34blazer

having a weird issue trying to dial in the closed loop operation. datalog attached

first issue, while on the highway the blm will move but sometimes it will get stuck and not budge. since replacing the NB HEGO ive had to tweak the O2 constants up but im afraid to dial them in, since at times the BLM wont move and it starts to lean out too much. check out the datalog and you will see what its doing

second issue, my AE in open loop is fairly close some spots are well on the rich side but i can deal with that. as soon as it goes into closed loop, and when im in tip in conditions, something is pulling the fuel trim way out and my WBO2 sees 16 and 17:1. i dont want to bring up the AE because in open loop it will flood it out an stall it everytime. im wondering if this is the culprit?

Loop Closed Param, Gain Factor Multiplier for Positive Errors

after a few seconds after tip in it returns to somewhat normal AFRs, with a stick car it does it everytime i shift, while driving normally. sometimes it will even produce knock too.

id keep it in open loop but with me driving thru drastic elevation changes and temps, its too hard to keep a close tune on it. and my IAT/MAT circuit is buggered up so until i have the time to find another 7730 im stuck with this.

thanks!

EagleMark

I played your data log with adx provided and it was in Closed Loop, BLM learn enabled and blm was never stuck?

Did notice the adx said V4 only which would be a ads, not adx, don't know if that has anything to do with what your seeing, what version TP are you running?

Ronny

iTrader: (1)

Is the NB a heated one? If not do you have headers?

Same with me(stick) mechanical knock drivetrain related. This spring I am going to disable it. If I change my SA I will enable it to determine if table is too advanced.

Seems the collant compensation temp tables need some work. Enlean the AE when at lower temps and enrichen when it hits CL temps(>160dF)

34blazer

@ eaglemark, if you watch the O2 readings and cross counts they stay static for a period of time.

@ronny, im using alot of the AXXD bin since its also manual. the problem im having is, ive been tuning in open loop for awhile now, so everything is fairly close according to the WBO2. AE tip in vs. temp is consistant. open loop vs. closed loop AE is way off. im wondering why closed loop tip in is lean but when in open loop its a little on the rich side.

im not producing any knock while shifting, its when the tip in leans out to like 17:1.

RBob

iTrader: (1)

If the AE is on the rich side during closed loop the INT will drop and then start to drag the BLM down with it. As the AE runs out the AFR then leans out with the low INT & BLM.

The trick here is to lean the AE some so that the INT increases a smidgen (that's a technical term). This way it doesn't have the lean lag afterwards.

RBob.

Ronny

iTrader: (1)

Is the ECU using the WB values? Like substituting WB for NB? If so there are some ground issues ECU vs WB that can be looked over.

17/1 would be expected if coolant compensation if off or AE is off. What is the BLM in CL?

RBob beat me. Does he ever sleep?

Tuned Performance

iTrader: (92)

RBob beat me. Does he ever sleep?

It's likes He's a RoBoT

EagleMark

I see that now, BLM learn still on, OS still ready, still closed loop but Rich Lean goes RICH and stays there when O2 cross counts stop. So maybe what RBob said is causing this too?

You got a VATS fail and Error 54 locked in Error 54 fail and error 54 passed fail along with Vol Mem Bombed. Don't know if they apply to your car? I changed error data Item List View to include everything from Bitmasks in case there was somthing missing... HTH!

RBob

iTrader: (1)

"Klaatu barada nikto", from a great movie...

RBob.

34blazer

Ill get another log when driving in traffic to get a better idea of whats actually going on.

theres no VATS in the mustang, although i enable it if im in a sh*tty neighborhood so it doesnt start lol.
not sure why im getting fuel pump low voltage, i spliced the bypass from the OPSU into the original wiring and to the factory relay, which is under the seat in the stang. only difference is the relay is the ford 5 pin.

@ Bob, the AE in closed loop is actually lean, open loop its rich. or are you saying its rich and the INT is being too aggressive and causing the lean tip in?

*just looked in the bin and found the "INT reset while AE in work" bit wasnt set. could this be my issue?

@ronny, the WB is only patched into the datastream. no simulated NB. im using the factory ford HEGO harness with a bosch NB, which is grounded to the block. the WBO2 is has its own ground on a different location of the block. the SLC has a ground test the forces 2.35v iirc and that checked out.

thanks for the replies!!!

EagleMark

But only alien robots know what that means, there is no translation to human english... so his is a RoBot!

Day the Earth Stood still! Both versions...

34blazer

OK this sucks! BLM is still not moving when it hits a certain point. attached is a new log that shows whats going on. not sure why its doing this but its getting annoying! also riched up AE for lean tip in, didnt work! the ECM still pulls the fuel way back to the point of knock and 16-17:1 AFR. Im starting to think S_AUJP is flawed

Street Lethal

iTrader: (16)

... also from Army of Darkness.

34blazer

EagleMark

He almost got it right?

You almost got it right too! Read this again!

He said to lean AE. You made it richer.

34blazer

LOL is there an inside joke im missing here


Ill get another datalog up soon. i have the INT locked during AE. also, i said that AE is rich in OPEN LOOP and lean in closed loop. any suggestions for why the BLM freezes? i posted another datalog last night showing whats its doing. thanks

EagleMark

Yes a little inside joking...

But there was a serious answer to your question too! Re read RBobs post or quote I just did. He tells you to rich of AE will cause what is happening, although your data or your butt dyno is showing it needs more AE, your data is wrong because on INT trying to lean out overly rich AE. He said lean out AE, you said you added more AE.

34blazer

ok thanks! ill give it a try, theres another glitch somewhere preventing me from getting a good tune as well. im going to swing by the boneyard saturday and grab another ecm, i think the IAT circuit is goofed because its stuck on 66*F. i had another thread about it so i wont go into it here.

but id like to figure out why the BLM stops moving in a certain area, its around 42kpa and thats where i spend alot of time cruising on my way to work. im not sure why it stops trimming the fuel but its killing my gas milage. i looked thru everything i could find in the xdf that would cause this problem but couldnt find anything.

*edit* @ RBob, i could be a customer of yours very soon lol

34blazer

RBob

iTrader: (1)

If you can dump the log file as a .csv with a single header at the top, that would be helpful. Makes it much easier for others to look at it.

Nothing wrong with that!

RBob.

34blazer

Ok heres the csv of the tip in and cruise. ive included everything so whoever looks at it can see everything. if certain things are wanted let me know and ill manually populate the spreadsheet.

RBob

iTrader: (1)

Missed the attachment. May need to zip it?

RBob.

34blazer

sorry, internet is hit or miss here. just moved so im using whatever wireless i can "borrow" until tomorrow lol. lets try this again....

RBob

iTrader: (1)

Looking at the tip-in log, I found one that did pull the fuel trims down. Which by the NB value it should. Check the tip-in picture.

Cruising along at 38 MPH and the throttle is opened to 42% (column N). This adds AE, along with increasing the MAP and changing through several BLM cells.

The NB O2 sensor voltage is in the high 800's, which is rich. The INT (column G) responds by dropping, reducing the volume of fuel. This pulls down the BLM (column E). When the BLM cell in use changes the INT is reset to 128 (see Sample #94).

There are several columns of WB data. Of which I'm not sure which one to use. Although, they all show lean. Which is the opposite of the NB sensor.

I guess the big question is: does the engine lag from a lean area?

RBob.

34blazer

yes once the afrs get into the mid 15 afr range i can definitley feel the power drop way off, and i start getting some knock counts. the ford HEGO seems to produce more voltage than ive seen. I find it weird that its like that, the circuit has its own dedicated ground on the block, actually, im still using the factory HEGO harness that came in the car. Ive also replaced both the NBO2 and WBO2 and did the free air calibration on the WB just in case. ive had to adjust the O2 constant tables up to match the output of the NB, i figured since the code originally used a non heated unit, and im using a heated unit that there would be some sort of difference. but apparently not. what constant or table can i adjust to keep this from happening?

RBob

iTrader: (1)

Sounds like you should install a GM 3-wire (AFS-74) O2 sensor. I've seen the same in the BLM-stuck log file. The ECM is getting cross counts and is in the mid 800's (milli-volts) on the O2 sensor. The INT is at 128 which will have the BLM sitting still.

RBob.

34blazer

ok something must be wrong with the adx, well, the live display. at ~42kpa it shows no more cross counts, freezes wherever its at. and thats the sweet spot where i cruise.

ill swap out the HEGO for the GM one you suggested. currently im using the Bosch 13953.
http://www.autozone.com/autozone/par...81_168259_1784_

RBob

iTrader: (1)

Do you have a particular location (by Sample #) in the log?

I have the area of Samples 800 - 825 up. INT is solid at 131, BLM sitting at 126, MAP at 40 to 42 KPa, MPH at 77. With the INT only 3 points away from 128, it isn't enough for the BLM to move.

And x-counts are occurring. Although the O2 is bouncing in the mid-700 mvolts. You did mention that the O2 window tables were increased, so that makes sense.

RBob.

EagleMark

I just did a car that was in 800s at idle and would not go closed loop, as soon as you drove within 10 seconds it went closed loop and stayed there. Then let idle and stayed in closed loop as INT and BLM was getting richer it went closed loop again, revved engine closed loop, let idle got richer and went open loop. It must have been cooling the O2 (non heated). I leaned out idle and it stayed closed loop! No more 800s on O2!

Since it's getting stuck I don't think it's the adx file? Does it come back later? Or is it rich and cooling O2 sensor going open loop? If it's truely tuned to a sweet spot and 128 INT then no changes are being made to BLM. Or do you have highway lean cruise? My cross counts stop when highway lean cruise kicks in. This may be ECM dependent though...these were both older 160 baud ecm and both times showed closed loop when this happened.

My newer PCM in LT1 still shows closed loop when WOT locker is installed in bin file. At WOT INT and BLM are locked at 128, 02 readings are in 850 to 925 range, no more cross counts but still says Closed Loop!

O2 cross counts are differant the way displayed in ADX on differant ECMs 1227747 shows going up and down, zig zag on TP ADX Monitor. I looked at one for 1227730 IIRC?... the other day and it went from 0 to 255, then dropped to 0, back to 255. On TP monitor it was a slow slope, then drop... rinse repeat.

Do you have that data log in TunerPro xdl file? If so what adx. I'd be curious to look at it that way and compare to what I'm used to... maybe zip your xdl and adx.

34blazer

im not too familiar with the monitors, ill try to interpret them a little better. however, i do understand that the INT/BLM can pause when no trim is required. thats not the problem im seeing, the problem is, at ~42kpa, constant throttle and speed, the O2 cross counts stop counting and according to the WBO2 the afr slowly drops into the low 13 AFR range. thats why im saying the blm is stuck, because the swing points in the table vs. whats being diplayed in the adx dash, is drastically different. the NBO2 volts will hang in the 840ish mv range, no cross counts, no blm movement to lean it out.

@eaglemark, im not able to upload my .xdl and .adx from the govputer but will do so when i get home. but if you want to see it my first post had a log with the .adx, iirc it shows whats happening. but i will post the newest log in a few hours.

thanks fellas

Ronny

iTrader: (1)

Possibility it is moving to OL due to 02 sensor cooling off? Might explain the going rich.
Occurred to me B4 I had the heated 02 sensor. Steady cruise no load NB cooled off. My car dropped out of overdrive at that point.

34blazer

that could be, but the CL indictor remains on the entire time. also, wouldnt the O2 volts be stuck at ~.450 mv?

Ronny

iTrader: (1)

I was going to ask about the CL indicator. It would appear you are in CL. Was uncertain if you mentioned that in posts made. When I am in OL I recall the 02 stops switching but NB02 shows changing voltage. I dont believe it moves to .45.

When you do cold start with datalog up and runnin check for OL indicator and whether you are reading volts at .8++ or .450.

EagleMark

Ronny just confirmed the same thing I was thinking/explaining above!

I missed a post, was same post time as RBob and he said he see's cross counts so that's what I was going to look for. Also suggested GM heated O2 so maybe back to stock O2 settings as well?

Where is this heated O2 mounted in exhaust? If heated it should not cool off even when that rich, unless it's really far back?

EagleMark

I gave examples of that above on differant ECM/PCM Some do some don't. I'll look at your log again...

JP86SS

iTrader: (1)

Double check what values you have in the O2 tables.
There is a spreadsheet in the package for V4 that you can enter your values. If you have them crossing each other or something goofy there it could cause a spot where thing don't work right.

I do believe the CL indicator could stay on and "Learn" be disabled.

JP86SS

iTrader: (1)

...and you're sure its not going into PE at 95 Kpa and 42% throttle?
Check those tables (normal and high coolant) and be sure you are not too close.

OK, that was just one segment of the log but figured I'd throw it out there.

34blazer

im hoping a GM sensor should cure these problems, i wasnt aware that different NBO2s, OBDI mind you, would be so drastically different. then again, id rather it be this than a problem in the factory HEGO harness lol.

im using a bosch 3 wire unit, OEM for 5.0 mustangs. Its sitting in the factory location, about 6 inchs back from the collector. i do have an off road h-pipe tho, but during cold startup i can tell the sensor is heating up because it starts to move within a few seconds after startup. ill get a cold start log in the morning if possible, but at the very least find out what the cold voltage is. got tasked to work late tonight and i go home for leave about noon tomorrow. thanks!

34blazer

thats what im thinking as well, CL is on but learn is off. i checked those constants to see if i could find something goofed with them as well. looks like theres the 3 tables and 3 constants? i didnt know i had to adjust those constants, ill plug it all in when i get home.

34blazer

wells heres the latest datalog .xdl, .xdf, and .adx. let me know what you think

JP86SS

iTrader: (1)

See if your problem lines up with how the O2 tables are entered.
They should not cross as you have them.
Also, having them that high is not helping your milage(unless that motor really needs that)
Now that its running ok, you may want to bring them back down to stock (or 50 higher) and try again with the VE tables.
Also your CL enable Hot threshold is set for 70 C. Seems kind of high and you could bring that down to 35 C and see if the BLM starts again.

EagleMark

JP86SS seems to have figured out some issues. Going back to stock settings would be a good start. Using a GM O2 with GM ECM would be good although I beleive Furd NB reads the same it's not something I would want to try or rule out while tuning. May be curious to test if car was tuned or stock engine... but not now...

I made a new item list named BLM Stuck in your adx for things I wanted to look at. First it has a reading from SIM NB LC-1 that is changing? If your using simulated NB all bets are off.

You O2 Filtered reading is always high in 700 to 800 range and not showing cross counts? Maybe that is what this O2 Filtered is for in this mask or the Super Version? Not what I am used to seeing in O2 voltage with cross counts.

Only time I saw cross counts stop was when Decel, In AE MAP or In AE TPS was active and vacuum shows to drop there to so seems to be working fine.

O2 Ready, Closed Loop, Learn Enabled was on the entire time with Rich/lean flip flopping so no indication of going cold/open loop. So back to first paragraph...

HTH!

34blazer

Alright fellas, since ive gotten off leave ive had time to figure this bugger out. Figured some things out. first of all, the filtered NBO2 values seem to be reading way off. i started to lower the O2 constants tables and everything started to fall into place nicely. the AE problem i thought i was having....totally my fault. the AE was a bit rich but still went lean. heres what was happening, during the period of time i was using open loop only, i had a weak fuel pump. initially i thought the open loop vs. map/coolant temp tables needed to be fattened up. it was lean during moderate and heavy loads when no in PE. so i added to those tables to richen things up. i fined tuned the VE tables, and reduced much of the table. since i replaced the fuel pump i forgot to return those tables to where they were at originally. so when i started to tune in closed loop the BLMs were high. before i wanted to mess with the VE tables i wanted to get the O2 constants figured out. but didnt realize that because i didnt add to the VE tables the ecm was playing ketchup most of the time. it would lean out during tip in and start adding fuel in but it lagged severly. it took lots of thinking to figure this out. so the other day i started adding to the VE tables and now the leaning problem is fixed, the ecm can now keep up with my driving. the O2 constants tables are getting close. floating in the 600 mV range now. weird thing is, the original values were higher in the low load range and leaner in the higher load range. my tables are opposite of that, oh well. fuel milage is way better and almost no leaning problems anymore.

that being said, since ive got it closer to being dialed in, i notice quite a bit more surging(bigger cam). im wondering what Prop Gain tables to adjust for light loads, it does it bad in the 30kpa range, not too bad in the 40kpa range. theres 4 tables im looking at now, just wondering which one is the best to play with? thanks everyone!

34blazer

ive been trying to figure out how to calculate the engines flow using gm/sec. im doing something wrong because none of my calcs are right. i need to adjust the prop gain vs. airflow table because of the cam in using, its jumpy at low load/rpm engine speeds and still a little laggy at highway rpms, about 50- 60 kpa.

im reading RBobs tuning paper and i cant seem to figure out what the constant is, been searching for the last hour. im horrible at learning shtuff like this lol.

im using RPM*MAP*Constant= gm/sec

so 2000* 40*?=?

34blazer

now im finding formulas that use:

rpm*map*iat air mass*engine disp(L)=g/sec but im getting large numbers that dont make sense to me.

Ronny

iTrader: (1)

What datalogging sostware? I would suspect TP-RT spits that out for you?

RBob

iTrader: (1)

Just put the airflow into the data stream. It is RAM location $6B

As for the large value you are coming up with, odds are it needs to be divided by 256.

RBob.

34blazer

ill look into that, didnt realize the ecm calculates that already. dividing by 256 makes sense now

Dominic Sorresso

3.4,

I had an issue last year where I was in C/L but my BLMs were frozen at 128. Turned out that my AFR v CTS table had a value ( at operating C) that was lower than my AFR Bias value in the constants. Now this is on an $D0A mask. Check that if there's anything like that in ur cal. Drove me nuts for a few weeks. Thank goodness for my Ostrich. Basically, played a game of battleship to figure it out.

34blazer

ok fellas got another question. first of all, thanks for the replies!

@Dom, i didnt see any of the tables you mentioned, maybe im not looking hard enough lol.

@Bob, still need to add the gm/sec into the datastream once i figure that out lol.

so i was reading the stickies (the O2 constants one) and saw Bob or Bruce mention tightening up the tables if not running cats. has anyone tried this? the separation from the upper window to the lower is ~80 mV. can i tighten them up to like 40mV?