Any GM OBD1 ECU experts out there?

Or anyone who has used them?  My LZ 3.9 V6 has one, schematic I found in the car says 1227703 ECM.  All the info I have.  Anyway, it started running like crap so I put an Innovate wideband sensor and gage on an extra bung.  It showed way lean at idle, like 20:1.  I did figure out the O2 plug for the ECU had a bent pin, it is still too lean but fattens up when I hit the gas.  Anyway, it looks like the sensor or something died, now it is back to 20:1.  I have a replacement sensor on order but remembered the innovate has 2 voltage outputs for an ECU, one looks like a narrow band voltage and AFR range.  The schematic also has an optional wideband input wire.  SO I am tempted to try one of these, but would prefer not to kill the ECU.  At least for now, a microsquirt is probably in the car's future as this current system is just too crude and the harness wiring is a complete unknown.  From what I have seen it could be the source of many future problems.

Those OBD1 systems are not very smart. 

It takes something really wrong to give any trouble. You don't really need a scanner or anything to work on them.

If it's reading lean like that and running like crap, first thing I would do is check fuel pressure. You probably have a plugged filter or failing pump.

If you have a scanner plugged in, have a look at your fuel trims. If they're maxed out and it's showing lean on the O2 than the system is trying to compensate but not getting enough fuel. 

That seems like a hyper specific ECM. From a pontiac lemans and some saturns? And it's running a v6? Weird. I would imagine it doesn't have a stock prom in it?

Are you asking about using the innovate as the narrowband for the ECM as well? I've heard of this being done when I messed around with 3rdgen fbodies long ago but never actually seen it work in practice.

Is there a CEL hooked up or obd port to pull info?

Poke around on Gearhead-EFI forum for GM systems

Like modern 3900 on an old school ecm? Why did someone do this and what is it in? 

Fuel pressure is 50 -60, changed to as low as 40, then it just ran really lean on acceleration.  I figured the bank fire injectors were running out of flow.  The age could be wrong, can hook up another to the fuel rail. There is an ECU port but I don't have a connector or software to look inside.  The connector only has 3 wires on contacts so I have a feeling data would be limited.   Since it is a stupid ECU I am not sure if there will be much more info than what my wideband shows.  Yes it is some sort of custom eprom.  But I have nothing but a schematic and the PO is not answering any questions.

What 3.9 v6 is it running?

That was originally a TBI 4banger ECM. Not that you couldn't repin/run a custom PROM but if the PO won't give you any info on it it's going to be hard to muddle through it yourself.

Does this car have a functional check engine light and an ALDL port?  If so you can check codes with a paperclip.  https://www.apexautoaz.com/gm-obd-1-codes.html

Also, don't forget that a missfire looks like a lean condition to an O2 sensor so the problem could be ignition.

i'd strongly consider figuring out a 3900 pcm that's compatible with hptuners and just making a standalone harness with the newer computer and then you get legit tuning capabilities.

I'm still confused about this. Are you sure you got that part number correct? Do you have pictures of the ECM? I don't know how an obd1 ecm can run that engine.

The Saturn forums have a lot of info on these ECUs. Good luck!

https://forum.sixthsphere.com

Backround. It is a 77 MGB with a pontiac LZ9 motor.  Build was done in 2012 and from then until last year it was driven about 500 miles.  I got a flash drive with some pics and a box of paperwork. It had some sort of aftermarket cam installed and the VVT removed.  I infer this because the stock pushrods caused valve clatter and it has hydraulic roller lifters.  Everything inside is fine so I got lifters that are .020 longer, which quieted the clatter completely.

The ignition comment is relevant, changing gap from the original .060 to .040 changed everything but not for the better.  It felt like too much advance then.  I put in some new plugs, that was when it went completely lean.  Then it killed the starter which means the driver side header has to come off.  So right now I am waiting on parts and sharing my confusion with all of you. 

Once it is back together I will play with plug gap and maybe wire in the wideband, right now if I blow up the ECU it would not upset me much.  Every comment seems to say this is not the right ECU for me.  A new harness and a microsquirt with some intelligence seems to be the answer.

The forum link is interesting but getting this system working seems to be a short term solution.  

In reply to iansane :

The 7703 ECM was used in pretty much every V6 J and L car through the end of OBD1. My 91 Z24 with a 3400 swap was running the same ECM, it was stock for the car but I swapped out the memcal for a PROM with a 3400 tune. Most swappers back in the day would just use the stock GM bin files and tweak the fuel maps for whatever they dropped in. The plus was it was pretty self-contained since it was a MAP system and, at the time, it was easier to work with than a full standalone system.

If it was a VVT engine, then it would definitely have to have a custom cam with that ECM because you can't control the VVT system and it doesn't play nice when left to its own devices. That was a big reason why I never looked at a VVT-equipped engine swap - I was a tuning newb and scared of the MS3 (at the time).

My Z had a wideband O2 and it worked fine with the OBD1 system. It was an Innovate LC1 IIRC. 

The fact that you get proper fueling when giving it throttle is weird. I haven't messed with GM OBD1 in 15+ years but I haven't seen that one before. Leaning toward sensor like you are and if it's the Innovate, it should come back to life when you replace it.

In reply to porschenut :

If you really want to play with it at all, grab an ALDL cable and some software. I used Tunerpro back in the day. Very simple to mess with things if you have a definition file and bin. (The definition file is basically the roadmap to the ecm and the bin file is the actual tune)

In reply to cobra17 :

That makes way more sense. Everything I was seeing showed that ecm only on the 1.6/2.0 tbi lemans cars.

In reply to iansane :

Thanks but I look forward to dumping this system.  I know the software and cable are cheap but the pain to deal with a 30+ year old system and wiring made by someone with questionable skills does nothing for me.  

Just ordered new wires, just in case.  The O2 sensor is the probable issue, when this all started it ran the same with it plugged in or unplugged.  

In reply to cobra17 :

When you used the innovate which O2 output did you use?  One looks like a narrow band range but I am not sure.

How does it run under wide open throttle?  The reason I'm asking is because it goes into open loop under WOT and ignores the O2 sensor. It also doesn't pay any attention to the O2 sensor during warmup and may even go back to open loop after it idles for a bit.

In reply to APEowner :

Depending on what operating system it uses, it also looks at the throttle position.

Generally speaking, and this is going from old memory, the engine should hot idle at 50 counts on the IAC and .5v on the TPS.  So you make sure the throttle is clean, adjust the throttle stop to that IAC count, and then adjust the TPS.

Yes, this needs a scantool like a Mastertech or other 30+ year old device to be able to read data.  And since it's a non stock application, you also need to know the OS (so you can say it's a 1991 Corsica or whatever when you connect) and pray that the car wasn't tuned around a poor initial setup.

Adjust WOT mixture with fuel pressure.  All changing fuel pressure will do at idle/cruise is change the fuel trim, since it will be in closed loop as soon as the O2 has a valid reading.

If you unplug the O2 from the computer, it won't do any fuel trimming.  There is an internal bias resistor that will read .450v if the sensor is unplugged, this failsafe is in the computer hardware.  If the computer was decently calibrated then this can help zero in the fuel pressure to run.

It's also a dumb system with little other failure management, if the coolant temp sensor is low it will dump fuel, AND if it's high it will also dump fuel to cool the engine.  Air temp sensor also trims fueling and needs to read correctly, especially in a non MAF car.  An unplugged CTS or IAT will read -40F on a scan tool.

In reply to APEowner :

WOT is an interesting question.  I should be able to answer it but the pedal is blocked at 50%, as pushing it harder just spins the tires in lower gears.  It does seem to run pretty well when the pedal is pushed, it is at a constant opening that the lean condition happens.

In reply to Pete. (l33t FS) :

Thanks.  You obviously spent time learning this system, something I am avoiding.  Sort of learning COBOL, it may help once in a lifetime but basically dead technology.

I did learn some stuff yesterday, ohmed the plug wires and one of them had the plug end pushed too high in the insulator so it was probably misfiring.  Which could have caused the lean readout.  So once the new starter is here I will do a trial start with the old O2 sensor(2 week delay on a new one) and maybe the output of the wideband.  

Thanks to all who shared their thoughts.  

In reply to porschenut :

It's always the ignition   And a misfire definitely will flatline an O2 to read lean

When I started my career, these were the vehicles I was maintaining, diagnosing, and repairing.  It's a nostalgic trip for me.

In reply to porschenut :

It's been a looong time since I messed with that car but by memory, it had two different circuits that were set up in the Innovate software on a laptop. I think mine had one set up for narrowband and the other for wideband. I'm pretty sure I've forgotten most of what I learned for OBD1 tuning - it's been replaced with Audi/BMW/Subaru useless knowledge. Most of what I remember from my GM days are the 12ohm injector resistance number and the ECM codes because I used those not long ago when a friend reached out with a few questions on a 3500 swap.

I always used Tunerpro RT with my old Z to do datalogging and, when I had the Moates Ostrich toward the end of my ownership run, on-the-fly tune updates. It was perfectly mediocre software for what it was, by no means great.

Pete's knowledge is more in-depth than mine. I was flailing around in my tuning journey using info from the third gen Camaro forum tuning section and a little bit from the Code53 forums. Mostly what you need to know is what bin mask your tune is using in order to see what it's got coded inside. Most of the 2.8/3.1 cars I think used $A1 so if your ECM was pulled from anything running a 60 degree V6, that's likely your bin mask. Once you set the proper mask up in Tunerpro, it will pull all the code and display it in a format that's intelligible.