So I've been working diligently on cleaning up my lumpy idle. So far I've done the following:

Replace numerous vacuum lines/AFM bellows
Dipstick seal
Filler Cap Seal
Magnecor Wires
Champion Plugs properly gapped
Pertronix distributor
Pertronix 3ohm coil
New fuel injectors
Cleaned throttle body
Verified throttle body gap
Cleaned various grounds
Verified ignition timing
Set idle a touch high (900rpm)

Next was to mess around a bit with the AFM. Disconnected the O2 Sensor, turned the AFM full clockwise. WOW! Idle went crazy smooth, no variations.

Now I understand that this "fattens" the mixture with the ECU as base specs and full rich on the AFM. I also understand that the XK motor does prefer a fatter mixture.

I have a 3-wire O2 sensor on my 86 XJ6. Do you think this has something to do with the throttle switch not bypassing the O2 sensor at idle?

Any thoughts on how I should proceed as I REALLY like this nice idle!!!?!?!

Thanks
Doug

 

Yeah, non-lambda mixture control via fixed fueling 'map' rather than the mixture being trimmed by the O2 sensor





The 3-wire type is simply a heated sensor. This allows the O2 sensor to start functioning sooner, that's all. For the purposes of this discussion, 1-wire or 3-wire doesn't really matter

I'm not aware that the throttle switch is intended to bypass O2 sensor at idle.





Idle aside, how does the car drive with the O2 sensor unplugged?

What happens if you leave the O2 sensor plugged in but keep the fuel rich AFM adjustment?

Cheers
DD

 

This is the post I saw that gave me the idea that at idle and WOT, O2 feedback is supposed to be disregarded by the ECU.

JagFORUM Logon

Performance with the O2 sensor unplugged does feel a tad sluggish. Even with the AFM fully clockwise and the O2 sensor plugged in, idle isn't so great.

 

Ahhhh......yes

I'll have to go back and re-absorb all of that.....

Cheers
DD

 

If I leave the O2 sensor attached and leave at full clockwise on the AFM adjuster, the idle goes lumpy again.

Any thoughts? I'm toying with the idea of adjusting the clockspring, but I've read mixed reviews on doing that so I may even get a used AFM to mess around with.

 

Just reading my last post made me think messing with the clockspring is just masking any problems. If I'm having trouble getting it to run richer, perhaps a better start is to focus on why it's running too lean rather than not rich enough.

For starters, I'm going to replace, not just examine the last of the vacuum lines I already didn't change. I wonder if I can have a vacuum leak just small enough to still allow me to adjust idle into specs, but still having it run too lean.

One thing I forgot to mention is that this was a California car, so it's my understanding that efforts may have been made to get the car to run even leaner for emission standards.

 

I'm still actually working on this! I went ahead and replaced the remainder of vacuum, AFM, cruise, air pump related hoses under the hood just to make sure any and all air leaks wouldn't clouding my tests. My goal is to replace each rubber item under the hood and I'm just about there.

That said, I've been tinkering with this idle. I also ran a compression test, all good there, with between 145 and 150 on all cylinders (plugs looked just right). I've concluded the idle is really the smoothest (remarkably smooth at 800 rpm) at 2 turns out on the AFM with the O2 sensor disconnected. Whenever the O2 sensor comes back into play, it goes back to lumpy.

OK, while I did read the link earlier in my thread from jag-lovers about bypassing the O2 sensor at idle, however, it's a bit confusing.

What do you all think is the best and simplest path for me to get the O2 sensor ignored at idle?

 

I'm learning and interested in this.. Do you know what your fuel trims are ?

 

As in "Long Term Fuel Trim" and Short Term Fuel Trim" ?

There is no on-board monitoring of any kind on the Series IIIs so 'fuel trims'...as we think of them in modern OBD II applications.... are not in the picture.

Cheers
DD

 

OK, so when the o2 is disconnected, although the idle is great, the car bogs down at full throttle on the highway. Pulled over, reconnected, and no more bogging down.

This is interesting because, from everything I have read, the o2 sensor is ignored at WOT. I guess it can only be three things: 1. my WOT doesn't actually reach what the ECU considers WOT 2. my throttle switch is acting up. or 3. that bypass doesn't actually exist on my car. It is an 86 California car, so perhaps it was some tweaking to pass emissions.

Any ideas on bypassing the O2 at idle, or do you think I should be reviewing alternatives?

 

I don't think that's the case.

It should have the same ECU and AFM as a variety of others, including some 50-state and Canadian cars. After '83 or so all the USA cars had the air injection, air switching module, 2nd cat convertor. The throttle switch was not California (or even USA) specific



You're the pioneer on this one, I think


Cheers
DD

 

Yikes, that's a recipe for disaster!!

Prior to working on any idle bypass strategy, would you advise testing for any other causes?

 

Checked the voltage on the O2 sensor disconnected two turns out on the AFM. In open loop, it's showing .91 volts steady. That's pretty rich as I can gather. With the O2 connected, it fluctuates as expected but in the higher range mostly .75 to .8.

Is this what you guys see in open loop? Does anything seem out of range for these?

 

OK, so I installed a microswitch at the gas pedal that essentially cuts the O2 feedback at idle thereby going into open loop where my idle is smoothest. This is strictly a testing phase to see if I'm pleased with the results, so be kind on the actual install. Truthfully, my idea was from Jaguar, who installed a WOT microswitch on the throttle body for non-US/Federal models. If I'm pleased with the functionality, I'll try to come up with a more elegant solution.

Moreover, my goal is to have this as a holdover as I continue to tackle this lumpy idle. I'm just too OCD for these Jags!! :-)

 

This is basically the microswitch I used, bought it from an electronics store:

 

OK, so the above method worked in terms of forcing the car into open loop at idle, however while my idle was in its new and improved state, I now had a hesitation upon throttle application, presumably from the rich condition. It runs very rich in this state, at least from an odor standpoint.

Not thrilled with this method, I went to the throttle switch. This switch has three connections on it, from left to right: one that closes at idle, one that signals the air pump valve?, and one that closes at WOT. If you look at the connector, only the air pump connection and the WOT signal are utilized. I decided to install a jumper to tell the ECU to go into open loop (or whatever fueling the system uses at WOT).

The results were virtually the same as my microswitch, albeit a bit cleaner on install. Unfortunately, I still had the hesitation when first applying throttle off idle. I have now removed the jumper. It was fun experimenting, but I would rather fix this lumpy idle the right way, whatever that means!

To reiterate, my idle is smoothest in open loop (i.e. with no Lambda feedback), two turns counterclockwise on the AFM.

Strange enough, I opted to replace the O2 sensor to at least eliminate one more source of potential issue. That night I connected the new sensor and went on a drive. I was blown away, throughout the temp range, the car was running smoother than ever including the idle. Knowing my success has been limited to date, I decided to assume it was a fluke. Woke up the next morning, went for a drive, back to lumpy idle!!! She's trying to get the best of me, but I won't let her get me down!!

I'd love to find a way to get this idle richer without going into open loop at idle...