twodoorhardtop

I have an 88 5.7 TPI IROC. Symptoms are 12mpg and a lean miss at idle. Measured .395mv at ECM from pin D7 (with D7 wire removed from connector) to tan wire D6. Same thing when going to chassis ground or neg. battery teminal. I replace the computer since reference was not the normal 450mv . New computer gave same lean miss and steady 390mv output. Cleaned up grounds and now .395mv.

Have I missed finding a bad ground?

The grounds I cleaned were: Negative battery cable at battery, block and fender. Two ring lugs at back of LH head. Braid attached to firewall at RH head. Couldn’t get to the other end of braid at RH head to clean but it was tight there. Chassis to firewall is 0.5 ohms with negative bat term disconnected and 7 ohms when connected. I find this strange.

Is the 395mv correct for my application or are they all 450mv reference?

Thanks!

With purple D7 (goes to the O2 sensor) and tan D6 (goes to ground) wires removed from connector, Pin D6 on ECM to ground gives 1.500V. Pin D7 on ECM to ground measured 860mv. Across D6 and D7 I get 640mv. I realize that 1.5v -.860v = 0.640v. When the D6 Pin is connected to ground ( which is what happens with the tan wire connected) I get 395mv. I assume D6 to ground completes the voltage divider circuit. I’m thinking the D7 should be higher than 860mv.

So what would cause this that is external to the computer?

There's good voltage at ECM D2. The reference for the MAF at ecm is 5.00V. I thought maybe something is finding ground through a sensor, bulb, coil, relay or motor.
I tried turning on different things like heater fan, wipers and lights but same 395mv. I also tried disconnecting just about every connector under the hood but it stayed at 395mv. When I started it I believe for a moment until it started that I got 450mv. I should have disconnected coil so it would crank for longer to verify.

Could it be a positive voltage leaking from somewhere to ground, raising the ground reference by 55mv? How do I go about finding that?

Vader

2-Door,

Welcome Aboard!

Have you tried a different oxygen sensor? The sensor element itself may be dragging down the bias voltage.

I've tested sensors that actually go to a negative 80-90mV output voltage when heaed in oxygen, and only go to about 800mV in an oxygen depleted atmosphere. I had determined that those sensors had been contaminated, and the bias voltage from the ECM was also being dropped.

twodoorhardtop

Thanks for the reply, Vader. I've been chasing the symptoms for a long time but now have it narrowed down. Help is very much appreciated.

Yes, it seems like somthing is dragging down the bias but I'm getting the steady 395mv with the O2 sensor disconnected. In fact, I disconnected it by removing its wire from the ECM connector and probing the pin directly.

BTW, with the O2 sensor connected, I've seen it swing from 40mv to over 1 volt. Its about a year old and believe its good.

I know I'm running lean from a buddies AFR sniffer on his chassis dyno. the .395mv from the ECM explains the lean condition. We could see the integrator lean it more even though AFR in the pipe was already lean.

I thought with the O2 wire removed from the ECM connector would eliminate wiring as being the problem and would definitely mean ECM was bad. Yet a new ECM and same 390mv. I not only checked the grounds again but cleaned most of them. I only picked up 5mv.

Is 450mv for all 165's?

Did I miss a ground? I mentioned which ones I tended to in first post.

All else seems to be working fine. One thing, I'm not sure of is that it goes into closed loop shortly after O2 starts working, even when CTS is about 100 deg F. I thought it shouldn't go closed until temp. is about 165. This occurs with old ECM and new.

Vader

You know, I've never actually checked mine. I've heard the "450mV" being thrown around as a theoretical bias voltage for many years. I guess I may have to check.

FWIW, the newer PCMs will report the bias voltage. I use AutoTap, and can read O² bias as a datum line. I've seen them under 400mV on an OBDII vehicle that is running "perfectly." Heater current is also reported as a separate datum.

Vader

Well, I just couldn't take it any more. I had to check my '86 (1227165 ECM) and guess what I found. I tried two different Flukes just to make sure, and discovered the same as you. My car is running flawlessly. I tried an engine ground, then went directly to the battery negative stud:

Vader

Incidentally, that was KOEO, 12.86V measured at the battery, and open circuit voltage (no sensor connected).

It would appear that whatever "problem" your ECM is experiencing, mine is having the same "problem". I'm guessing you need to look elsewhere. If you add 50mV for an oxygen sensor operating at full temperature in an oxygen-rich atmosphere, you'd arrive at the oft-quoted 450mV bias.

twodoorhardtop

Vader, thanks for checking the bias and for the pic! Incidently, that Fluke looks a lot like mine. Are you saying it has a 390mv bias and 450mv reference for stoichiometric? Don't a lot of cars actually have true 450mv reference and bias?

A week ago, when I was at my buddy's shop checking out my AFR and using his scanner, the Snap On guy was there to demo a new troubleshooting handheld computer. He entered my car info and looked up O2 sensor and it stated that it uses a 450mv bias (or did it say reference?). In any case, one would think they are telling you that so you can check it. I can't see a mechanic needing the 450 value for any reason unless he's going to write a program to check cross counts.

Could there be a special 395mv O2 sensor for my application? Maybe the O2 sensor I got for replacement is the wrong one and is for a 450mv bias instead of 395mv.

I've checked or replaced everything else I could think. New ECM, wires, plugs, cap, rotor, IAC, EGR, recent accel injectors and O2 sensor. Checked for vacuum leaks, fuel psi, MAF, CTS, MAT and CSI for leaking. Set throttle plate opening, TPS and timing. Tried going back to OEM prom.


If you load the engine and then let it idle, it takes about five seconds for it to start to miss about once a second. No miss when cold. Watching the AFR you can see it go lean and then miss. After clearing the computer, the integrator leans it out more even though its already lean in the tailpipes.

Not sure where to go next.

Vader

No, you cannot have my meter. It's almost an antique.

What I am saying is that the ECM evidently biases the sensor at around 390-400mV. A good, working zirconia oxygen sensor will generate about 50mV when heated to operating temperature in an oxygen depleted atmosphere. That same sensor will generate about 1.020V in an atmosphere with 21% oxygen while at operating temperature.

The frequently quoted "450mV" that were hear may well be the point at which a zirconia sensor is when heated and in an atmosphere of combustion gasses from a completely stoichiometric fuel mixture.

Honestly, I think the sensor and system are working in your case, and the problem you are experiencing is something other than a faulty ECM or O².

Going a bit deeper, what engine/year is this, and what PROM code are you running?

twodoorhardtop

Your're right, it’s almost antique!

Thanks for clarifying. It makes sense now. I agree, something else is wrong.

Engine is 88 5.7 TPI stock with 134k miles and prom code is ABYA. I tried this one but now have a hyper back in that I got around '89. No difference between them.

If there’s a miss from any source, could this make the ECM try to lean things. I can actually see the O2 line go to 1v occasionally. Maybe when it misses? To me this means “didn’t fire” rich and zero oxygen burned. Percent relative to atm is therefore same as atm at 21%, so you get 1V. This occasional rich reading could skew the average so ECM leans. Right?

Maybe its something with diverter valves? Or injectors?

I don’t think it’s engine mechanical. All the old plugs looked good after many thousands of miles. I’ve got AC’s in there now. I had a broken head bolt at about 100K and had a multi-angle valve job done on heads before I reassembled. Guides were good. Very little ridge in cylinders. Sucha pain to do a leak down test. Runs good when cold. Maybe rich only compensates for another problem?

It’s been driving me nuts! Its not bad running but annoying, plus MPG is bad.

Vader

If there is an ignition misfire in the left bank, the O² would detect the unburned fuel and likely try to lean the mixture, even when the other 7 cylinders are at perfect F/A ratios. Remember that the ECM is only watching 4 cylinders, so one dead one makes a 25% change. The other bank of cylinders is merely a slave to the left bank. Anything that affect the EGO on that side affects the whole engine.

BTQ - Remember, waaaay back to the early Fluke DMMs? The cheapest, lowest-featured, entry level DMM was the 8010. The one on the left is just a step or two above that. It's about all I could afford in 1980, since it cost more than my first car. My first car is long gone, but this old girl is still working fine.

88TPI406GTA

I am not sure how the measured voltage of .450 or .390 verifies how the O2 sensor is working. Why? Because to accurately measure an O2 sensor, the ECM needs to see its input as switching...which, when the O2 sensor heats up enough to start producing a signal, the stock O2 sensor generates a DC signal between 0-1 volt. Hypothetically, .450 millivolts is the middle ground voltage at which the signal crosses from high to low or vice versa.

What you need to see to make an accurate determination is some datalogs of what the ECM sees in open loop, closed loop and WOT...

The O2 sensor cross counts in closed loop has a certain spec (which I cannot remember right now) that will be a determining factor in how it operates.

I think that you need some datalogging. Check out the DIY PROM board for some good free options, like Moates software.

HTH,

twodoorhardtop

Vader, looks like your one had fraternal twins. I have the grandfather, a sibling to yours, and two cousins with the name Micronta. Can you imagine, I found the Fluke in my mom's outside trash with a bunch of junk.

I'm sort of back to square one with my car, but thanks to you it could rule out one bank. Probably can't mess with it anymore for a couple of weeks. I think I'll do a cylinder balance check next with the IAC disconnected.

Good point 88TPI. It does run good cold until a little after O2 starts swinging. Not sure exactly at what point it starts. I'd like to know what's happening first, the lean or the miss? Will also check for exhaust leak near O2 sensor. Then I will probably make a cable and datalog.

Vader

88TPI is correct about logging run-time data. I had presumed you were already doing that on the chassis dyno.

The fact that the integrator was building leaner tables indicates the ECM thinks the mixture is rich, even though the WB in the EG analyzer says otherwise.

The information you present points to either gas contamination of the sensor, since the on-board O² is indicating "Rich" while the external sensor says "Lean", or the on-board sensor is not seeing the whole picture (and we already know that is the case). Since it only monitors the left bank, any exhaust leak, AIR diverter leak, vacuum leak, or other condition that would cause a lean EGO on the right bank would not be detected by the on-board O², but would be detected by the tailpipe sniffer O². The ECM could be detecting a rich condition on the left side, while the right side just gets leaner and leaner due to the ECM's "corrective" efforts.

If you are certain the O² is good (and it may be, since the normal failure mode is toward a lean signal, not rich) you may have multiple problems. In any case, one of the problems is apparently a rich condition on at least one of the cylinders in the left bank. Potential causes of this are:

1. Misfiring spark plug;
2. Leaking fuel injector;
3. Streaming fuel injector, creating fuel puddling instead of atomization;
4. Sluggish injector, with propagation delays that create a "sticking on" condition.

I would suggest just what you propose - A power balance test of the injectors. Make sure you bypass the EST for the test and disconnect the IAC to remove any ECM idle correction. If you determine there is an injector problem (I'm guessing in the left bank) you may also want to check fuel pressure after shutdown to help indicate a leaker.

If you do discover injector problems, I'd suggest contacting Rich at Cruzin' Performance for options. The '88 models probably still had RP/Bosch injectors that are worth servicing. If they are Multecs, they may not be worth it. In either case, Rich will give you the best advice.

Please keep us posted. I'm interested in what you discover.

88TPI406GTA

As usual, Vader has a good course of action

So check:

1. Exhaust leak
2. Injectors
3. O2 sensor


Are you running a GM o2 sensor? The DIY PROM guys seem to find that a GM O2 sensor is the way to go since some of the other replacement brands have some issues in how they work with the ECM. It is entirely possible that the O2 sensor has been contaminated too...

Follow Vaders advice...the dark path is the best way

twodoorhardtop

The dyno is fairly new for my bud. He's more of general auto repair and race engine builder not a tuner, yet. He only has AFR diigital meter. No other gasses or logger, just a scanner.

Dumb question: How accurate is the AFR since it's meassured after the cat? Cats convert CO to CO2 with H2O as a by product. I know they are simply a honeycombed ceramic coated with platinum and other precious metals needed for the chemical reaction. Where do those extra O molecules come from? Are all AFR meters O2 based? I still think I'm lean but just curious about this.

My O2 sensor swings the full range but I know that still doesn't mean accuracy. It's a Bosch that is about a year old. It didn't make any difference at the time of replacment but I had found bad spark plug wires and EGR since then. I thought they were going to be the total fix for the miss. It ran better but didn't get rid of this minor miss. So it's possible that I traded problems.

Before replacement of these, I had a long-cranking-before-starting problem where fuel psi wouldn't hold so I replaced the injectors with new Accels. It solved the problem but thought it would also take care of my miss. They probably have only 20K on them and I keep the fuel filter changed so I don't "think" they're likely.

I've actually had the minor miss for years; around when the head bolt broke at 100K, now I have 135K. Not exactly sure when the bad MPG started but definitely way after the miss. It could have begun about a year ago. Hmmm, that's about when I replaced the O2 sensor. Well see!

I believe we all are on the same page.

Thanks guys! I'll update in a couple of weeks.

88TPI406GTA

Check out this thread for what a non-GM o2 sensor can do. It isn't the first one, BTW....

https://www.thirdgen.org/techbb2/sho...ighlight=bosch

HTH...