Issue with hesitation on cold start

[babywag]

Voltage dropping after a while isn't really a concern. Many alternators reduce charge somewhat as they heat soak.
Long as it's in the 13.6-14.6 range you're okay. I also wouldn't be concerned with the difference from ecm battery voltage data.
It was previously really low now it seems you have resolved that issue.

Where is the o2 sensor mounted? Data reported suggests it's dropping out sometimes possibly cooling off? Or bad wiring/sensor.
Many very low readings reported in the data log.

Still looks a little lean. What's your fuel pressure?

[Theodore]

Hey, babywag, good deal on the battery voltage - that was something else to find/fix. Looks like the hacked up harness was the culprit there. Am looking to get a new EFI harness from AFI, just wanted to get it running right first, so i wasn't changing too many things at once.

re: fuel pressure - reads 12 steady. Have installed a schrader valve in the supply line coming in w/what's become a permanent 'temporary' fuel pressure gauge.

re: O2 sensor mounting location: have included a picture for reference.


Also, re: O2 sensor dropping off - i grabbed a 30-minute data log (attached) this morning, with the Jeep idling in the driveway, pulled the O2 sensor data & plotted it on a chart over time, and saw an interesting pattern - that shows what you're talking about. i would've thought there'd be more points clustered around the middle vs. way out on the edges.



Am heading outside now to trace the O2 sensor wire & see what i have. If it's anything like the rest, the wire probably changes colors twice before making it to the ECM.

[Theodore]

am back from taking a look at the O2 sensor mounting location & getting a better picture; as well as, checking the connections on both ends. Connections look to be good. Also, started & ran the engine a few minutes to compare data log readings vs. multimeter. The cycle speed of the multimeter shows it to be even leaner still at idle in Open Loop. Once Closed Loop, i saw quite a bit of cycling above .5V, that wasn't evident in Open Loop. Hm. With fuel pressure being ok, we back to finding a vacuum leak?

The O2 sensor is mounted 9 1/2 inches below where the exhaust pipe mates to the exhaust manifold. Apologize for the bad light, but i can affirm that the end of the tape measure is against the end of the exhaust pipe.

[babywag]

If you look @ the log a couple things jump out @ you.
When cold it is running lean or the o2 sensor isn't working properly, mV should be much higher.
Still seeing a LOT of o2 mV readings in the single and low double digits.
A properly working o2 will swing wildly up/down but in my opinion you should be seeing higher swings, and seeing far to many really low mV swings.

Line 190 is starts closed loop, and the o2 mV first few data points 52.08, 4.34, 0, then seeing random double digits occasionally.

Start with the simple, does your engine have a ground strap?
Is the o2 mounting surface clean? Try and measure engine grounds and o2 ground to see if there is a lot of resistance.
Also verify NO exhaust leaks!!!

Your BLMs in the log continually rise and climb into the 150s.

[Theodore]

Hey, babywag, et.al., was hoping to have this data pulled/uploaded yesterday, but i ran into a snag - I ran the Jeep out of gas during the test... The gas gauge is off by about 1/4 tank - meaning, when it shows empty, i can only add 3/4 tank worth of gas, so i misjudged a little - ugh.

Anyway, after doing the gas can thing last evening, I decided to get out early before work this morning & follow your advice & remove/check/clean the O2 sensor, and its mating surface(ground). So far, i've only run it at idle in the driveway, but it appears to have had a positive effect.
Cleaning involved: using DeoxIT D100 electrical contact cleaner and 220-grit sandpaper on the threads/mating surfaces.

Before cleaning the O2 sensor was sooty & rusty.


And, so was the bung. Apologies, I don't have after pix.


Charting the O2 sensor data - before cleaning (CLOSED loop occurs at 190 on the X-axis). Average over the time period is 510, with a larger gap between lows & highs vs. AFTER, while the average of CLOSED loop values is 496.92; Min=0, Max=802.9.


vs. O2 sensor data - after cleaning (CLOSED loop occurs at 205 on the X-axis). Average over the time period is 437, with a smaller gap between lows & highs vs. BEFORE, while the average CLOSED loop values is 516.14; Min=65.1, Max=950.46.


Overall, based on the chart, things look better. Can't wait to make a test run tonite & see how it runs & what the logs look like. It seems like we're getting there.

Also, I didn't have time to look at the engine to frame ground strap. Have seen comments that it's located near the passenger side motor mount. Given the Frankenstein nature of this vehicle, my money's on there not being one.

This morning's data log is attached - Cold Start Idle to Hot - run time ~8.5 minutes.

More later...

Thanks.

[babywag]

Hard to really garner a ton from just a short idle log, but making progress.
Are you confident no vacuum leak(s) and/or exhaust leak(s)?
ECM is adding a bunch of fuel according to BLM/INT...
BLM @ end log still pretty high, IAC actually went to 0 counts then recovered.
Idle speed seems erratic and little high?

Take it for a drive warmed up, and closed loop and get some good data if possible.
~20minutes of real world driving is best.

[Theodore]

apologize for the delay - life got in the way. Finally had a chance to take a 30-minute test drive & pull a data log. Running-wise, i couldn't tell any difference: engine died when first put into gear, hesitated when accelerating away from house, and back-firing at speed on the Interstate, when rolling at 70, let off, coast to 60 & get back in it.

I did have a chance to look at grounds, and as i'd suspected, there's not one i can find between engine & frame. There's one between battery & A/C bracket on block, and one to passenger fender. I've ordered a new set of battery cables & ground strap to enable replacing what i have, ridding ourselves of a few more splices, while adding block to frame ground. Will pull another data log once i've got that installed.

You'd asked about vacuum/exhaust leaks:
- re: Vacuum: I tend to think i'm good here. Have used 2 1/2-3 cans of Carb cleaner, plus a home-built smoke machine to find/fix 10-12 leaks to date. With that said, all vacuum lines have been replaced, as well as, TBI rebuilt. TBI gasket & Intake gasket have not been replaced. Min.idle, IAC & TPS set up procedure has been done in tandem w/OBD1 scanner.
- re: Exhaust: I don't hear any ticking, and see no evidence of an exhaust leak.

For interest, i plotted scatter charts for the main sensors, as well as, uploaded the latest data log they go with - CLOSED loop occurs at data point 214.

IAC:


Battery Voltage - ranges from 13.7-14.1


O2 Sensor:


Block Learn Multiplier:


Integrator:


TPS:


MAF:

[babywag]

backfiring out of throttle body or exhaust?
if you mash the pedal when it's cold/hesitating can you produce a backfire out throttle body?
most stockish AMC engines using stockish GM programming are lean @idle/off idle.

It is adding a LOT of fuel in idle/off idle areas. What injectors are installed? are they 5.7 injectors or possibly 5.0 injectors?
Something possibly a mismatch actual hardware(injectors) vs. .bin file parameters or programming isn't great?

Possibly the o2 sensor isn't functioning ideally, or there could be a small leak upstream of it
Even a tiny leak sucks in air, o2 picks this up, commands MORE fuel, ecm stuck in fruitless attempt to manage fuel/afr due to inaccurate reading.
When the "sample" isn't accurate, the ecm can't adjust properly because it expects "x" to happen but "y" happens.

No VSS so really hard to get a picture of what is happening. Still seeing a large chunk of TPS @/around idle in log.

Just my opinion...if it was mine, I have always either installed new injectors, or had a used pair cleaned with flow data to input into .bin
My 88 is 14psi, my 90 is 15psi. The stock regulators seems to be 10-12psi. My opinion is fuel delivery/flow accuracy appears to be more consistent a tad higher psi than stock.

This thing controlling timing? What vacuum devices are hooked up? Charcoal can/EGR/??

[Theodore]

Re: backfiring: is out the exhaust. Have never had it backfire thru the throttle body - is only at highway speed when in throttle, let off & coast to 55-60, and get back in it - it immediately starts - there's a push/pull type feel, sounds like popcorn popping behind you.

Re: injectors: not sure on that. Will have to look into it. Any pointers as to how to tell what i have?

Re: computer-controlled timing?: no, from what I understand, this is one of the early AFI systems, circa 2006, that uses a hybrid AMC distributor stalk, with a GM cap w/ignition control module built in. Uses a separate coil.

Re: vacuum connections: PCV, MAF, are all that are coming off the throttle body - other ports are cappped off. EGR isn't connected, vacuum port is capped; Charcoal canister is gone, Air pump is missing as well; vacuum switch that controls the EGR - that installs next to Coolant Temperature Sensor is missing/capped. Otherwise, have traced/replaced all vacuum lines, starting w/PCV, MAF on throttle body & fanning out from there. Saw quite a bit of improvement in how the engine ran.

Given the age of the system, i don't know what the .bin has defined. Have been wondering about what it takes to read the chip vs. forum research & making changes based on it. Am not really interested in burning chips myself, but would love to know how things are supposed to be set, based on the programming.

Also, i've thought about taking it to a shop & having a full vacuum test run (intake & exhaust), and see about putting the leak question to bed. Given the conversations around running 'lean', it seems it's a question of not enough fuel, or too much air. Will wait to see what we get w/installation of battery cables & engine to frame ground strap.

[babywag]

sounds like timing control is enabled from description.
post a pic of the distributor...

injectors are usually color coded and have part numbers stamped into top(very hard to read).

What 4 digit code is on ecm/chip if any will give an idea of stock programming.
Can't speak for AFI but have read a few Howell chips and they were almost identical to stock .bin very little changed.

[Theodore]

sounds like timing control is enabled from description.
post a pic of the distributor...

injectors are usually color coded and have part numbers stamped into top(very hard to read).

What 4 digit code is on ecm/chip if any will give an idea of stock programming.
Can't speak for AFI but have read a few Howell chips and they were almost identical to stock .bin very little changed.

Here's the distributor:


And a pic of the ECM, showing 1227747 & AKMH. ? - is this the code you're looking for, or do i need to take the access panel off & check there?


I'll take a look at pulling/checking injectors later on today. Work day has begun.

[babywag]

That system/distributor is controlling timing.

AKMH is unfamiliar to me some AKxx chips are 4.3 v6 bins so probably not much help there.

[Theodore]

That system/distributor is controlling timing.

AKMH is unfamiliar to me some AKxx chips are 4.3 v6 bins so probably not much help there.

Maybe i'm confused: i've heard the term 'computer controlled timing', which in my mind, says the ECM is controlling the timing vs. what i'm thinking here is the ICM (Ignition Control MOdule) is controlling the timing, true? or, are both versions 'computer controlled timing', since there's no vacuum advance?

Later on, when i check the injectors, i'll pull the cover to the ECM & see if there's any markings on the chip itself, that may dispute the AKMH tag.

Thanks for the help - i feel like i'm heading down a rabbit hole - not sure how much further it is to bottom.

[babywag]

The computer controls the timing...

An ignition module transmits/receives signals and fires the coil.
A module alone cannot "control" timing.
It has a sorta "limp" mode that just dumps like 20° advance under certain malfunctions.

I would ask AFI for more details on your chip based on their tag#.
You said they told you it wasn't timing controlled which it clearly should be. What did the "tech" base that from?
If your chip doesn't have timing control enabled it will never run right.

[Theodore]

got an update re: injectors: GM5235206*RPD /7311 GM

Saw this over at GearHead-efi: http://www.gearhead-efi.com/Fuel-Inject ... rt-numbers that says these are for 5.7L/350 engines flowing 61 pph.

Numbers are hard to read in pic, took about 12-15 attempts to get it in a readable form.

[Theodore]

Thanks for playing along - to say this stuff is new to me, is an understatement. Reading/learning more everyday. Have read quite a bit that EFI systems are susceptible to flakiness due to bad electrical connections - i suspect a lot of the issues i've seen to date have been due to a large number of bad splices & the voltage drop associated with them, not to mention, a myriad of vacuum leaks i've found/fixed to date.

When i started this journey last summer, ECM reported battery voltage was ~12.1-12.5V. Today, it ranges 13.8-14.2V. Am planning on replacing battery cables & adding an engine to frame ground strap this weekend pulling another log, and going from there. Am thinking a good next step is having a formal vacuum test run to ensure i have no leaks in intake or exhaust, then proceed to looking at pulling the .bin from the ECM and looking for mismatches, like the EGR's enabled, but doesn't exist, wrong injectors defined, etc. Make sense?

[babywag]

You may want to pull/send injectors for cleaning and get flow data?
Can also buy new but kinda pricey imho.
2006 if accurate is quite a while ago and who knows what yours are doing @this point.

I would also ask AFI for a copy of the .bin they used?

[Theodore]

Making progress: Today, I swapped out battery cables & the ground strap. Earlier, i'd mentioned that i didn't believe i had a block to frame ground, turns out i had one, but it was cracked/corroded & terminated on a painted metal surface on the frame - so i wonder how much good it was doing. After sanding paint off to bare metal around the mounting bolt, and finishing the installation, we now have the highest voltage ever reported by the ECM, 14.5 (+.4V). At this point, we're up 2.3V from where we started.

Took the Jeep out for a test run, butt dyno says it's running better; am still seeing similar symptoms, though they aren't nearly as bad as before. Am planning on making another test run tomorrow without having had the battery disconnected & see what we have.

First of the week, am planning on having a smoke/vacuum test run first of the week to ensure no leaks, & contacting AFI to get a copy of the original .bin.

Attached today's data log, along with a few charts of key metrics.

RPM - noticed a roaming idle for 30 seconds at cold start, but it may be due to the battery cables having been disconnected. Am hoping this improves the more it's driven.


Battery Voltage (up .3-.4V from before)


IAC


O2 Sensor still shows quite a few measures below 100.


BLM (down a couple points from last run:135.83


INT


TPS


MAP

[babywag]

Still needs more fuel @idle and low rpm...can you adjust the fpr on it?

[Theodore]

Hey, Babywag, just got home from passing a smoke test; no vacuum or exhaust leaks.

As far as I know, I have no way to adjust fuel pressure. In an earlier post, youd mentioned having injectors cleaned & flowbenched; am looking at sending them off to MrInjector, as well as pulling the bin file to compare whats there vs what I have. Wondering if I have a mismatch between the tune & reality...?