3.2 CO (mix) Adjustment

[CrossT]

Got brave and bought a voltmeter to test and adjust the mix on my 86 coupe. I suspect my process is flawed here, so correction is what is needed.

I brought the engine to operating temp.

I attached the voltmeter to the O2 sensor - to the female contact (#1) and the #2 male contact.

I set the meter to 200mV and read the output of the sensor at idle.

The only readings I could get were jumping between about 50 mV and 1V. I was shooting for a jump between approx .2V and .8V. This reading was regardless of the mix screw position. I didn't seem to be impacting the voltage output of the O2 sensor by adjusting the mix screw.

Do I have this procedure correct? Or am I off somewhere?

thx for advise

todd
=-=-=-=-=-
86 cpe

[CrossT]

Ok, I am an idiot. Some people aren't meant to be trusted with this type of work. My problem was that I was connecting the voltmeter to the WRONG END (Doh!). Once I realized that, I was able to get readings that were as expected, varing between hi and low. I will mess with it some more, but was able to get it reading between about .2 and .9 volts and the thing runs. Still fluctuates a bit but runs. More fiddling to do . . .

todd
=-=-=-=-=-=-
86 cpe

edit: The turns of the screw are currently at 7 from fully in. Any indications of what is 'normal' ?

[stlrj]

Unfortunately you will not be able to set your CO with a volt meter only.
The correct way is with a CO probe inserted before the cat and the 02 sensor disconnected. Another way would be to use a scan tool and adjust your CO to maximize your crosscounts.

Joe

[CrossT]

Thanks Joe,

Looks like I'll be taking it in to a wrench. As it is now, it fluctuates at idle and frequently stalls.

I just this last weekend installed a new perf chip and cone filter/intake kit -- even more variables.

thx
todd
=-=-=-=-=-=-
86 cpe

[stlrj]

Todd.

Before you take it to the wrench, try leaning it out to eliminate the idle fluctuation and stall that running too rich will do. And not to worry, you'll never be able to run it too lean to where it would hurt anything. Most of the time they're set too rich by well meaning wrenches who never read the factory specs of less than 1% at idle to run right.

Joe

[cowtown]

Hi Todd, I set up my inital mix with an open-loop O2 sensor also, but I've just bought a Gunson Gastester to do it right, so I'm agreeing with Joe.

I wanted to let you know that you can get a Gunson for $139 through a discount at www.gwagen.com. If you feel like buying the tool instead of paying someone to do it.

Colin

[CrossT]

Thanks guys,

My current lame diagnosis

I suspect the continuing fluctuation is because I am still too rich

I think my stalling is because the idle bypass is set too low [or too far screwed in]

The condition is driving me nuts and making me not want to drive the car

more fiddling tonight . . .

todd
=-=-=-=-=-
86 cpe

[Steve W]

Todd, you measured the voltage wrong. Put the positive(red) end into the female end - the O2 sensor - and put the negative end of the meter (black) to a metal ground point of the motor. I could be the aluminum intake manifolds for example. Voltage should fluctuate evenly between 0.2 and 0.8 or a stoichiometric mix. You can adjust at the idle mixture screw in the air flow meter accordingly. I would not mess with the fuel quality switch at the dme. Ben Parrish, for example, who did have to lean out his overall mix with the switch, had to do so because he was rich running at his altitude up there at 5700', in Colorado, where the oxygen is something like 40% less than at sea level. A CO meter is great, but unfortunately not all of us have this luxury. The O2 sensor reading can give you a good reading where you are with respect to the stoichiometric ratio.

[stlrj]

Steve, the voltmeter method is flawed because you could have it set too rich and still get that lazy idle fluctuation that Todd is experiencing as the DME is trying it's best to ramp down from an over rich condition. Most EFI systems, DME included, have a much easier time ramping up from lean with increased crosscounts to where the idle fluctuations would disappear.

Joe

[CrossT]

You guys showing interest in this is much appreciated. Fortunately, these problems are clearly of a solvable nature.

I messed with it again last night. Turned the mix screw out one more turn - now out 9 turns from all the way in. I may me getting close on the fluctuation. I'll try it with the voltmeter hooked up the right way this pm and see where I am. If I can get it to stop fluctuating to an acceptable point, good enough, otherwise off to the wrench.

The stalling issue, however, is driving me nuts. I 'upped' the idle bypass screw in hopes that would cure the stalling and it did not. With contacts b&c bridged, the idle is now somewhat higher than without it being bridged. It stalls only under a certain set of circumstances. If I pull up to a normal stop there is no problem; the car acts normally and doesn't stall; the idle doesn't seem to dip low below normal idle. If I inch forward after stopping, say at the the light or in a drive thru, it dips too low and it stalls. Is there something that stands out as a cause for this?

Items I replaced a half year ago or ~2k miles [head temp sensor, O2 sensor, plugs, wires, cap, rotor]. Items I replaced last weekend [cone filter/intake, perf chip from Steve]. I must say, Steve's chip has made the car feel notably peppier and smoother across the rev band. The idle valve was replaced at 137k and the car now has 162k. Engine was rebuilt at about 153k. Alternator at 150k. No ref to the AFM ever being serviced or the position sensor or the speed sensors for that matter.

Does this narrow it down so that anything stands out?

thx,
todd
=-=-=-=-=-
86 cpe

[stlrj]

If I inch forward after stopping

There could be an issue with the vacuum line for the brake booster causing the stall.

Joe

[CrossT]

I'm getting more confused here. I hooked up the voltmeter to the O2 sensor output.
- pos to one of the female contacts [there are two]
- neg to the ground point on the intake manifold left].

I get no reading from the sensor. I attach the volymeter meter to the battery, thinking I'm either setting it wrong or it's defective, and it reads fine; in this case, 12.6v.

Defective O2 sensor? Am I doing it wrong?

the car's undriveable and I'm getting more and more confused . . .

thx for suggestions

todd
=-=-=-=-=-
86 cpe

[Steve W]

From the O2 sensor there is one black wire, and two white wires. The white wires are for the preheater element and is not what you care about in this situation. The black wire is the important sensor output. If you follow the black wire of the O2 sensor from where it's at on the cat behind the left rear wheel, through a rubber grommet in the engine sheetmetal, there will be the connector you want to measure voltage from. It's a round rubber boot, and it's the female connector within, not the male from the DME. That's were you want to measure positive voltage from, and attach the ground terminal from the meter to the car as you did to the left intake manifold. If you measure below 0.1 to 0 volts, you are idling way too lean. Above 0.8 and it's too rich. Hope that clears thing up.

[CrossT]

Thanks Steve,

You're a valuable resource here!

I still get nothing from the sensor. Sits on 0.1v or 0.0v [i.e., no sweeping at all] - regardless of the position of the mix screw in the AFM.

The sensor's less than a year old, so it seems it would be ok, but . . . .

I suppose I could buy another new one. Hmm, wonder if they are warranted.

todd
=-=-=-=-=-=-
86 cpe

[Steve W]

Hmm, maybe you do have a air leak of some sort, like the vacuum line to the brake booster as Joe suggested. You could try disconnecting it from the manifold and plugging up the hole temporarly to see if it makes a difference. Are you sure you have no other air leaks, such as the rubber lines which attach to the back of the throttle body? Excess air being sucked in at idle will cause an erratic idle, and because the mixture is so lean, you would get almost no voltage output from your sensor. Here's a couple of pics of a manifold and a throttle body showing the different hose nipples. Since your sensor is so new, I doubt it is defective. You could verify that by measuring the voltage off of it while you are driving, sensor disconnected from the DME plug. You should be able to see the voltage swing all over the place. Try the spraying the engine starting fluid at different suspicious points again, this time with the plug disconnected from the idle control valve, and the O2 sensor disconnected. Double check around the intake manifold gaskets, and if there's no rise in rpm, you are ok there. Intake manifold gaskets on a 3.2 are notorious for shrinking with age, and even getting partially sucked into the intake port. Also while you have that multimeter handy, double check that the idle positioner microswitch on the throttle body is properly engaging when the butterfly is closed. On the throttle body, you see a small black microswitch with a small red button. It tells the dme that the throttle is closed and informs the idle valve accordingly. While you are at it, check that the full throttle microswitch is also engaging. That is the big black contraption on the side. After about 80% throttle opening, it should engage and tell the dme to activate the chip's full throttle fuel and ignition maps, ignoring the O2 sensor's output, for maximum power. Adjust if necessary. About half of the 3.2s out there do not have this switch engaging at all, due to age and wear.

[Lorenfb]

The only way to set the CO is via a gas analyzer (CO machine).

The adjust screw on the AFM bottom only sets the idle CO and
not the off-idle mixture (a common misconception). The "wheel"
in the AFM must be adjusted for 1.5% before the cat. at 2000
RPM w/o the O2 sensor. I've use this method for the last 10yrs
on Porsches and BMWs with good results.

The voltage is a quick approach but not too accurate .


Good luck
Loren
Sytemsc.com

[DJB]

I recently had a similar problem, my 86 Euro spec is also chipped. It was running rich at idle and prone to cut out.

My wrench inspected the DME and found that the adjustable richness control (hope this is right - a six? position swith in the DME) was set to 4 and with an after market chip he suggested that it should be set to the lowest position. He said he changed this and the problem is 100% improved.

Hope this helps.

[Jeff Alton]

Steve, how does one check the full throttle microswitch??

Jeff

[Steve W]

Jeff, just use a continuity tester or multimeter to verify that the contacts on the microswitch are closing past around 80% of throttle opening. If you see that black module in the picture above to the left of the throttle body - that's the switch. Disconnect the plug from it and probe the contacts with your meter.

[john walker's workshop]

a bizerk voltage regulator fluctuating between 14 and 17 volts is very common on the 90A alternators installed on carreras, and will give you that same off/on idle surge, wierd running, and dying problem. 17V and DMEs don't mix well. check the charging voltage before spending a lot of time on engine adjustments. watch the meter at various RPMs.