O2 sensor replacement - ambitious for newbie?

[Ronnie's.930]

Originally Posted by aoncurly

Make sure your O2 sensor is really the root of your hunting problem. If you disconnect it and your car stops hunting, then the sensor is the problem and needs to be replaced. If it still hunts after disconnecting, it might be your fuel mixture and you need a CO analyzer for that.

No offence, but this is not 100% accurate (I know from experience) as a "too rich" condition can cause a hunting idle situation even if the O2 sensor is good - disconnecting it simply takes the O2 sensor out of the loop, but does not prove it is bad.

[ColinH]

Ronnie - so disconnecting the O2 sensor would make the engine run leaner? What you're saying (if I understand you correctly) is if this appears to solve the problem this doesn't necessarily mean the O2 sensor is bad? I think it is running rich but I have no data to support that.

I'm in the process of accumulating some tools and courage.

Cheers

[Ronnie's.930]

Colin - I am only familiar with the K-Jetronic (CIS) fuel system and I think your 86 has Motronic, correct? If so, perhaps what I am saying doesn't apply to your car. . .

With the K-Jet system, the O2 sensor reads the oxygen content in the exhaust and then the system attempts to compensate for a rich condition - this can lead to a hunting idle. When the sensor is disconnected, the K-Jet frequency valve goes into a fixed duty cycle (no longer trying to compensate) - so the rich condition is still there but the idle no longer hunts.

Again, this might be apples to oranges for your fuel system . . .

And of course, the only negative consequence of experimenting with replacing your O2 sensor, is you might find you didn't need to, and that is definitely not the end of the word (not a $$$$$ part, in other words).

[spuggy]

Originally Posted by rusnak

I think the sensor wires can not be reliably soldered. IIRC, either the heater or sensor wires must be crimped.

I thought the "no soldering" edict was odd the first time I read it, so researched it a bit.

The O2 sensor needs a reference for "fresh air" - which it apparently gets by sucking fresh air down the center of the wire. If you solder and heat shrink the splice, this will block the path of fresh air. So it will then have no reference.

From the Bosh document for the LM11 http://www.bosch.com.au/content/language1/downloads/sensors_oxygenlsm11.pdf

The Oxygen Sensor is a “Galvanic Cell”, a type of battery. The sensor
contains two porous platinum electrodes with a ceramic electrolyte
(Zirconium Dioxide) between them. The Oxygen Sensor generates a
very small voltage, ranging from as little as 100mV (0.1 volts), up to a
maximum of 900mV (0.9 volts) dependent upon exhaust gas oxygen
level. The sensor references atmospheric oxygen, typically approximately
21%, to the varying amount of excess oxygen in the exhaust. The
larger the differential in oxygen between the atmosphere and exhaust
the higher the voltage output of the sensor.

And

• The Oxygen Sensor compares the amount of oxygen in the exhaust
  gas against the amount of oxygen in the atmosphere.
  These differing amounts of oxygen will allow the sensor to produce a
  voltage output to the vehicles fuel management system.

The wire is stainless steel for "resistance to corrosion and thermal stress", according to that document.

But the installation instructions (page A-28) do describe installing via the supplied posi-lock connector - and go on to state

Important: Under no circumstances, must the cable be soldered!

[ColinH]

So a little update. My idle is pretty high (1000-1100) rpm so I realize that's an issue. However, I started it up today and got a steady idle at about 1000 rpm(sometimes it surges right from starting sometimes it doesn't) and I tried the oil cap trick to see if the rpm would drop. It did. I replicated it several times and each time the rpm dropped about 100 rpm or so. That said, there is a hose that puzzles me (this will really show my igmorance!). The part just before where the air splits into the two intake manifolds there is a hose that exits towards the front of the car (about 1cm in diametre) that is not connected to anything. It is plugged with a bolt and there's a hose clamp keeping it in place. The free end of the hose is kept in place somewhat loosley by a cable lock that's going around one of the hoses that goes into what I think is the ICV (but the hose in question is not hooked up to the ICV). Anyway, it looks weird.

Cheers

[ColinH]

Oddly, after I did the "oil cap test" my surging problems seem to have disappeared. Over the past two days I have had a constant idle. Weird, eh?

[scarceller]

If you suspect the O2 sensor is causing idle hunting just unplug the O2 sensor. These engines run fine with it unplugged. The O2 sensor and it's circuitry are only for minor mixture fine tune. The actual procedure to set mixture is with the O2 unplugged.

Idle hunting is usually the result of an air leak or improper mixture settings. So eliminate the O2 sensor from the equation by simply unplugging it.

Also be sure that the idle switch on the throttle body is working. If the switch does not close when the throttle plate returns to the idle position then the DME (computer) will not enter the idle maps.

[scarceller]

And if you plan to cut and splice the O2 sensor wire be very aware that the green signal wire is a shielded wire with a center conductor and an out braided shield. I worked on a car a few years back where someone decided to splice a after-market O2 sensor and they soldered the center wire to the shield! The end result is that they effectively grounded the signal line which tricks the DME to always see a lean mixture and then the DME just richend the &*^$ out of the mixture! It was impossible to set the base mixture! Took me 8 hours to find the problem!

[scottb]

Originally Posted by ColinH

My car is hunting at idle and I spoke to a mechanic and he suggested that the O2 sensor should be replaced.

I had a similar problem with my 84. I replaced the O2 sensor, but it didn't fix the problem. Then I had my mechanic adjust the mixture. Problem solved.

You might consider a mixture adjustment before throwing parts into the car. Easy (and inexpensive) fix.

HTH.

Scott

[rusnak]

I have an LM-1 air/fuel mixture gauge. It reads out in % Lambda.

What you will notice is that the "mixture adjustment" is taken over by the DME/ O2 sensor once the sensor heats up. All you're doing when you "adjust" the mixture is setting the baseline further or closer for the DME to adjust.

I think a faulty or worn O2 sensor is a very logical diagnosis when the mixture can be measured and is clearly rich. If lean, then you'll be suspecting either a huge vacuum leak or fuel delivery issues.

Surging idle is caused when the DME is trying to adjust fuel mixture with fuel pulses, but is stuck in a faulty feedback loop, such as a false air or vacuum leak. Or the mixture is outside of the DME's ability to correct. "Adjusting" the idle merely re-sets baseline. You'll be at this baseline when your O2 sensor is unplugged or the O2 sensor is cold. Of course, the pre-programmed map is also in effect too.

My problem with unplugging the O2 sensor on a US car is that the system is still searching for voltage from that wire. Any stray voltage can trick the DME into thinking that the car is running lean.

[scarceller]

I've done tons of tuning on the Factory DMEs and know first hand that surging idle in these engines is almost always a result of rich base mixture. I know this because I can easily force this to happen by just enriching mixture at idle, I can do this on the fly with my EPROM emulator and my laptop. It also does not need to be off by much to cause the issue. Base AFR below 13.8 often causes this issue especially at cold start.

You can even try a test yourself by simply unplugging the CHT sensor shortly after cold start. When you unplug the CHT sensor the DME sees (thinks) extremely cold CHTemp and instantly richens the mixture well below AFR=13.0 and the engine should immediately start surging.

I suggest a very simple adjustment, on the AFM turn the air screw out about 1 turn (keep track of the original position so you can put it back) with it out 1 turn you will have leaned the idle mixture and the surging may go away. But the only way to properly set base mixture is with a CO meter or a WideBand O2 if using the WideBand target AFR=14.2 at idle.

[scarceller]

Originally Posted by rusnak

My problem with unplugging the O2 sensor on a US car is that the system is still searching for voltage from that wire. Any stray voltage can trick the DME into thinking that the car is running lean.

Rus, this is not really the case. The DME circuitry does not just simply float this signal when the O2 is unpluged. The signal line has several resistors built into the circuit and the end result is the O2 signal line goes to about 5vdc via a voltage divider created out of a few resistors. Once the line goes 5vdc the DME program code completely bypasses the O2 closed loop code. The Euro DMEs work with excatly the same code as the US DMEs but the Euro DME simply hard wires this input line to 5vdc.

Bottom line is that if you unplug the O2 sensor in the US DME it runs just like a Euro DME without an O2. The only thing that could happen in a US car is that somehow the O2 signal line get accidentally grounded. So long as the O2 line does not get grounded by accident then the DME runs fine.