There's even some sample bins you can look at!

[95GMC]

It definitely idles a lot better in open loop. I've only actually driven it in open loop once and it ran ok but I was mostly just trying to see if it would help with lean surge on decel at highway speeds. I found out something interesting today, I figured if the truck was actually able to idle right in open loop maybe I could just run open loop at idle so I did some research and found "stay in open loop idle temp threshold" in my bin and change it to max temp but the thing still goes into closed loop somehow??? Thanks for the thread link I'll definitely check it out

[ralmo94]

It definitely idles a lot better in open loop. I've only actually driven it in open loop once and it ran ok but I was mostly just trying to see if it would help with lean surge on decel at highway speeds. I found out something interesting today, I figured if the truck was actually able to idle right in open loop maybe I could just run open loop at idle so I did some research and found "stay in open loop idle temp threshold" in my bin and change it to max temp but the thing still goes into closed loop somehow??? Thanks for the thread link I'll definitely check it out

If I remember correctly, there might be a flag for open loop idle. You could also change maximum and minimum IDLE BLM and INT to 128

[95GMC]

Finally got the open loop idle to work by using one of the bin files from the thread tayto linked and doing the exact same thing I already tried with raising the temp qualifier... I guess somehow the bin I was using was corrupted or something? At any rate I transferred all my VE tables etc. over to that bin and I have a decent idle for now. Before I did this I tried several things and still couldn't get the closed loop idle working well, I'm not sure what's going on it's a very mild engine in comparison to some I've seen on here and other forums that idled well in closed loop. I also drove the truck again and still have the lean surge only on decel from say 30mph and above that gets worse with more speed. I guess for now I'll just drive it and do some of the WOT tuning. I guess I could always cheat and enable the DFCO earlier so it can't physically run lean since it's not technically running but I really want to fix this the right way, whatever that entails.

[ralmo94]

Finally got the open loop idle to work by using one of the bin files from the thread tayto linked and doing the exact same thing I already tried with raising the temp qualifier... I guess somehow the bin I was using was corrupted or something? At any rate I transferred all my VE tables etc. over to that bin and I have a decent idle for now. Before I did this I tried several things and still couldn't get the closed loop idle working well, I'm not sure what's going on it's a very mild engine in comparison to some I've seen on here and other forums that idled well in closed loop. I also drove the truck again and still have the lean surge only on decel from say 30mph and above that gets worse with more speed. I guess for now I'll just drive it and do some of the WOT tuning. I guess I could always cheat and enable the DFCO earlier so it can't physically run lean since it's not technically running but I really want to fix this the right way, whatever that entails.

You might think about switching to a 4 wire o2 from like a 98 or something, and run the signal ground right to the pcm wire that grounds on the engine, and not ground that on the engine, to make sure you are getting accurate signals. Otherwise the signal ground is on the exhaust pipe.

[95GMC]

That's a great idea and heck for the price I'll definitely give it a shot

[ralmo94]

Make sure you DON'T connect signal ground to heater ground

[95GMC]

Oh yeah for sure definitely don't want to try to pull 5+ amps through the signal ground haha I imagine that'd fry the driver in the ecm pretty quickly

[tayto]

disable DFCO and see if the lean surge goes away. it is one of those things you should enable AFTER you get your VE table dialed in. Since you are running headers,.you will probably have to mess with proportional gains. A 3 or 4 wire is a good idea for headers IMO.

[ralmo94]

disable DFCO and see if the lean surge goes away. it is one of those things you should enable AFTER you get your VE table dialed in. Since you are running headers,.you will probably have to mess with proportional gains. A 3 or 4 wire is a good idea for headers IMO.

There may also be a Decel ENLEANMENT

[95GMC]

My truck does have a factory 3 wire o2 but I'm going to do the 4 wire swap anyways, I like the idea for sure. I have dfco disabled when this happens but I'll make sure to disable decel enleanment before I take it out again.

[tayto]

if you have 3 wire i don't 4 wire is really necessary. you have a tune problem not mechanical

[ralmo94]

if you have 3 wire i don't 4 wire is really necessary. you have a tune problem not mechanical

The O2 low grounding from the stock place at the thermostat housing then to the threads of the O2 low sensor COULD cause funky readings, weather or not that IS the problem, a 4 wire sensor wired direct, would eliminate that, and ensure a good signal.
That said, I am highly suspect of it being Decel ENLEANMENT.

[95GMC]

if you have 3 wire i don't 4 wire is really necessary. you have a tune problem not mechanical

Ok cool I didn't order anything yet so I won't worry about it. I was looking into the proportional gains and I'm trying to make sure I'm looking at the right tables, are they "proportional flow gain factor vs MAP vs RPM" and "proportional counts vs slow o2 error" ? If so what am I looking to do with those? Get more cross counts when the issues occur? How do I know whether to raise or lower the numbers? Sorry for all the questions I haven't messed around with any of this stuff before.

[tayto]

The thermostat ground IS NOT for the O2 sensor or the ECM. It is for the AC clutch! Grounds for ECM are on the back of the passenger side head. I have put headers on a few TBI vehicles and only used a 3 wire because it would drop out of closed loop at idle. Have also used a 3 wire on stock when cat removed or high flow cat used. The stock 1 wire o2 on a TBI GMT400 truck is after the Y pipe merge right before the cat. No issues with ground there FWIW. He is also running shorty headers......

[ralmo94]

The thermostat ground IS NOT for the O2 sensor or the ECM. It is for the AC clutch! Grounds for ECM are on the back of the passenger side head. I have put headers on a few TBI vehicles and only used a 3 wire because it would drop out of closed loop at idle. Have also used a 3 wire on stock when cat removed or high flow cat used. The stock 1 wire o2 on a TBI GMT400 truck is after the Y pipe merge right before the cat. No issues with ground there FWIW. He is also running shorty headers......

Maybe so, but my 94 stock had all sensor grounds at the t stat, including the one to the 02 low
All I was saying is that a direct wire from pcm to O2 makes sure of a good signal. Agreed that both 3 and 4 wire HEATED O2 is better than 1 wire non heated.

[tayto]

I suppose. they only went to 4 wire O2s on the newer OBD2 era PCMs. I have a feeling it was to do with noise more than a good ground signal, but I am not an electrical engineer. I also suspect the older TBI stuff was less susceptible to that sort of stuff. I have seen lots of that era GM with multiport with shielded pairs and whatnot, stuff you don't see on the older stuff.

I have torn apart many GM TBI harness and the ground was always on the back of the head. What colour wire was on the thermostat stud?

[ralmo94]

I suppose. they only went to 4 wire O2s on the newer OBD2 era PCMs. I have a feeling it was to do with noise more than a good ground signal, but I am not an electrical engineer. I also suspect the older TBI stuff was less susceptible to that sort of stuff. I have seen lots of that era GM with multiport with shielded pairs and whatnot, stuff you don't see on the older stuff.

I have torn apart many GM TBI harness and the ground was always on the back of the head. What colour wire was on the thermostat stud?

I don't remember now. I tapped in at the pcm for mine and ran a new signal high and low from sensor to pcm.
I think you may be right about noise. It's not really a ground per day, but the low signal. What was explained to me is that the pcm reads the difference between the high and the low.

And to be clear, I'm not sure you don't have a tuning issue, but my understanding has been that you always rule out mechanical issues first before trying to FIX the tune. A 4 wire tied directly to the pcm, Might not fix your issue, but it would rule out interference, or the mechanical aspect. Then once ruled out you can move on confidenty.

That's just my 2 cents.

[tayto]

I did a bit of research, looks like the P4 TBI stuff did indeed have the grounds on thermostat housing. most of the stuff i did was the C3 ECM , f-bodies, b-bodies and early C/K TBI stuff.

[ralmo94]

I did a bit of research, looks like the P4 TBI stuff did indeed have the grounds on thermostat housing. most of the stuff i did was the C3 ECM , f-bodies, b-bodies and early C/K TBI stuff.

I know some of the buy back stuff the grounds got moved to the head.

[95GMC]

Still battling this even with decel enleanment and DFCO completely disabled though I've sort of just been living with it by dumping a bunch of extra fuel in that area so it only has time to climb from say 11:1 to 14:1 before I'm back in the throttle so i at least don't get a bad surge from climbing to 17:1. Tried adjusting some of the proportional gains but I have no idea what I'm doing so it didn't seem to help much though I did try small adjustments and large adjustments just to see if I was being a nancy about it, didn't seem to have much of an effect either way.

Had another issue this morning which was the first cold morning I've had the truck out in a while. I've been having an issue with running lean when the ambient temperature is low so I went through the trouble of installing a MAT/IAT sensor, cut the flag on in my bin, and set the MAT/CTS blend to .500 but it didn't appear to help at all this morning. I'm wondering if the flags for MAT sensor high/low, code 23/25, have to be enabled for the ECU to use it in it's fuel calculations? I thought I had read of people just enabling the flag itself for MAT and not the high/low flags but I'm going to cut those on just in case. Yesterday when I drove the truck home all the trims looked great and didn't notice any lean areas. Should I drop this number lower to add more fuel when cold? It almost seems like the MAT input isn't working but I know the ECM sees it because I can get the data on my adx file and it's reading correctly. I don't see any other tables I can adjust to adjust fueling across the board at different ambient temps, just AE adjustments for different temps. Otherwise I would have just adjusted that table than go through the trouble of installing the MAT sensor.

[ralmo94]

Still battling this even with decel enleanment and DFCO completely disabled though I've sort of just been living with it by dumping a bunch of extra fuel in that area so it only has time to climb from say 11:1 to 14:1 before I'm back in the throttle so i at least don't get a bad surge from climbing to 17:1. Tried adjusting some of the proportional gains but I have no idea what I'm doing so it didn't seem to help much though I did try small adjustments and large adjustments just to see if I was being a nancy about it, didn't seem to have much of an effect either way.

Had another issue this morning which was the first cold morning I've had the truck out in a while. I've been having an issue with running lean when the ambient temperature is low so I went through the trouble of installing a MAT/IAT sensor, cut the flag on in my bin, and set the MAT/CTS blend to .500 but it didn't appear to help at all this morning. I'm wondering if the flags for MAT sensor high/low, code 23/25, have to be enabled for the ECU to use it in it's fuel calculations? I thought I had read of people just enabling the flag itself for MAT and not the high/low flags but I'm going to cut those on just in case. Yesterday when I drove the truck home all the trims looked great and didn't notice any lean areas. Should I drop this number lower to add more fuel when cold? It almost seems like the MAT input isn't working but I know the ECM sees it because I can get the data on my adx file and it's reading correctly. I don't see any other tables I can adjust to adjust fueling across the board at different ambient temps, just AE adjustments for different temps. Otherwise I would have just adjusted that table than go through the trouble of installing the MAT sensor.

Have you filed in your idle ve?with trimalyzer you can filter for idle flag and get it dialed in pretty easy only using data from when it's running off of the idle ve table. If you put an AFR error in your ADX, you can use that in trimalyzer and tune open loop. Since you have wideband, I would log a full open loop run and see how far off it is. The O2 only has a limited amount of time to figure out how far off it is.

[95GMC]

Have you filed in your idle ve?with trimalyzer you can filter for idle flag and get it dialed in pretty easy only using data from when it's running off of the idle ve table. If you put an AFR error in your ADX, you can use that in trimalyzer and tune open loop. Since you have wideband, I would log a full open loop run and see how far off it is. The O2 only has a limited amount of time to figure out how far off it is.

Just the sections where I'm actually at idle, pretty much just the 30-40 kpa range at 600-800 rpm. I haven't used trimalyzer, is it something in tunerpro or something I need to download? How do I go about putting an AFR error into my adx? I'll definitely do that, sounds like a better approach than what I've got going on right now

[tayto]

prop gains are only for closed loop, you are still running open loop correct? it sounds like you need to increase your VE table in those areas. has ASYNC been disbaled? I had a car that would do this weird jerking on decel, it was because async was coming on. Also just to make sure you have no EGR valve? charcoal purge line has been blocked (can hook up in final stages of tuning).

cold start still very much has to do with your VE table. get your IAT working if you wish but focus on tuning the VE table first before you start fussing with stuff like cold start. it is easy to get overwhelmed tuning.

[95GMC]

Well I had done my initial VE adjustments in open loop and thought I was pretty close minus the decel issue so I switched closed loop on to see if it would fix the decel issue but didn't make a difference. Also when I went back to closed loop my tune was lean across the board, now I've been fighting all that so I probably need to start over yet again. I just checked and ASYNC is enabled right now. That's right on the EGR valve it's gone and disabled. I haven't actually disconnected the charcoal purge but I will do that.

I think I explained the cold issue poorly, it is only ambient air temps changing that makes it run lean and not really a cold start. It can be say 30 degrees outside and the tune will be running lean across the board even after the engine is up to temp but then if I leave the tune the same and drive it in 60 degree weather it will run perfect and not lean at all. If I add fuel across the board to fix the lean issue at the colder ambient temp, when I drive again in a higher ambient temp it will be running super rich. I thought maybe the IAT would help with that but it hasn't seemed to make a difference so far. I definitely agree with easy to get overwhelmed tuning, I'm definitely there lol

[ralmo94]

Just the sections where I'm actually at idle, pretty much just the 30-40 kpa range at 600-800 rpm. I haven't used trimalyzer, is it something in tunerpro or something I need to download? How do I go about putting an AFR error into my adx? I'll definitely do that, sounds like a better approach than what I've got going on right now

Trimalyzer is a program for analyzing data logs a member on here, Steveo, wrote. It is pretty cool because you can use it to combine more than one data log and makes tuning ve a copy paste affair.

I wrote a how to write up on using it with AFR error, and included an ADX with AFR error in it. Since you said your wide band uses a different scale, Maybe you can either change the scale in my ADX, or look at it and see how I did it. Not with computer now.

Write up is here
http://www.gearhead-efi.com/Fuel-Injection/showthread.php?10776-Files-Included-How-To-VE-Tuneing-OPEN-LOOP-with-WIDEBAND

Basically you export your log file as a csv and filter stuff out. I think you may even be able to filter out canister purge. Not sure.
But would definitely look at using it, it's an Awesome program

[tayto]

https://www.thirdgen.org/forums/diy-prom/448152-choke-afr-crank-run.html

here's a good thread to read. guy is using MAF code but starting tables are the same. I would try what dimented24x7 said in the last post ”if youve changed your injector flow, the crank PW table will need to be scaled to maintain the correct crank fueling.”

so take your original injector flowrate and devide by the new flow rate: (example)
61 / 78 = 0.78

so now multiply that whole table by 0.78.

[95GMC]

Thanks for all the info guys, I'm going to get the Trimalyzer set up and start fresh on this sometime this weekend hopefully.

[95GMC]

Well so far using Trimalyzer is great I absolutely love how it's set up. It has helped a bit with my lean decel at highway speeds but I still can't seem to get the MAT/IAT sensor to influence fuel trims. I will log on my hour long drive home from work with ambient temp around 75 F and everything will be great and all trims within 5%. Then the next morning with ambient temp around 40 F, having changed nothing in the tune, I will notice lean stumbles and upon arriving at work and running the log through Trimalyzer it will be telling me to add 10-15% fuel across the board. If I add the fuel it wants I'll then find the truck running super rich in the afternoon with warmer ambient temps but it will then be ok on the next morning when it is cooler.

It seems fairly obvious to me that for whatever reason my MAT/IAT is not influencing fueling for whatever reason. I've tried changing the MAT & CTS Blend Multiplier vs Airflow table values with no luck. I even found a thread on this forum where a guy was having the same issue and simply turned on the MAT flag and put .750 or .800 everywhere in the blend table and never had the issue again. At this point I'm thinking of retrofitting a heat pipe to my right side header to get back to the exact factory setup unless I can figure out how to get the MAT sensor working. Has anybody ran across anything like this before?

[ralmo94]

Have you reviewed to see if it reported a different temp when lean compared to when rich?
What research I have done on the subject says tbi should use a MAT instead of an IAT, due to the fact that the fuel cools the manifold temps. Where is your sensor mounted?

Maybe someone else has more experience.

[myburb]

For years I have run my BB chev in open loop. To keep it happy with oat temp changes I have used the mat sensor ticked and the blend of .852. My sensor is at the base of a non stock air cleaner so your needs might be a bit different.

[tayto]

Carb/throttle body icing is a real thing where i live . everyone always deletes the stove pipe, as if it's a performance hindrance? it helps with cold startup drivability and gets the car up to operating Temp faster..if the system is operating properly i can't see how it affects performance? I haven't tried one but also heard there are good gains from the "salad bowl" mod. my TBI caprice had an IAT "bird cage" style that was in the air cleaner. i was always told the MAT sensor (same as a CTS) was not as good because it responded to changes slower. there was actually a bird cage style one you could use from a 6.5L diesel....

[ralmo94]

Carb/throttle body icing is a real thing where i live . everyone always deletes the stove pipe, as if it's a performance hindrance? it helps with cold startup drivability and gets the car up to operating Temp faster..if the system is operating properly i can't see how it affects performance? I haven't tried one but also heard there are good gains from the "salad bowl" mod. my TBI caprice had an IAT "bird cage" style that was in the air cleaner. i was always told the MAT sensor (same as a CTS) was not as good because it responded to changes slower. there was actually a bird cage style one you could use from a 6.5L diesel....

The argument about may vs iat was that as fuel is atomized it takes heat and can cool I don't imagine faster reaction time would matter that much, but I don't know.

[tayto]

op, i noticed you are running the edelbrock performer. i ran a vortec performer rpm which had provisions to run coolant under the carb/throttle body. vortec heads also have NO exhaust crossover. i am wondering if you are having issues until the manifold gets heat soaked? this may be a good candidate for the TBI MAF code if you are running a 4L60E or manual transmission.

EDIT: sorry to keep rambling here, but go look at the ultra expensive GM performance TBI vortec intake. it has coolant passages running under the TBI like a stock intake....

[95GMC]

Thanks for all the info everybody and I reall appreciate the ideas. I ran to the local pick n save near my work on lunch a couple days ago and grabbed a factory air cleaner and hose clamped a piece of pipe to the cylinder 2 header pipe and ran the flex hose from there to the air cleaner. I also had some more weirdness with the MAT sensor. When I drove home that afternoon after putting on the factory intake I realized I was still running on the chip with the MAT flag enabled and fault codes enabled and the MAT unplugged but it didn't throw any codes or affect fueling in any way which I though was odd because the MAT was reading -40 or whatever the minimum it reads is since it was unplugged.

The really crazy thing was the next morning I installed the newest chip I burnt with the MAT flag off and MAT codes disabled and all of a sudden I was running super lean in only one area of the VE table, almost like the MAT was working but only in one area? Very weird. Anyways that's only been 2 days ago so I've really only had a chance to drive the truck once but the factory air cleaner definitely has seemed to help. I had the MAT/IAT located in pretty much the same area as the factory intake flap on the factory air cleaner (I took the whole assembly apart and installed a mounting plate I made for the temp sensor). I attached a few pics of what I did and some things I noticed in regards to the temp at TBI unit vs the rest of the engine so I can definitely see what you're saying about wanting to have it set up as a MAT instead of IAT ralmo94. I may look at where I can mount the sensor closer to the TBI unit if it's still giving me some grief. I guess I just hadn't really anticipated this big of a fit out of such a mildly built engine but I guess it's still quite a bit different from stock. Also tayto I've eyeballed that intake several times but usually right before I break down and order it my wallet slaps me in the face and says keep working with what you have lol

17849

17850

17851

17852

17853

17854

17855

17856

17857

17858

17859

17860

[ralmo94]

Everything you shown looks good, but I feel it should be clarified in case not already understood, what you have is an IAT, or intake Air Temp, Not a MAT, or Manifold Air Temp. If you get the hiccups figured out it will probably work well. As stated previously ( I think someone mentioned it on this thread ) a MAT is threaded in to the intake manifold, and is the same part number as a coolant temp sensor. The difference is MAT would be able to sense temperature after fuel evaporates and atomizes, Wich lowered temp, IAT, picking temp up before the fuel has been injected can not. But if you get your IAT functional and calibrated it should work just fine for you

[95GMC]

Oh absolutely I understand the difference, I was trying not to confuse anybody that may be looking to this thread for help by going back and forth between IAT then talking about MAT flags and MAT blend table (I actually work on cars for a living, just not tuning obviously) but I see that just added even more confusion. At this point there's really nothing else I can do as far as the IAT because I can't get it to change fueling reliably no matter what I put into the blend table or even get it to flag a code with it unplugged, unless there's another table I'm missing. The temp the ECM reads via the IAT always seems to be similar to what I would expect on any other car, slightly below ambient temp when cruising and at ambient temp or slightly higher at idle due to some heat soak so it seems like the resistance curve is acceptable and the ECM is interpreting it otherwise I would think it would be showing some erroneous value. I'm definitely no programmer or anything so if the flags and tables in the tuning program don't work I'm pretty much clueless after that unfortunately, I don't know how to go in and actually see if the changes I'm trying are actually being implemented, though if I had to guess since the addition of the sensor does nothing and no faults set with it unplugged something isn't working right somewhere. I tried updating to the newest version of tunerpro, starting with a fresh .bin file, and copying numbers off of a BPLH bin but none of that worked, that's pretty much the extent of my knowledge on this type of stuff.