Tootie Pang

iTrader: (2)

Stock 1989 305 TPI with TH700R4

Here's what been changed:

Every sensor and IAC.
ECM
Complete Ignition- coil, cap, wires, plugs, pickup, module
Timing set, damper
Valve stem seals and valve springs
Water pump, radiator, cooling hoses
Injectors with Bosch D3
Fuel Filter
Recent Fuel Pump
Vacuum lines
Intake plenum/runner/tb gaskets
Air filters

Cleaned/checked:
Throttle body cleaned and idle set
EGR
PCV
Tank Evap
Fuel system 37-47 psi on 15 minute drive around test, 47 psi held steady at part and full throttle, no bleed down, regulator good, pump and lines seem good.
Timing base 8 deg
No vacuum leaks- spray test and vac reading, 17-24 in
IAC set
Idles at 600 RPM, does not hunt
BLMs are around 128 but then sometimes move way up. Does not correlate timing-wise with stumble.

Runs near perfect, no codes, seems to to have full power.

Problem - stumble, missing and longer crank time to start (did not used to do this)

When:

Stumbles mildly occasionally warming up and warm, at both idle and part and full throttle acceleration. At full acceleration, when it stumbles, you can feel and hear the engine kind of go flat and then it comes back a second later.

Idle. Idles nice for about 10 seconds, 600 RPM, then starts with the stumble and idle drops further to just above 500. Seems to jump around a little between 500 and 500 when missing.

When it does not stumble, it is buttery smooth.

Typical pattern:

Normal driving: smooth. Coming to a stop, idles down to 600, very smooth. Holds smooth for about 10 seconds, then "bump", "bump". about one every 3-4 seconds, seems to gain in frequency to about every second the longer I idle.

Acceleration- once stumbling at idle, can stumble a bit on acceleration. When it doesn't it is smooth acceleration.

Other things noted:

1. When I checked my timing with the ICM enabled, it looked to be around 20 at idle, and then when unplugged, 8 degrees was the base. However, when I reconnected the plug, the timing did not return to being advanced. It stayed at 8. I had to shut off and restart the engine and then it ran normally.
2. I changed the knock sensor a while back, could the wrong knock sensor cause this? I have not checked my knock count and have no codes.
3. Car takes about 7 seconds of cranking to start. Even after priming the fuel pump and rails.

Scanners:

I have an OTC 4000 scanner and an ALDL bluetoothscanner from 1320electronics.com is going in next week.

Tootie Pang

iTrader: (2)

Drove around yesterday.

INT stayed between 130 and 124

BLM was 150.

Tootie Pang

iTrader: (2)

Got a Code 45.

Tootie Pang

iTrader: (2)

More clues:

So I was goofing around trying to get my new ALDL Bluetooth working with ALDLDroid and the car was idling for a good amount of time. Definitely warmed up.

After not having much luck with the ALDL, I decided to swap out the ECM with the one I had replaced a month or two back thinking it was causing the trouble with the hunting idle and poor running.

It had been sitting in a box since then so it was completely reset.

I fired it up and the car ran like brand new. I was just beside myself. Then, after about 20 minutes of cruising and punching it and punching it and cruising, I was like "man this was it, that 87 Trans AM ECM must not have been perfectly compatible".

A few minutes after that, it started again. Idle got really low, stumble, stumble on acceleration.

Now that I think about it, I noticed, that once that started, my tach was jumping with my turn signal.

I wonder if my alternator is doing something funky. Well, I'm going to keep looking.

I need to get that ALDL working so I can log all of this stuff.

But boy was I beaming for a while.

20 minutes tells me a few things:

1. Maybe the BLM is done learning and it starts doing something weird.
2. Maybe my alternator is doing something funny. It was new on the car when I got the car but its likely a cheapo Duralast or something. Maybe the voltage regulator? Come to think of it, the battery was kinda new too. Also crappy. Failing battery could surely bring things down.

Hmmm. I'm really starting to think alternator regulator.

Tootie Pang

iTrader: (2)

Update of the update:

I just got back from an early morning drive. Cool air, no headlights, light electrical load. Car ran near perfect. Except at low idle after the car completely warmed up and driven for a good while. I noticed just putting the turn signal on caused the stumble to get noticeably worse and my gauges would move with the turn signal click. I also noticed the white sticker label on the alternator looked "toasted"- burnt brown. If I put it into neutral or park while idling the problem want away when the idle went up 100-200 rpm. I am now thinking this whole thing is a crappy alternator. Of course I talk myself into these things all the time. That's how I've ended up replacing every dang thing practically. New Delco alternator and high quality battery going in today.

Tootie Pang

iTrader: (2)

Well, voltage is higher across the board and at low idle. Car starts quicker. But stumble is still there. I don't think it's a closed loop thing. It can stumble cold, warm, at idle, or under power, on acceleration from a stop, or acceleration in a held gear. It is intermittent but once it starts it stays. It may not happen for a whole ride. Or it may happen for the whole ride.

I am still trying to get my ALDL system working. Once that's in place I'll see if there's anything funky according to the ECM.

Tootie Pang

iTrader: (2)

...deleted...

Tootie Pang

iTrader: (2)

Got my ALDL Bluetooth working with ALDLDroid. Super cool!

My BLMs are way too high. 158 solid. Idle or steady driving. Dips down on acceleration but then comes back up. INT's are fine. My data file is attached.

The code 33 is from when I was goofing around with the MAF and I had not reset the code.

Tootie Pang

iTrader: (2)

Found this excellent explanantion from "Injuneer" Fred from New Jersey in 2008. The link doesn't work anymore unfortunately.

------------------------
The ECM calculates the fuel injector pulse width on the assumption that it knows all about your engine... displacement, injector size, mass air fow, etc. It uses an equation that includes the terms:

..... x (BLM/128) x (INT/128).......

The BLM is the long term fuel correction. This is what is stored in the ECM in volatile memory. The INT is the short term fuel correction, and its not "stored". The ECM recalculates the short term correction for every cycle of the calculation - about 9 times per second. It is intentionally forcing the closed loop A/F ratio to swing back and forth between very slightly rich (14.65:1) and very slightly lean (14.75:1). It has to do that for the cats to work.... when its lean the cat stores the excess oxygen, when its rich the stored oxygen combines with the unburned hydrocarbons (HC) and carbon monoxide (CO).

If the ECM finds that the long term average of the INT's is adding fuel (over 128), it ratchets up the BLM's to try and get the average of the INT's back to 128. It stores those values in the volatile memory..... that's most of what the ECM "learns". If the BLM's are above 128, the ECM is responding to what it BELIEVES is a lean condition. IF the BLM's are below 128, its responding to what it BELIEVES is a rich condition.

The BLM's are developed during part throttle/part load conditions, when the ECM is operating in closed loop (O2's are being used to modify the fuel rate). They are not learned when you are running at WOT, so they are not an indication that the engine is leaning out at peak power output. They are used at WOT to calculate the fuel rate (but only if they are above 128). A BLM above 128 indicates that something in the fuel system is not responding the way the ECM thought it would.

For example, if your fuel pressure is too low, the injectors don't flow the amount of fuel the ECM is expecting, it runs lean, and the O2 feedback elevates the BLM's above 128. If you had a vacuum leak, same result. If you have an MAF sensor and its dirty and reporting less mass air flow than is actually entering the engine, its a lean condition. If you have a speed-density setup, and have modified the intake and exhaust to improve the engine's breathing, but have not had the programming modified with changes to the volumetric efficiency (VE) tables, it would run lean. These are all "TRUE" lean conditions.

But sometimes, the O2 sensors give the incorrect feedback, indicating a "FALSE" lean condition. The main examples of this would be 1) faulty O2 sensor(s), 2) misfires, 3) exhaust leaks before the O2 sensors. All of those trick the ECM into thinking the engine is running lean, and it adds a bunch of extra fuel, using the BLM's, that the engine does not need.

I have this online writeup that includes the A/F management aspects of the LT1 engine management system, but its pretty much the same for the 3rd Gen fuel injection systems:

http://members.aol.c...ZZ/ScanMast.htm

Tootie Pang

iTrader: (2)