so.. what if I can't screw a different O2 sensor all the way in?

[daeron]

Im trying to swap another used O2 sensor into my car... and I cant screw it all the way in. The threads on the original one from the car were totally buggered; I managed to get my first used replacement to tap all the way in, but this second one isnt gonna do it.

 

By the way, the one that WAS in my car, and not returning any voltage after being heated by a good spirited drive, sure passed the bench test with a propane torch easily enough. So did the one that had been in the car; the one I am trying to put in NOW took the longest to heat up and display any voltage; but it displayed right at 0.65 volts, where the old old one didnt get up over 0.4, and the newer old one (that I just took out and tested ten minutes ago) went rapidly up to 0.8-0.9 volts.

 

I had checked the resistance between the body of the O2 sensor and the chassis of the car (testing to see how well the O2 sensor was grounded) and got VERY low impedance, virtually zero ohms.

 

Thoughts? clicky for the thread describing my issues in detail, but be forewarned that is already a very long thread. I'm gonna start a new one if the replacement O2 fails to do anything, with a brief summary.

 

Should I try to go get a new manifold/cat?? That costs as much as the ECU, $25 at the boneyard. Mine DOES have a few leaks, but most of those are after the O2 sensor anyhow, and I am not worried about the leaks.. (no noise/fume problems to speak of; I like my engine being as loud as it is..)

[Gloyale]

Im trying to swap another used O2 sensor into my car... and I cant screw it all the way in. The threads on the original one from the car were totally buggered; I managed to get my first used replacement to tap all the way in, but this second one isnt gonna do it.

 

By the way, the one that WAS in my car, and not returning any voltage after being heated by a good spirited drive, sure passed the bench test with a propane torch easily enough. So did the one that had been in the car; the one I am trying to put in NOW took the longest to heat up and display any voltage; but it displayed right at 0.65 volts, where the old old one didnt get up over 0.4, and the newer old one (that I just took out and tested ten minutes ago) went rapidly up to 0.8-0.9 volts.

 

I had checked the resistance between the body of the O2 sensor and the chassis of the car (testing to see how well the O2 sensor was grounded) and got VERY low impedance, virtually zero ohms.

 

Thoughts? clicky for the thread describing my issues in detail, but be forewarned that is already a very long thread. I'm gonna start a new one if the replacement O2 fails to do anything, with a brief summary.

 

Should I try to go get a new manifold/cat?? That costs as much as the ECU, $25 at the boneyard. Mine DOES have a few leaks, but most of those are after the O2 sensor anyhow, and I am not worried about the leaks.. (no noise/fume problems to speak of; I like my engine being as loud as it is..)

 

Spark plug hole tap. should be the same size

[daeron]

Spark plug hole tap. should be the same size

 

I already thought of that, but until someone else told me, I had forgotten that i DO have one of those!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

 

(BTW, did anyone else realize that you cant capitalize the letters when you type out a smiley code? )

[daeron]

well... *pant*pant*pant*

 

after a good amount of labor *pant*pant*pant*

 

I got the threads chased on the cat, and the "new" O2 sensor is mounted flush.

 

*pant*pant*pant*

 

and a recommendation.... if youre malnourished from a diet of PBnJ, and Ramen, with occasional bits of meat and other real food in between...

 

*pant*pant*pant*

 

Refrain from cutting threads into steel with a tap or thread chaser.

 

My arms feel like the ramen noodle that has been sustaining them FAR too much since I lost the job at the BBQ place. (BTW, Sonny's is NOT guaranteed to "make you feel good")

 

Now I get to lay on my back in the rain and re mount the exhaust, and quite frankly I doubt its going to make a difference. I WAS gonna get some video of me driving the car today, but that will have to wait until tomorrow. The exhaust might wait, too. I am already exhausted. (har, har)

[Gloyale]

Now I get to lay on my back in the rain and re mount the exhaust, and quite frankly I doubt its going to make a difference. I WAS gonna get some video of me driving the car today, but that will have to wait until tomorrow. The exhaust might wait, too. I am already exhausted. (har, har)

 

I don't know, I'm pulling for ya here. The new sensor put out Quite different numbers in the bench test. that could be a sign.

[daeron]

OKAY....

 

The ECU "learns" how to correct A/F ratios from the O2 sensor, right?? so REALLY to CLEAR my problem, i should need to clear the memory on the ECU?? SO, if I just fixed my problem, it should show by a marked improvement in performance, but if its still a little off clearing the ECU should fix it??

 

Imn running to get some fuel now, and also driving to heat up the O2 sensor... im gonna get back and check the voltage output of it, anxious to get feedback on these questions.. i am PRETTY sure its running better..... it definitely isnt running as bad as it was yesterday, but it wasnt always consistent with its problems. I HAVE thought I had fixed it before... but it MAY be fixed!

[daeron]

Well i dunno nothin about nothin. It isnt running anywhere nearly as bad as it has been running.... so MAYBE i ought to go get a brand new O2 sensor. I have $65 bucks left to play with (got some cash from my brother for working with him painting at his house) and I just dont know.

 

Voltage readings out of this O2 sensor were more like what they should be.. i THINK.. but i think the battery in my DMM needs charging. at first I was reading 0.12 volts out of thin air, with no probes plugged into the meter..... (i DID get it down to like, 0.7 before testing anything...) so to say my voltage test results just now were "inconclusive" doesnt even BEGIN to touch the tip of the iceberg. I got like, 0.3 or so at first.. then I tired unplugging a spark plug wire, and it went to ZERO! not to 1 volt, to Zero volts.... hitting the gas with one plug undone generated MAYBE a tenth... and then, plugging the spark back in didnt bring it back up at all for a little while, then it brought it back to about 0.2 volts.

 

The car is CERTAINLY running better... BUT on the drive to the station, it started being REALLY bad, acting like it was starving when it was NOT starving on fuel...... I unplugged the battery cable while I went in and paid, and pumped, so the battery got unplugged for about five minutes or so... would that "reset" any memory of the O2 circuit patterns saved in the ECU? should i "clear memory mode" with the white and green connectors? Who IS Eric Cartman's Father?? Tune in next time, for another exciting episode of "daze of our soobs...."

[Gloyale]

Well i dunno nothin about nothin. It isnt running anywhere nearly as bad as it has been running.... so MAYBE i ought to go get a brand new O2 sensor. I have $65 bucks left to play with (got some cash from my brother for working with him painting at his house) and I just dont know.

 

Voltage readings out of this O2 sensor were more like what they should be.. i THINK.. but i think the battery in my DMM needs charging. at first I was reading 0.12 volts out of thin air, with no probes plugged into the meter..... (i DID get it down to like, 0.7 before testing anything...) so to say my voltage test results just now were "inconclusive" doesnt even BEGIN to touch the tip of the iceberg. I got like, 0.3 or so at first.. then I tired unplugging a spark plug wire, and it went to ZERO! not to 1 volt, to Zero volts.... hitting the gas with one plug undone generated MAYBE a tenth... and then, plugging the spark back in didnt bring it back up at all for a little while, then it brought it back to about 0.2 volts.

 

The car is CERTAINLY running better... BUT on the drive to the station, it started being REALLY bad, acting like it was starving when it was NOT starving on fuel...... I unplugged the battery cable while I went in and paid, and pumped, so the battery got unplugged for about five minutes or so... would that "reset" any memory of the O2 circuit patterns saved in the ECU? should i "clear memory mode" with the white and green connectors? Who IS Eric Cartman's Father?? Tune in next time, for another exciting episode of "daze of our soobs...."

 

I think I've read in the FSM that afgter replacing either the MAF or the O2 sensor that the ECU will need possibly a few hours of driving to "relearn". I would *clear memory mode* just on principle. Then drive until you think you might have a problem, then see what you get with U-check and Read modes and see where you're at.

 

As far as unplugging a Plug wire making no volts. Wet, unburnt fuel *quenching* the O2?

[daeron]

I think I've read in the FSM that afgter replacing either the MAF or the O2 sensor that the ECU will need possibly a few hours of driving to "relearn". I would *clear memory mode* just on principle. Then drive until you think you might have a problem, then see what you get with U-check and Read modes and see where you're at.

 

As far as unplugging a Plug wire making no volts. Wet, unburnt fuel *quenching* the O2?

 

Thanks for sparing me the time to read that in the FSM, i think thats more or less what it says to do, now that you mention it.

 

And your "quenching" theory fits what I saw perfectly; it WAS just a 2+ mile drive to the gas station, and a 2+ mile drive back.. I hit full throttle up to 70MPH with the AC on for about 3/4 of a mile on my way back in; but REALLY a drive up to the shop, or my parents place, or somewhere ELSE would probably be the best way to check things out.

 

I dunno, tomorrow is another day, the battery in my multimeter will be fully charged so I shouldnt get any false voltage, I may just go up to the shop to use the multimeter that I know is up there... BUT i dont know what kind of battery it takes; and knowing my dear little brother, THAT multimeter has no battery in it right now.

 

and quite frankly, I ought to go BUY a cheapo analog meter just to test my TPS again. any clue what I could get one for from home depot? I still need to check TPS and MAF function at the ECU end to confirm that i have no wiring fault; but i have no codes related to either. Maybe the landlord can find his semi-fried analog meter.. (he blew the voltage aspect out testing something once, but IIRC the resistance test on it still works fine)

[Cougar]

As far as your meter goes I suspect that it may be ok. If you see a small amount of voltage on the display in the voltage mode short the leads together and see what the display shows then. Hopefully it will be zero volts then.

 

The voltage you measured at the O2 sensor while the plugwire was removed doesn't seem right. Measure the sensor again doing the same test but also remove the ECU wire to the sensor. See if the sensor voltage goes higher then. If it does then it appears the ECU has a problem and is loading down the sensor output.

[joostvdw]

your ECU needs about 1 hour to fully reset! so leave the batt cable off and go watch some TV

 

after you've reset the computer, go have some fun, the ECU needs to learn the values (and most importantly, the extremes) of the sensors to fully and correctly control the injection system, after a bit of driving you'll notice it running smoother

 

also getting everything up to operating temp isn't a bad thing either, the old O2 sensor may need some time to "restart" so to speak, the engine will need some time to burn out all the crap that settled there when it wasn't running right.

 

so fill er up (that's where the $65 is going ) and enjoy the ride!

[Gloyale]

your ECU needs about 1 hour to fully reset! so leave the batt cable off and go watch some TV

 

 

I haven't read that anywhere in the FSM or any other literature. I don't think it takes hours. Just run "clear memory mode". It may take a few hour of total drive time, but the learning process of the ECU is continual, no need to unplug it for hours. Joostvdw, where did you read this?

[daeron]

My multimeter needed a charge on the battery; there was no "zeroing" it at all.

 

I tried again today and again got very inconclusive results... totally inconsistent voltages. It seemed to climb up to 0.2-0.3+ range when I had the plugs all plugged in, and hit the throttle.. but other than that it was minimal voltage. And ALL measurements on the O2 sensor output voltage have been unplugged from the ECU... should they NOT be? I hit the "new" old O2 sensor with a blowtorch really hot before I ever put it in the car.

 

The car DEFINITELY runns different now.. and its definitely better, but not completely healed. It also now has an occasional tendency to sputter WORSE than it was, usually when I come off the brake.. This new sputter is EXACTLY as if I had just taken a fast left turn, with an almost empty tank, and starved the fuel pump. Most all my problems are relegated to low RPM running now.. and fiddling with the pedal seems to make a difference. I *really* need to get an analog multimeter and check for smooth operation of the TPS; I haven't detected ANY problems with it using my digital meter but it is acting EXACTLY as it would if it was seeing an occasional discontinuity in the TPS output. I also need to pull my plugs again after tooling around some tomorrow and see what they look like. My brother mentioned that a cracked insulator on a plug could cause a similar problem, and when I inspected the plugs last time I was just examining the state of the electrode for burn characteristics; I didn't check the body of the insulator.

 

If I can find an analog meter, and confirm proper TPS operation with that;

AND inspect the plugs to ensure they are 100%;

AND confirm proper ECU-end readings from both TPS and MAF;

AND driving around a little bit after clearing the ECU doesnt do anything;

then I am going to buy a brand new O2 sensor. (by Fuji, I hate being broke... )

 

Unfortunately, all this has to wait until probably wednesday, because the landlord and I have scads of work to do around the house and today and tomorrow are his days off. (among other things, we are cleaning the house.. I mean, literally, cleaning the house.. bleaching the brick fascia on the outside walls, and the roof shingles.. stupid mold.)

[Gloyale]

 

then I am going to buy a brand new O2 sensor. (by Fuji, I hate being broke... )

 

The generic single wire Bosch ones are only about 20 bucks. They seem to work fine.

[daeron]

yeah, i had intended to call around getting direct quotes. I wonder if my friend Ralph at the local small chain store might be able to help me out on this...

 

It helps when the general manager of the store down the street is of the opinion that you are the only person in the world who could cut his old pomeranian's toenails without the dog freaking out, or bleeding, or anything.

 

Veterinary Technician FTW!!

[Cougar]

I see from your reply that the O2 sensor voltages you gave were with the sensor output isolated so there was no possibility of it being loaded down by the ECU circuit. Assuming the sensor is good it still seems to indicate the engine is running lean. This could be due to a intake leak after the MAF sensor or a leak in the exhaust manifold which you said you had. It puzzles me why the sensor didn't change at all with the plugwire disconnected.

 

One thing you could try to do in order to check out the TPS is use a small bulb, like a dash light, in series with the wiper lead of the sensor. If there are some trouble spots you should see them with the light.

[RenaissanceMan]

Ok, just some info on the conventional O2 sensor operation to clarify any misconceptions. Read carefully and hope this helps everyone here.

 

The oxygen sensor generates voltage by a difference in oxygen content. Literally the voltage is created by oxygen molecules travelling from the outer "reference" area, through the material, into the inner part where the absence of oxygen is.

 

When the car is working properly in closed loop operation, it will switch between 0.2 volts to 0.8 volts. Only test this with a DMM (high impediance) an analog meter which in most cases is considered low impedance can draw too much current and damage the oxygen sensor. 0.2 volts being lean, 0.8 volts being rich (these values may travel to more extreme in either direction on well controlled systems)

 

When you remove a spark plug wire from your engine, you are yes, introducing unburnt fuel into the exhaust, HOWEVER you are also introducing unused oxygen from that cylinder into the exhaust as well, which is why it will read lean (low voltage).

 

The proper way to test oxygen sensor operation is to use a propane source. While the engine is idling and the computer is in closed loop operation (warmed up) you should notice the sensor alternating between .2 and .8 volts every 1-2 seconds (depending on vehicle) This is the ECU controlling the amount of oxygen and carbon monoxide into the exhaust, which allows proper 3 way catalytic convertor function (if you want to know why the cat needs this, send me a PM). Back to the O2 sensor... introduce propane (not lit of course) into the intake of the engine. This is adding extra fuel causing a proper rich mixture and the voltage should go high. the computer immediately starts compensating, and eventually you should see the o2 voltage start alternating again. Now remove the propane source and you should see the o2 sensor read real LOW for a while. The engine may stumble, and the computer after a few seconds should adapt to the normal fuel mixture again, and the o2 sensor should go back to its normal .2 to .8 fluctuating.

 

Please note on older computer-carbed subies, these results may not be as "spectacular" because of the limited ability those older computer systems had.

 

conventional O2 sensor thread size is M18x 1.5

Spark Plug thread size is M14x1.25 (there are spark plug sizes of course such as M10 or M12 used in motorcycles and other small engines)

 

Please note that any exhaust leak between the O2 sensor and the engine (really common on subies) will cause it to read lean, therefore causing your engine to run rich, because the computer's block learn is compensating for it

 

the oxygen sensor is not used until it warms up. On modern systems with an O2 heater, this may be only 30 seconds. On older 1 wire O2 sensors, this may take 5 minutes, or longer in cold weather. (if your drivability problem occurs right from the start before your engine even warms up, your o2 sensor is NOT the problem, unless you have a bad o2 as well as a bad engine temperature sensor that failed to low resistance, but thats quite a circumstance)

 

Exposure to coolant vapor or other wrong chemicals can destroy an oxygen sensor.

 

Always test the o2 sensor while its connected to the ECU, and using a propane torch to "benchtest" an o2 sensor is risky business as it may get too hot and destroy it.

 

-Mike

[joostvdw]

I haven't read that anywhere in the FSM or any other literature. I don't think it takes hours. Just run "clear memory mode". It may take a few hour of total drive time, but the learning process of the ECU is continual, no need to unplug it for hours. Joostvdw, where did you read this?

 

I haven't read it in any books or FSM's, but I have read it several times on xt6.net and usmb.org and heard it from some old subie techs, so I guess there is some truth in it

 

on the other hand, it can't do any harm to leave it unplugged for an hour or so (I didn't say or mean hourS, 1 is enough)

[daeron]

The oxygen sensor generates voltage by a difference in oxygen content. Literally the voltage is created by oxygen molecules travelling from the outer "reference" area, through the material, into the inner part where the absence of oxygen is.

I've got all this figured out, but thanks for re-iterating anyhow

When the car is working properly in closed loop operation, it will switch between 0.2 volts to 0.8 volts. Only test this with a DMM (high impediance) an analog meter which in most cases is considered low impedance can draw too much current and damage the oxygen sensor. 0.2 volts being lean, 0.8 volts being rich (these values may travel to more extreme in either direction on well controlled systems)

All tests done with a DMM, and I never read anything over 0.35 volts.. although I never hit it with propane into the air intake.

When you remove a spark plug wire from your engine, you are yes, introducing unburnt fuel into the exhaust, HOWEVER you are also introducing unused oxygen from that cylinder into the exhaust as well, which is why it will read lean (low voltage).

 

The proper way to test oxygen sensor operation is to use a propane source. While the engine is idling and the computer is in closed loop operation (warmed up) you should notice the sensor alternating between .2 and .8 volts every 1-2 seconds (depending on vehicle) This is the ECU controlling the amount of oxygen and carbon monoxide into the exhaust, which allows proper 3 way catalytic convertor function (if you want to know why the cat needs this, send me a PM). Back to the O2 sensor... introduce propane (not lit of course) into the intake of the engine. My cat is gutted; any difference there?This is adding extra fuel causing a proper rich mixture and the voltage should go high. the computer immediately starts compensating, and eventually you should see the o2 voltage start alternating again. Now remove the propane source and you should see the o2 sensor read real LOW for a while. The engine may stumble, and the computer after a few seconds should adapt to the normal fuel mixture again, and the o2 sensor should go back to its normal .2 to .8 fluctuating.

 

Please note on older computer-carbed subies, these results may not be as "spectacular" because of the limited ability those older computer systems had.

 

conventional O2 sensor thread size is M18x 1.5

Spark Plug thread size is M14x1.25 (there are spark plug sizes of course such as M10 or M12 used in motorcycles and other small engines)

Huh. My plug thread chaser worked perfectly for the O2 sensor bung, using the larger side of it.. same size, same threads, i carefully checked before trying it and it was confirmed by the O2 sensor screwing in mostly by hand after cleaning the threads.

Please note that any exhaust leak between the O2 sensor and the engine (really common on subies) will cause it to read lean, therefore causing your engine to run rich, because the computer's block learn is compensating for it

 

the oxygen sensor is not used until it warms up. On modern systems with an O2 heater, this may be only 30 seconds. On older 1 wire O2 sensors, this may take 5 minutes, or longer in cold weather. (if your drivability problem occurs right from the start before your engine even warms up, your o2 sensor is NOT the problem, unless you have a bad o2 as well as a bad engine temperature sensor that failed to low resistance, but thats quite a circumstance)

Well, my problem is straight up, starts (such as it is) as soon as i start my engine. Would a new O2 sensor be sending money after bad, then?

 

Exposure to coolant vapor or other wrong chemicals can destroy an oxygen sensor.

 

Always test the o2 sensor while its connected to the ECU, and using a propane torch to "benchtest" an o2 sensor is risky business as it may get too hot and destroy it.

 

-Mike

 

 

Regarding your final comment, about always testing it connected to the ECU... this question has been the only bit about O2 sensor operation that I didnt understand.. (although your post was appreciated, most all of it was well within my grasp just from having read the FSM thoroughly several times)

 

What difference is made checking it with the ECU plugged in?? My understanding is that the electromotive force (voltage output)is totally dependent upon the amount of oxygen catalyzing across the tube inside the sensor.

 

BUT, what you are saying is basically that my problem does not lie with the O2 sensor. The chief reason I am even concentrating on it is because that is the only code thrown by the ECU relevant to any run problems (other codes are VSS, neutral position, park position) In other words, what you are telling me is that once I go confirm proper MAF and TPS readings at the ECU plug (already confirmed at the sensors themselves) AND check my spark plugs again, then I am left with needing to replace the ECU; OR some sort of major internal engine damage, which I HIGHLY doubt. The way it tends to come and go in my mind eliminates possibility of misjudged cam timing on one belt or both.

 

I've got sixty bucks. I can afford to get a new O2 sensor, replace the ECU, OR buy an analog multimeter to test the TPS with as opposed to my DMM. Any ONE of those I can float; any TWO will SEVERELY cripple my efforts to get a job, and all THREE will leave me with an empty tank (and no cigarettes) by the time I buy them all; if not before.

 

Please, post away.. I dont care how marginal the posts may be in terms of actually HELPING me, I need mental stimulation here.

 

I am off to eyeball the sparks and run the wiring checks I mentioned. after that I will drive it, warm it up good, and check the O2 sensor once more, this time plugged into the ECU and see what difference that makes. I have yet to see my sensor indicate anything but "lean" conditions; and the leaks arent THAT bad until after the sensor; most of the leaking is within the body of the gutted cat where the weld lines are rusting.

[mikeshoup]

Regarding your final comment, about always testing it connected to the ECU... this question has been the only bit about O2 sensor operation that I didnt understand..

You should test it without it plugged into the ECU... just unplug the sensor in the engine bay and plug your multimeter's positive lead into the O2 sensor side of the connector, and ground the negative lead.... that should render the actual voltages the O2 sensor is giving as it generates its own voltage.

 

You can do it with the O2 sensor plugged in as well. Find the check connectors under the dash. One of the pins should be the O2 sensor output. Then you can open the throttle a bit and watch the O2 sensor cross hatch.

[Gloyale]

I am still suspiscious of that ECU. But there are other things too. But everything you are describing seems like not enough fuel being injected. (which would make sense with the lean readings on the O2)

 

After all this it could be as simple as the fuel injector itself. Have you done any tests to confirm it is working right?

 

The other possibiltiy is the fuel pump. Apparently that is what was wrong with Caboobaroo's coupe. I believe it is an 87 SPFI as well. Pump was working, but not up to par, according to the shop he took it to. I guess we should wait to here if it's "fixed" first before assuming that really was the problem. But perhaps?

 

I know, everything I say is "perhaps?"

[daeron]

 

Tell me those sparks dont look perfect. I mean that; without my haynes book i am worthless for reading plugs; but to my best understanding that is exactly what they should look like.

 

I have changed out the fuel pump AND pressure regulator with JY units; and I dont have a high pressure gauge to confirm proper pressure.. but my lines all blow through easily enough, the pickup screen isnt clogged, and its a fairly fresh filter. I know this is not scientific diagnosis, but outside of driving to my brothers shop and taking time out of HIS day to confirm this, i cant think of anything else. Besides, what lean condition!?? I mean, am I wrong in saying the sparks look fine?

[daeron]

okay.......

 

I just went to check the TPS, and the fragment of the 89 service manual i think is not right for my car.....

 

I wind up getting resistance values that vary properly between pins 28 and 25 on the large yellow ECU plug, but the 89 FSM says that pin 28 should be for "self shutoff??" Anyone care to confirm that I am checking the right pins, and while you're at it let me know what pins should be for the MAF on my 87 SPFI.. I will check according to what I have in the 89 FSM, and I should be able to find out on my own hopefully.. but i figure may as well ask THAT now while im asking for confirmation on my TPS pinout.

[daeron]

Well.. TPS checks out at ECU end with my DMM, and i dont have an analog meter to test it with. MAF operation checks out ath ECU end. Sparks don't have any physical damage.

 

I am horribly torn between ECU and O2 sensor... so I think what I will do is clear out the memory on my battery, and play a waiting game.. god help me if the loyale in the boneyard disappears.. I couldnt use an ECU from a turbo, could I? there IS a turbo that just got dropped off out there (and I havent seen one of these yet, im anxious to go look... )

 

anyhow, thats where I am at.

[mikeshoup]

ECU has to be from an SPFI... probably even an 87. I know there's been wiring changes from 88-89, and 90+... I dunno if that would have an effect on whether or not you can use a later ECU.