Edit.... symptoms have changed since I originally posted this. Read #2 (below) for current status. I'll underscore the differences. Thanks.

I've got the strangest driveability problem and at a loss for an explanation. Here are my observations:

Before installing the turbos, the engine ran great (with headers & no power adders). Throttle response was crisp, idles smoothly, etc. In fact, I passed emissions. In open loop, my wideband showed an A/F ratio close to the cold engine AFR tables. In closed loop, it reads at or close to 14.7 under normal driving conditions.

After installing both turbos with related hardware, including a heated 3-wire O2 sensor located close to the oil filter (same proximity as a header or stock exhaust manifold's O2 sensor), the engine runs very lean, 17-18 to 1. Weird thing is that this occurs in open and closed loop. The throttle response is sluggish and the idle hunts. I then read that the O2 should be closer to the turbo so I welded another bung, about two inches from the turbo. Result: No difference. I was thinking that I might have purchased a bad heated O2 sensor so I substituted back my original one-wire. Result: Takes longer to heat up. Other than that, no difference -- very lean.

So, I removed the driver's side turbo and reinstalled the original GM exhaust manifold (I sold the headers) with the heated sensor. Result: A/F ratios are back to "normal". I have no clue what's going on. To me, the only difference is the O2 sensor location and possibly turbulence in the exhaust flow. Are there any others? I did note a significant difference in exhaust temps. With the headers, the idle temp was >1000 deg F. With the cast turbo manifolds, the temp hovers around 600 deg F at idle. I am unable to check the temp with the OEM manifolds (no bung).

Please correct me if I'm wrong but in open loop, aren't all the engine sensors used except the O2? Or does the ECM run strictly off tables, bypassing all sensors?

Any ideas from the experts here? I really could use a clue....

Willie
aka "Clueless"

 

I've got the strangest driveability problem and at a loss for an explanation. Here are my observations:

Before installing the turbos, the engine ran great (with headers & no power adders). Throttle response was crisp, idles smoothly, etc. In fact, I passed emissions. In open loop, my wideband showed an A/F ratio close to the cold engine AFR tables. In closed loop, it reads at or close to 14.7 under normal driving conditions.

After installing both turbos with related hardware, including a heated 3-wire O2 sensor located close to the oil filter (same proximity as a header or stock exhaust manifold's O2 sensor), the engine runs very lean, 17-18 to 1 when in closed loop. The A/F ratio seems good in open loop. In closed loop, the throttle response is sluggish and the idle hunts. I then read that the O2 should be closer to the turbo so I welded another bung, about two inches from the turbo. Result: No difference. I was thinking that I might have purchased a bad heated O2 sensor so I substituted back my original one-wire. Result: Takes longer to heat up. Other than that, no difference -- very lean.

So, I removed the driver's side turbo and reinstalled the original GM exhaust manifold (I sold the headers) with the heated sensor. Result: Closed loop A/F ratios are back to "normal". I have no clue what's going on. To me, the only difference is the O2 sensor location and possibly turbulence in the exhaust flow. Are there any others? I did note a significant difference in exhaust temps. With the headers, the idle temp was >1000 deg F. With the cast turbo manifolds, the temp hovers around 600 deg F at idle. I am unable to check the temp with the OEM manifolds (no bung).

I just re-installed the turbo stuff and it's back to running extremely lean. The only changes I can think of is exhaust configuration and O2 sensor location. The only way to get the ratio to a reasonable number is to stab the gas pedal. It temporarily drops to the 14's, then back up to 17-18 to 1.

Any ideas from the experts here? I really could use a clue.... Weird....

Willie
aka "Clueless"

 

Interesting.... When in closed loop, the O2 sensor voltage remains basically fixed at max, 1091mV. It does NOT oscillate, and therefore does not produce any cross counts. So this sensor thinks my engine is running extremely rich, reducing injector pulse width because of it, causing it to run really lean... So why does it think the engine is running rich with the turbos when it isn't?????

 

Is it possible your turbo headers don't have a clean enough ground? You could try running a ground wire staight to the o2 bung and see if that helps. It's my understanding that turbo DSM's came that way from the factory. Worth a try at least.

 

That's a great suggestion. I actually thought of this last night and clamped a battery jumper cable to the O2 sensor and negative battery terminal. I wanted as good a ground as possible! Result: No different.

I performed one more experiment this afternoon. I tried the stock non-heated O2. Result: No different.

The O2 voltage isn't completely fixed as I eluded to earlier. It does vary but only from about 1030 mV to 1091 mV.

Willie

 

ur having the exact opositte of what mine did,mine always read lean,even with the single wire 02 2-3 inches away from the turbo outlet.but the switch to a heated o2 fixed my problems.im thinking maybe both o2 sensors u have are no good.if a motor runs contiously rich it can burn up the o2 sensors.

actually come to think of it a few months ago i had the same exact problem as u ,but no turbo.all the sudden my engine started running dead lean.and was throwing codes for o2 rich.to make a long story short heres what happened.i had an a/f ration gauge hooked up to the same sensor that pluged into the ecm.well th4e gauge went bad.and was backfeeding 1.1 volts down the line back to the ecm.i unhooked the gauge and it ran fine after thatwhat i sugest u do is unhook the o2 and start the car see hwo it runs.while its running with the o2 unhooked hook a volt meter to the o2 sensor wire comming from the ecm if it still reads 1volt u have a short somewere,if it dosent make sure the wire from the o2 to the ecm isnt grounded out anywere.also while the car is running check the voltage at the 02 again with it unhooked from the harness and see what it reads

 

Thanks for the reply Dave. I could go testing the voltages as you describe, but the plain fact of the matter is both sensors work as they should when screwed into the factory exhaust manifold. It's only when I install them into the turbo downpipe that the symptom occurs. This leads me to believe that there is nothing wrong electronically.

I just thought of a simple test. See if this makes sense. Unscrew the sensor out of the bung but keep the wiring harness connected. Ground the sensor body. Run the engine. When the ECM goes to closed loop, the sensor, being in free air, should read super lean and compensate by fattening up the injector pulse width. This should in turn be reflected in my wideband as a rich mixture instead of the 17 to 1 it's been displaying. If it does, I'll know the sensor is working. If it doesn't, can conclude the sensor is not functioning? I can try this with both sensors. Does this seem like a valid test?

Willie

 

Hey Dave,

I did a quick test as you described. I disconnected the O2 sensor and probed the connector (ECM side) for voltage. With the ignition on and engine not running, I got 0.467 volts -- good!! With the engine running, this voltage rises to just over 1.0 volt -- bad I think! These values are the same with the O2 sensor connected. So, this means the O2 ECM input is getting slightly over one volt from somewhere, doesn't it?

General question: Is the O2 voltage supposed to be in the 0.450 volt range in open loop and remain there until closed loop?

Willie

 

im not positive but i dont think the voltage from the ecm should change from .45 if there isnt an o2 hooked up.1 volt at that wire would indicate a rich condition to the ecm.which in turn the ecm will lean out the engine causing ur lean condition.im leaning heavily twards a bad ecm here.but since u say it runs right without the turbo im really not sure.is this a maf setup or speed desnity setup?how much vacum are u getting at idle.do u have any major vacum leaks?any chance u still have the egr hooked up?

i will dbl check my voltages with the o2 disconected tommorow just to see what mine does since i know mine runs right

 

....1 volt at that wire would indicate a rich condition to the ecm.which in turn the ecm will lean out the engine causing ur lean condition.

Exactly what it's doing!

I also thought a bad ECM so I replaced it (got two!). Same symptoms! Believe me, I've been through the wringer and tried everything I can think of. It was a last resort that I post my problem here.

My car is an '87 (MAF) but has been converted to SD (749 running $58 code). I don't think I have any vacuum leaks. My gauge shows 18 inches at idle which is darn good for 3100 feet elevation. The EGR has been disabled.

I appreciate you checking your voltages for comparisons sake. Actually I could do the same thing with my other '87 Camaro but it's in storage about 15 miles away and has no battery....

Willie

 

Okay, check this out.... I tested the O2 circuit for an open, using the diagnostics for Code 13, "Open Oxygen Sensor Circuit".

With the engine running in open loop, disconnect the O2 sensor:
Before disconnecting: 1091 mV
After disconnecting: 1091 mV

Jumper harness circuit to ground. Scan should display O2 below 200 mV:
Voltage dropped to 22 mV!

This indicates either a faulty O2 sensor connection or sensor.

Is it possible that I have two faulty O2 sensors? I suppose anything is possible. Maybe now I'll remove the turbo stuff and repeat all these tests with the OEM exhaust manifold.... Those heated O2 sensors are not cheap!

Willie

 

It sounds like the O2 sensor ground (at the back of the cyl head) is not connected or has a bad connection. It is more probable since the original O2 sensor worked before the install. Yes, it does happen that new O2 sensors can be junk but not very likely.

Check the tan O2 ground that goes to the back of the cyl head. You can do a visual and also ohm the cyl head to the O2 ground pin at the ECM (ECM unknown??).

The O2 ground is on a seperate lug/terminal and sometimes are forgotten when the other grounds are removed from the back of the head. A quick test would also be to run a wire from the ECM O2 ground pin to any chasis ground. You would see the O2 mV change when connecting the ground if that is what the problem is. Then you could fix it properly afterwards.

 

Update. Problem appears to be solved!!

Here's what I did:

1a) Removed/disconnected O2 sensor pin from ECM connector. O2 sensor connected to other end of this wire. Because it is an open circuit I should see zero volts. Ignition off: 0 volts. Ignition on: 0.62 volts. What??

1b) Disconnected O2 sensor connector. Rechecked voltages. Ignition off: 0 volts. Ignition on: 0.62 volts. This wire is shorted, receiving voltage from somewhere!

2a) Installed new pin/wire to O2 sensor at ECM connector, leaving the sensor end open. Checked voltages on this wire. Ignition off: 0 volts. Ignition on: 0.445 volts. Lookin' good!!

2b) Connected O2 sensor. Rechecked voltages with scanner. Ignition off: 0 volts. Ignition on: 0.445 volts. Started engine: 0.445 volts!! Before replacing this wire, the voltage was 1.091 volts! A/F ratio in the 12 to 14 range in open loop. Good so far! But the real test>>

3) Ran engine until closed loop attained. Watched voltage. Yes, there were cross counts but the voltages were not far from stoich, 0.450v. I suspected a cheap O2 sensor. It is a no-name brand.

4) Picked up an A/C Delco heated sensor. What a difference. The voltages swing from as low as 100mV to 900mV and back in a flash! I believe the problem has been solved...

Thanks to all of you for contributing to exercising my brain cells. This one was a tough one and I've been battling this for about two months. I guess you never stop learning about these darn old cars and their darn crustified wiring harnesses!!

Willie

 

I am glad you found the culprit Willie.

-J

 

You and me both J... It's been frustrating to no end. I've put in a lot of hours in a superheated garage trying different things.... then trying them again.... BTW, it was 110 here a couple of days ago. Add 15 degrees to get my "in-garage" temp.

To make one further point, this solution in no way explains why the O2 sensor worked differently with the OEM manifold. I'm not even going to attempt this because it really doesn't make sense. I'm just very content that it's now operating as it should.

When it was running in closed loop, my BLM in idle was 148. It would go to 128 when giving her a shot of accelerator. Before installing the turbos, my idle BLM was in the 127-130 range. Is this result expected? The only difference besides the obvious is the ambient temp. I last ran the car in January when temps were in the 60's. I'll post this question on the DYI/PROM forum.

Willie

 

The O2 sensor is a differential input at the ECM. That means that chasis ground has little to do with the O2 ground. If the O2 input is left floating and you have a .62v differential between O2 ground and ECM ground then you would see .62 volts at the O2 input (since it is floating). The technique for testing differential signals differs from testing signal ended signals (like CTS, IAT, TPS, etc).

So, which wire was the problem? The O2 signal ground or the O2 signal wire itself?

 

The O2 signal wire was the problem. I tested it first and if it proved to be good, the ground wire would have been next.

Willie

 

nice job on figuring it out.now that its running i would check to see were that o2 wire was picking up the voltage from.u could possibly have a few wires melted together or shorted somewere else in the car.could possibly cause more problems down the road.i got lucky mine just happened to be my gauge backfeeding the 1.1 volts.id trace the wire back all the way threw the harness in ur case.and make sure none of the wires are melted or have cut insulation.

 

Just curious if you found out where the short wa in the O2 to ECM wire. This kinda tech is so invaluable

Glad you got it fixed..