Hot Rod Forum banner

OxySensor Fiasco

3K views 31 replies 6 participants last post by  B-1 
#1 ·
Hi Everybody, I've never needed to replace an O2 sensor till now. It ain't coming out. Spent nearly a couple days and the only progress I've made is discovering what I would have done to engineer a better system.

The car wouldn't idle in gear without a slight power brake manipulation.
Unhooked O2 sensor and now I don't need to keep one foot on brake & one on gas to idle in gear. I've used a torch and a wide assortment of tool combinations to get the sensor to brake free. Now the hex on the sensor won't hold a socket without slipping.

At this point I'm down to figuring I'll leave it in for now, braze up the top to stop an exhaust leak where the guts/wire of the sensor broke out and fool the ECM into thinking the sensor is working (see below).

Surely someone has done this too. I was hoping for advice on the best way to rig a variable resistor up to the single wire going to the ECM. That way I could fine tune the fuel to rich or lean according to what I want for any senario; and I was thinking this would be beneficial to maybe injecting hydrogen and I could adjust the voltage for a leaner gas mixture for that purpose without a MAP enhancer???

It's on an '88 Regal -4 speed auto - 2.8 liters - no room to get a straight shot at the sensor or turn a tool against the firewall for a solid bite.

Any advice welcomed to steer me into a better direction. If I pull the engine or slightly hoist it at some future point, I'll replace the sensor unless what I'm thinking works better for my purposes.
 
See less See more
#4 ·
Hi CDJr. I don't have an electric welder, only an oxy/acet torch. Using a size "0" welding tip is taking too long to heat, so I'm going to get a #4 at Harbor Freight today or tomorrow. Since a #4 is rated for welding 1/4" material, I may give that a try. I am
hesitant though, 'cause I don't want to do any damage that would require repairing the exhaust manifold. Better to leave as is and do the mod I mentioned in the beginning. That way I can still use the vehicle and get it inspected here in WV. Repairing the manifold is going to require removing the head for sure. I can't sacrifice that much time and pain. Don't care about sacrificing a tool to get a job done though. Thanks for the idea. I haven't got that far in thinking of alternatives yet.
 
#6 ·
Hi cobalt327. Thanks for the link. I'll spend more time on that later. Good stuff! I started by renting an O2 sensor tool kit from AutoZone and following the Service Manual instructions by getting the engine hot. That didn't work. I shoulda known that 99% of the time a 24 year old 02 sensor is not going to come out that way. By this time the ceramic broke out of the sensor so I jerked the wire out and went for a 7/8" 6 pt. impact socket and a 3/4" drive breaker bar with a 3 ft. cheater pipe. If I had a torch at the time, that would've probably worked, but instead started slipping because I needed a MAC universal inpact swivel. The angle was too great to get a solid turn on the bar. Then I went to a 22mm 6 pt. impact socket. I could keep that socket on a while before it started slipping. Then got the torch & tried heating the sensor and hammering a 21mm socket on. Couldn't get a solid blow. Yesterday was pipe wrench day! I've got a dozen pipe wrenches ranging from the 1940's era to modern design. Not enough room. Tried removing a heat shield from the firewall to gain some clearance, but the nut who designed that had steel hydraulic line routed through a channel in it. Only way that is coming out is bend and disconnect lines and looked like I would still have to cut it in half to get it out. Nothing is going to happen from underneath the car. this is purely a top-side job!

Gotta tend to some business today, before trying again. Still would like to hear of a way to bypass the need for an 02 sensor though. I have a rheostat, just need to envision this idea better. Definitely don't want to damage the manifold. That looks like the way it's headed in the process. Any 02 sensor bypass ideas?
 
#7 ·
Hi ap72. There's barely enough room to get a torch or tool in the area of the existing sensor, let alone look for another bung location. I don't care about a false reading occasionally. I'm immune to seeing the engine light on this car and it runs better without the sensor than it did with the defective sensor hooked up. My main concern in the last couple of days was to stop the exhaust leak so I could get it inspected. Anything more in the direction of a "Normal" repair is icing on the cake!, but I would still like "electrical mod" ideas for now to save time & sweat.
 
#8 ·
Well, I know it wouldn't be a good idea to take voltage from any other sensor or ECM wiring for a mod. That means running a wire from higher voltage and stepping it down from 12v to 1 volt and then having a rheostat to adjust from 1 volt to .4 or so?
 
#9 ·
F-BIRD'88 said:
If this is a 1 wire sensor:
You can mount a new sensor bung, further down the exhaust pipe where install is easier
by converting to a 3 wire or 4 wire heated type sensor.

The heater circuit just needs a simple 12v+ fused circuit switched by the ignition key.

Yes F-BIRD'88, it's a single wire sensor. Thanks for that idea, but wouldn't it be more apt to give false readings further away from the cyl. head? And where would one hook the other two wires on such a wide band sensor in an OBD1 setup?
 
#10 ·
F-BIRD'88 said:
it don;t work like that. Some people should just stay away from cars.

The ECM is getting no feedback from an 02 sensor as it now stands and for the most part it runs just fine on the preprogrammed tables on the chip. Sometimes the engine light doesn't come on, but the fuel is rich.

What about a PWM to vary the voltage in the appropriate range? I think what you are talking about would be more critical on a wide band OBD2 system. As I said, I haven't had much reason to consider the workings of an 02 sensor in the scheme of things until now; weighing my options as it were - slanting towards something less "cheater bar" fashioned.
 
#11 ·
F-BIRD'88 said:
This is not WIDE BAND it is narrow band. The only difference is the heater circuit.

it is the same narrow band sensor with a heater circuit added.

1 wire is the sensor output, the other two wires are for the heater.

The heater maintains the required minimum sensor temp so it can function without being on the ex manifold

any fused 12V circuit that is switched by the key will run the heater circuit.

You are lost on all this stuff. If you had just used a proper O2 sensor socket tool in the first place and don it on a dead cold engine, the thing would have come right out.

if the exhaust system has leaks the sensor will not function.
Read my other posts about tools and techniques used so far!
Of course you wouldn't remove a cold rusty 22 year old sensor
without some kind of tactic!
 
#12 ·
F-BIRD'88 said:
You are being an idiot. (And it won;t work) Just take the car to an exhaust shop and they will get the old sensor out for you.

Well, by you saying it can't be done is good reason to believe it can be done. Thanks for the 02 schooling though, just don't apply for a tech school teaching spot to brow beat the kiddies.
 
#13 ·
Would the voltage fluctuate that quickly? I can imagine fluctuating between readings given between 1000 degrees F and 1200., but in a closed loop idle, why such a quick fluctuation as mentioned. Yes, I am curious. That's how the human mind works sometimes. One thing I'm certain of, a muffler shop will never lay eyes on this car. Where's the learning process in that?
 
#15 ·
B-1 said:
Would the voltage fluctuate that quickly? I can imagine fluctuating between readings given between 1000 degrees F and 1200., but in a closed loop idle, why such a quick fluctuation as mentioned. Yes, I am curious. That's how the human mind works sometimes. One thing I'm certain of, a muffler shop will never lay eyes on this car. Where's the learning process in that?
The learning process is the same as DIY neurosurgery... you learn that sometimes wishing and hoping you can do something still doesn't change the fact that you can't.

If you want it done right remove the damn manifold, then replace the sensor. If you were really about the learning process then that would have been done already.
 
#17 ·
ap72 said:
The learning process is the same as DIY neurosurgery... you learn that sometimes wishing and hoping you can do something still doesn't change the fact that you can't.

If you want it done right remove the damn manifold, then replace the sensor. If you were really about the learning process then that would have been done already.

I don't need to learn what it's like to drill out and heli-coil manifold bolt holes. Thanks for the offer anyways. I know there are mature mechanics on this board. Age has nothing to do with it, but stability helps.
 
#18 ·
B-1 said:
The ECM is getting no feedback from an 02 sensor as it now stands and for the most part it runs just fine on the preprogrammed tables on the chip. Sometimes the engine light doesn't come on, but the fuel is rich.

What about a PWM to vary the voltage in the appropriate range? I think what you are talking about would be more critical on a wide band OBD2 system. As I said, I haven't had much reason to consider the workings of an 02 sensor in the scheme of things until now; weighing my options as it were - slanting towards something less "cheater bar" fashioned.
The problem w/"tricking" the ECM by feeding it a middle-of-the-road signal is the air/fuel ratio will be correct only a tiny portion of the time. I'm sure you could get the engine to run, and possibly run well enough that you don't kill it immediately- but the only sensible thing to do is remove the old one or weld a new bung and install a fresh sensor there.
 
#19 ·
cobalt327 said:
The problem w/"tricking" the ECM by feeding it a middle-of-the-road signal is the air/fuel ratio will be correct only a tiny portion of the time. I'm sure you could get the engine to run, and possibly run well enough that you don't kill it immediately- but the only sensible thing to do is remove the old one or weld a new bung and install a fresh sensor there.

Thanks for the reply cobalt327. Yes, that is the sensible and desirable thing to do. This car is a work car. I mean very often pulling a half ton load. It works so well at doing it that my F-250 is still setting in the garage stripped down to a short block waiting to be fixed. As a journal note - The car is a clunker in most peoples eyes. I looks fairly good, but it has held up to very much and will be used as long as it doesn't take up too much time to maintain. If all else fails, cobalt327, I'll do the job right, when I get the time. Too busy to get too overly involved right now, but always time to consider things and experiment. I mean, it's a 1988 Regal and there are other things to be repaired, like the house, guitar, truck, bike, customer PC's. hence the need to get this car to pass inspection and repair better when time permits. I know if I end up pulling the head to repair broken manifold bolts and stuck 02 sensor, that will tempt me to do a total rebuild. Probably nobody but me can relate to that. Thanks a bunch for your sane and proper answer!
 
#20 ·
When I get this PC job out the door and the rain passes over, I am going to get the beefier torch tip and give it another go, but I'm going to get some brazing rod too so I can legally drive it AND occasionally work on it till I get it in decent order. If time runs out on me before inspection, I'll slap the "Farm Use" magnetic signs on the door until I get a proper repair. hehe
 
#22 ·
The O2 will change the a/f ratio VERY rapidly, it's far from stationary except at WOT. You'd be better off leaving the sensor disconnected than to feed the ECM false info. That's just not gonna work. If you can braze, just fit a new bung, seal off the old sensor and be done w/it. But w/o an O2 sensor working, ain't no way you're gonna pass inspection.
 
#23 ·
cobalt327 said:
The O2 will change the a/f ratio VERY rapidly, it's far from stationary except at WOT. You'd be better off leaving the sensor disconnected than to feed the ECM false info. That's just not gonna work. If you can braze, just fit a new bung, seal off the old sensor and be done w/it. But w/o an O2 sensor working, ain't no way you're gonna pass inspection.

Here in WV, as long as the brakes, tires, lights, glass, exhaust are in tact with no body parts flapping in the wind; it will pass.
Hard to start and rough running engine raises no eyebrows. THANK GOD!
 
#25 ·
By the way, this engine runs smooth much of the time; idles easy without the 02 connected, except sometimes pulling a heavy load it will run a little rough because of the rich fuel mixture I guess. Pull in to get gas and restart to a smooth idle, zippy take-off. Haven't had any plug trouble. All of this has led me to believe that this OBD1 system is not as finnicky as I originally thought. That's where the "fooling the ECM" idea came in. I thought I might join the idea of manually leaning out the fuel mixture and injecting HHO. Mind games.
 
#26 ·
Here is how it works, without getting into too much detail:

The 02 sensor is just that, it senses how much residual oxygen in in the exhaust stream. There is a material in the sensor tip, that generates a voltage between 0v and 1v, roughly, when exposed to oxygen. The sensor has to be at or above 660* F to work, which is why the old 1 wire sensors were mounted in the manifold, so that the exhaust would heat them quickly. At .450-.500 volts, this is the "ideal" mix. Above and below, rich/lean. This cycles in the milisecond range. A fixed voltage will not work, and you cannot manually switch it fast enough.

fbird's idea with the 4 wire sensor mounted furthur down the pipe is a good one. The sensor will be every bit as accurate, likely even more so, as the original. The heated sensors were brought into use because the 1 wires had a tendancy to cool off below threshold, and drop the fuel system into open loop, increasing emissions. The solution was to add a heater circut that was ECM controlled to maintain sensor themp above threshold. In reality, all the early ECM did was turn on the 12v feed to the heater when the car was stared, and turn it off when the car was shut off, so a switched power realay feeding 12v to the heater circut does exactly the same thing.

Your methods will not work. It is time to be rational, and realize that by the time you "engineer" a complicated jimmy fix to fool the ECM, you will have spent more time and money than the car is worth. If you do not want to put in a heated sensor down stream of the original one, you need to remove the original sensor, and replace it. There is no in between.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top