Stalls out after sitting in traffic too long

[sixsracing]

Replacing injectors one at a time is perfectly acceptable.

Voltage at the 02 sensor should be at 0. That's the way the system works, the 02 sensor adjusts voltage to lean/rich mixture and the ECM is forced to match it. Negative voltage from the O2 sensor with a positive response from the ECM.

4 black plugs means WAY too much fuel. Can you swap the fuel distributor out with a known good one? I can't imagine a temp. sensor causing this. You mentioned that you unplugged the thermo-time switch, but did you check the cold start valve to see if it was shutting off as well?

[tolusina]

...
Voltage at the 02 sensor should be at 0. That's the way the system works, the 02 sensor adjusts voltage to lean/rich mixture and the ECM is forced to match it. .........

4 black plugs means WAY too much fuel. ........

Oops.....
With the O2 sensor unplugged, voltage on the O2 signal wire from the ECU should be 0.45 VDC, call this Vref (fot reference voltage). In closed loop, the ECU varies the mixture as reported by the 02 sensor via frequency valve duty cycle changes , The ECU continually strives to keep the O2 sensor voltage as close to Vref as possible.

Since it is running far too rich, have you tried this yet? http://dana.ucc.nau.edu/~kaw5/CISidle.htm

If the duty cycle mixture adjust procedure fails to work for you, it's definitely time to check fuel pressures, system, residual, cold and warm control. You'll need a CIS specific pressure gauge set, J.C. Whitney has such a set for $50 - $70.

[gull]

See above -- I checked the pressures already, they're good. Cold start injector isn't leaking. The injector spray patterns look mostly OK, although #4 is a bit dubious. Since all four plugs are black, though, that's not my primary problem.

Since fuel pressures are good and the injectors are at least passable, my plan now is to verify the O2 sensor and Lambda system are working properly and then run through the mixture adjustment procedure you linked to. I didn't want to touch the adjustment until I'd verified everything was in proper working order, because I've learned through watching others' hard experience that adjusting the mixture to compensate for something else being wrong just results in two problems instead of one.

The car has never quite run right in the time I've owned it, so it's possible I'm seeing the cumulative effects of wear and tear and the system is just due for a mixture adjustment. The car still has the original anti-tamper plug in the fuel distributor, so it's probably never been adjusted.

[gull]

Well, I think I've pretty much got it cornered. Gather 'round and you'll hear a tale of wiring harness failures and bad test equipment.

I disconnected the O2 sensor and hooked up one DMM to it, and another to the Lambda test connector. I started the engine. The O2 sensor came to life quickly and started indicating a slightly rich condition, as expected. The duty cycle initially started out near 50%, as I'd expect in open loop mode, but when the engine warmed up a little and it decided it was time for closed loop mode, the duty cycle again dropped to 5.2%. The car started idling poorly and blowing gray smoke, and the O2 sensor voltage rose to just under 1 volt. I thumbed the full throttle switch and got a steady 36%. Wait a minute...that should be more like 60%. A little light went on -- my cheap DMM's duty cycle function was reading BACKWARDS! 5.2% wasn't the Lambda unit calling for full lean...it was calling for full RICH. No wonder the car was blowing so much smoke!

The second light went on when I started wiggling the O2 sensor connector and the car suddenly ran better for a few seconds. Hmm, interesting. I shut it down and went on a scavenger hunt for the Lambda control box, eventually finding it above the cardboard under-dash panel on the passenger side. I disconnected the connector and took off the connector cover so I could back-probe the connector. (Never wedge a test probe into the front side of an FI connector, unless you like troubleshooting bad connections.) Sure enough, pins 2 and 4, the O2 sensor signal and shield, were a dead short. Wiggling the O2 sensor end of the wire would cause it to open. I chopped off the O2 sensor connector and put a new one on, soldering it and using heat-shrink tubing. (Neatness counts with low-voltage signal wires.) I put everything back together, and suddenly the car was behaving itself.

Now that I had a functioning Lambda system, I let the engine warm up and the adjusted the mixture. I took the car for a drive and it ran great -- no more black smoke, and a smoother idle than I'm used to it having.

When I got back I let it sit and idle, to see if it would stall. It didn't, but about 15 minutes in, the main fuel pump got loud and it started misfiring on acceleration, like it was starved for fuel. In-tank fuel pump, maybe?

[gull]

Thought I'd post a little update on this one.

The car continues to run quite well except when it sits in stop-and-go traffic for long periods in hot weather. It will still occasionally cut out under those conditions. Interestingly, this sometimes occurs when the traffic clears and I accelerate away -- it'll buck and cut out a couple of times, then pick up again. I'm hoping next time to get a good look at the tach while it does it, because that'll tell me whether it's a fuel or ignition problem. The problem is so rare and intermittent, though, that it's hard to troubleshoot.

Side note: The car is due for an emissions test in February. I'm really curious if all the FI system tuning I did last summer will result in better numbers than last time.

[kamzcab86]

Check the fuel pumps.

[gull]

Check the fuel pumps.

I'm suspicious of them too, but I'm not sure how to determine whether one of them is bad, and if so, which one, given how intermittent the problem is. Fuel pressure tests OK, and the car doesn't normally have any fuel starvation symptoms, except when cornering hard with the fuel level below 1/4 tank.

I'd hate to spend $300 replacing both pumps only to have it not fix the problem...

[Briano1234]

Check the fuel pumps.

I'm suspicious of them too, but I'm not sure how to determine whether one of them is bad, and if so, which one, given how intermittent the problem is. Fuel pressure tests OK, and the car doesn't normally have any fuel starvation symptoms, except when cornering hard with the fuel level below 1/4 tank.

I'd hate to spend $300 replacing both pumps only to have it not fix the problem...

In-tank ca-ca....doo doo that is a sure sign that the pump is week, or that your hose above it is deteriorated........Had a few cars do that prior to the in-tank going out.

You would have to be 35-45 degrees to have the fuel out from under the strainer I am a thinking.

[Calimus]

It's going to be one of the pumps. As Brian can attest, I'm dealing with the same issue in the wifes rocco. This car first ate the main pump on the underside of the car. Once replaced with a spare, it ate the in tank pump which likes to die quietly and unannounced. This caused a strain on the main pump and it died as well. I replaced the in-tank with a brand new one and got another spare from Brian. Within 3 months, now the main pump is acting up again. Can't blame Brian though, 10 year old pump for free, you can only expect so much and just hope it hold up. So it looks like I'll be replacing the main pump again.

[gull]

OK, I'll plan on replacing the in-tank pump and strainer before spring, then. (I don't drive the car much in the winter.) Hopefully that'll take care of it.

I miss the way these were wired on my old Volvo 240 -- each pump had its own fuse in the fusebox. You could pull the fuses and use an ammeter to test the current draw of each pump separately, as a health check.

[tolusina]

...... The O2 sensor voltage seems to drop to nothing when I plug it into the Lambda harness.

Measure voltage on the ECU side of the O² connector, O² sensor disconnected. It should be a steady 0.45 VDC. That's the vRef (reference voltage) that the entire closed loop system strives to attain and maintain when in closed loop.

Your residual pressures seem just fine and reasonable, I'd not spend excessive time worrying them.

The "Basic Oxygen (O2) Sensor Testing" section on http://cabby-info.com/cis.htm pretty well covers lambda control function testing.

[gull]

...... The O2 sensor voltage seems to drop to nothing when I plug it into the Lambda harness.

Measure voltage on the ECU side of the O² connector, O² sensor disconnected. It should be a steady 0.45 VDC. That's the vRef (reference voltage) that the entire closed loop system strives to attain and maintain when in closed loop.

Yeah, if you read the rest of the thread you'll find that my Lambda problems turned out to be a shorted wiring harness. It took me longer than it should have to find because I misinterpreted what my DVM's duty cycle mode was telling me -- the readings were inverted compared to what's given in the manual.