TPIShark

Hi all,

Been a long time since I've been on here, but as the title says, I'm completely out of ideas. Been working on my neighbor's '87 IROC for quite a while now. It suffered many setbacks from being left outside over last winter, without any attention. Bad fuel pump (only 2 years old), stuck injectors, bad IAC, the list goes on and on. Most of it I can attribute to the ethanol-blended gas in use these days. Without stabilizer, it kills everything in a short time. The fuel pump looks like it was in a lake for ten years. Anyway, I'm at a stage now where the car runs rather well again, except for two issues: It's hard to start cold, and it will not go closed loop, ever. It's a 305, auto, running the "165" ECM. I've been through all the sensors, replacing the MAF and the O2 along the way. There are no codes, and using two different scan tools, everything seems to be working properly. (Coolant temp, IAC movement, airflow, spark advance, etc.) The only data that's odd, is the O2 value and block learn / integrator values never move, but I believe this is a product of the O2 never coming "online". I've probed the O2 lead back to the ECM, and checked all my grounds as well. (It's an un-heated stock-type sensor, in the stock exhaust manifold.) I've been through all the diagnostic charts in the GM manual, and even swapped-in two additional ECM's I had lying around. It's worth noting I removed the cold start harness when we first put the car together, and was running a modified '89 bin file for the enrichment. I honestly don't even remember what PROM is in there now, as the ECM got wet during it's stint outside, which started this whole process. I dug out my old PROM burning equipment last night, to start futzing around with that problem. My main concern is it not going closed loop. I consider myself a pretty handy TPI guy, but this one has me stumped. Appreciate any input you guys might have. Thanks!

KyleF

From my reading I have an understanding that 3 conditions must occur for Closed loop


Oxygen sensor has varying output voltage
Coolant temperature above 105°F.
A specific amount of time (pre-programmed into the MEM-CAL) has elapsed.

I would start verifying these. It seems the problem lies with not seeing the 02 sensor voltage move. New part doesn't = good part. Possibly the new O2 sensor is also failing to operate properly.

burnout88

I was leaning toward a bad ECM. But you say you swapped it. So that actually then leads me to the wire itself that runs from the O2 sensor to the ECM. tap the wire coming out of O2 after the connector with a meter check the reading coming out of it. If it’s good and varying then the O2 is fine. Then tap the wire before it goes into the ECM. If you have no reading or it’s static. Then you have found the problem. Run a new wire from the O2 to the ECM. Problem solved.

TPIShark

Thanks for chiming in. I'm kind of laughing at the amount of views vs. responses, as I know it's a tough one. I did trace the O2 lead all the way back to the ECM, and it seems in good shape, including the pin at the ECM connector. Yes, I had two additional ECM's available. While I was honestly unsure of their condition, they both worked, but with the same issue, and no other issues. I'm figuring the odds of having two others with the same fault are not likely, but I suppose it's possible. I don't have my notes with me now, but I want to say a scan (and the voltmeter) showed the O2 stuck at something like 300mv.

SbFormula

Usually a Tune doesn't "detune" itself. Bad "new" 02 sensor? You should see around 451v before start-up and a gradual increase in voltage as the 02 sensor gets warm. Here's an example with a heated 02 sensor (Rec#117). Non heated should take more time. Close loop starts after 120sec in this case. 02 Ready after 7-8 secs (Rec#121). 02 Rdy turns "1" from "0". If it stays at "0" it won't enter close loop. Cheers!

TPIShark

Thanks SBFormula. As much as the car shouldn't require a heated sensor, perhaps I might try that. I should have mentioned that the issue persists well past ten minutes of run time, with coolant temp at least 180 deg F, so the timers and temp should not be a factor. I will check the calibration to see what it's calling for just the same.

Tuned Performance

iTrader: (92)

Check d6 at the ecm it’s a dedicated ground for the o2.
have you measured voltage at the o2 while engine is warm ?

SbFormula

Heated O2 is nice, specially with headers when you have to install O2 sensor way down the pipe. It prevents it from exiting CL at idle. Make sure the heating part is activated with fuel pump if possible. Mine is activated with car in RUN which will activate it if I leave the key in RUN. Not good if the engine is not running. I damaged one O2 like that as I had left the key in RUN for a while during testing.

nice little wire adapter from TPI Parts
https://tpiparts.net/inc/sdetail/2167

Tuned Performance

iTrader: (92)

Cheaper on ebay
https://www.ebay.com/itm/TPI-TBI-3-W...sAAOSwmgJY2rey

you can use a acdelco afs74 o2
hooking power from the pink/blk of the esc module or fuel pump feed works.

SbFormula

Cool!!!
NGK21003 will fit too. Whatever is less expensive

Vader

There are a few more things involved in "closing the loop" in a 1227165 ECM with a typical AU/AR/BM/BU EEPROM codes. This is a partial (nearly complete) list:

AUM/BUA Closed-Loop Operation Parameters
CTS < 14.7ºC (58ºF) for 75 seconds;
CTS >14.7º (58ºF) and < 40.7º (105ºF) for 51.4 seconds;
CTS >40.7ºC (105ºF) for 12.5 seconds;
O2 >0.699V & O2<1.99V for 10 seconds;

Coolant Temperature Sensor Related Parameters
BLM enabled between 50ºC (122ºF) and 140ºC (284ºF)
Cold spark advance disabled above 56ºC (133ºF)
Hot spark retard begins above 116ºC (240ºF)
Highway Mode spark advance > 59.8ºC (140ºF)
Knock sensor disabled below 66.5ºC (152ºF)
Power enrichment at base A/F ratios > 56ºC (133ºF)
Target IAC idle RPM >80ºC (176ºF)
IAC multiplier at 1.0 (base) > 32ºC (90ºF)
Knock Control enabled > 67ºC (153ºF)
EGR Duty-cycle enabled at 56ºC (133ºF)
EGR Duty-cycle at MAX >80ºC (176ºF)
TCC lockup enabled >50ºC (122ºF)
SHIFT light enabled >50ºC (122ºF)
Diagnostic communication enabled at 70ºC (157ºF)
Code 13 (oxygen sensor fault) enabled above 70ºC (157ºF)
Code 32 (EGR fault) enabled above 30.5ºC (87ºF)
EVAP canister purge enabled above 70.3ºC (158.5ºF)

Vader

Just for additional reference, what is the target idle RPM when the engine is at temperature (above 176°F)?

Also, if you suspect the interface software is not communicating and/or displaying status properly, you can always jumper the ALDL to see if the ECM is actually in closed loop or not.

Field Service Mode
On the OBDI ECMs, you can jump 'A' and 'B' terminals on the ALDL while the engine is running.
WARNING! This must be done after the engine is running. This is called "Field Service Mode" and will not harm the ECM.

If the ECM is in Open-Loop mode, the SES light will flash rapidly, about 2½ times per second. If it's in Closed-Loop mode, it will flash about once per second. When in Closed Loop mode, flashing less than once per second indicates the ECM is enriching the mixture above the 128 count base line. Flashing more than once per second indicates the ECM is leaning the mixture below the 128 base line.

TPIShark

Just wanted to close the loop on this. (Sorry, couldn't help myself! :-D) After another round of careful diagnosis, and digging into the harness, I traced the tan wire (CKT 412, O2 ground) back to a broken lead behind the engine. Not sure if it originally shared the ECM grounds at the back of the head, but I crimped a ring terminal on it, and attached it to the firewall, sharing the braided copper ground that goes between the firewall and block. It now goes closed loop after the timers kick in, but keeps pulling itself back out of closed loop. I'm guessing the overly-rich condition may have fouled the O2 sensor, and the plugs as well. Going to address that, followed by some more scanning at this point to check the tune. I purchased a "G1" package from Moates to make that a little easier to work with. I was trying to hook and old laptop with a serial port up to my OTC scanner, to get a better view, but it seems the software available now (in some cases free) makes that a better choice than fighting with old equipment...