K-Jet issue

[dmc12_uk]

I need help with a K-Jet issue that's been plaguing me for months and despite my best efforts to fix it i'm quickly running out of ideas. I've owned my car for about a year and for the first 9 months the car has run perfectly, then one afternoon I broke down on the highway (engine died and would not restart). The exhaust tips were black with soot so i pulled the Lambda sensor and the plugs, all were heavily fouled. I cleaned and reinstalled everything, got the car started and discovered a vacuum leak between the Idle Speed Regulator and the Cold Start valve. I fixed the vacuum leak and assumed all would be well again, well to put it plainly, the car hasn't run properly since this incident.

Symptom: the K-Jet will enrich the system and cause exactly the same problem as before, fouled plugs and Lambda sensor.

Diagnosis done so far:
replaced all vacuum pipes,
checked all vacuum pipes are routed and connected properly, (Including the PITA pipe in the valley),
sprayed all vacuum pipe fittings with carb cleaner to double check for leaks,
hooked up a fuel pressure gauge at the warmup regulator, measured approximately 40PSI which held for 30 minutes without dropping,
checked that the Idle and WOT switches are functioning correctly and operating as intended with the throttle spool,
cleaned the engine grounds including the ground to the frequency valve,
cleaned all the multi-pin connectors under the coil cover,
replaced the Lambda sensor and cleaned the spade connector,
cleaned the electrical connectors on the black and silver ECU boxes.

Now for the mystery! when I hook up my Actron dwell meter, start the car (AC OFF) and wait for the the system to go into closed loop, all is normal. Dwell meter reads a hair over 20 with the frequency valve buzzing away. When the Lambda warms up and the system goes into closed loop, I get a nice dwell meter needle bounce between 15 and 20. All is well so far. After about 20 minutes of idling, all of a sudden the dwell reading will hold steady at around 18, frequency valve quits making noise and it will remain this way indefinitely. Switching on the AC has no effect on the dwell reading. No slight adjustments to the mixture screw make any difference with the CO plug reinstalled and a quick blip on the throttle. If i turn off the car and let it sit for a while to cool, I can repeat the dwell test and get the same result. seems to be when it hits the 20 minute mark, something is going wrong. I considered maybe one of the ECU's has gone bad but I gather they are reliable and seldom fail?

I'm running out of ideas!

[Bitsyncmaster]

My first guess is the lambda ECU goes bad with heat. Not common for that ECU to go bad so you symptoms just suggest it.

[Ethan Rode]

My recommendation would be to make sure that the Lambda ECU isn't losing power and ground while the car is running.

Am I understanding that after going into closed loop, and after the frequency valve stops, you get a steady reading, like the ECU is in open loop? Where are you hooking up your meter? Should be the orange wire on the blue connector.

Check the wiring diagram. You'll see power is delivered to Lambda ECU on Pin 8 from the Lambda Relay, via fuse 7. Remove the cover of the ECU's connector and you'll be able to back probe the wiring. Using a voltmeter or test light, do you have power after your frequency valve quits? If no, there's an issue with the lambda relay, or the wiring to the relay or ECU. if yes, then check grounds. The ECU is grounded in three places: pin 5, 16 and the ECU's case. I believe all these are bonded on the back wall of the luggage area, above the relay compartment. Run a jumper from the negative side of the battery and jump to each of the terminals. If the frequency valve starts working, you've found a bad ground. If you've established that you have good power and grounds, and you're still getting any signal on your dwell meter, you've got a bad ECU.

[David T]

Fouled (sooted) plugs and black soot in the exhaust indicates it is running VERY rich. One possibility is the fuel plunger in the mixture unit is getting stuck. Take the mixture unit apart and make sure the plunger can move smoothly without sticking.

[dmc12_uk]

My recommendation would be to make sure that the Lambda ECU isn't losing power and ground while the car is running.

Am I understanding that after going into closed loop, and after the frequency valve stops, you get a steady reading, like the ECU is in open loop? Where are you hooking up your meter? Should be the orange wire on the blue connector.

Check the wiring diagram. You'll see power is delivered to Lambda ECU on Pin 8 from the Lambda Relay, via fuse 7. Remove the cover of the ECU's connector and you'll be able to back probe the wiring. Using a voltmeter or test light, do you have power after your frequency valve quits? If no, there's an issue with the lambda relay, or the wiring to the relay or ECU. if yes, then check grounds. The ECU is grounded in three places: pin 5, 16 and the ECU's case. I believe all these are bonded on the back wall of the luggage area, above the relay compartment. Run a jumper from the negative side of the battery and jump to each of the terminals. If the frequency valve starts working, you've found a bad ground. If you've established that you have good power and grounds, and you're still getting any signal on your dwell meter, you've got a bad ECU.

Thanks Ethan! i'm going to test these power and ground feeds over the weekend. I'll report back my findings.

[dmc12_uk]

Fouled (sooted) plugs and black soot in the exhaust indicates it is running VERY rich. One possibility is the fuel plunger in the mixture unit is getting stuck. Take the mixture unit apart and make sure the plunger can move smoothly without sticking.

Thanks David, i'm going to start by checking the power grounds to the ECU first then. If everything checks out then i've give this a try.