Phil Silano

Over the last few weeks the car ( 1983 automatic) has intermittently run poorly. Missing like crazy, with really bad vibration around 1,800 RPM. It didn't like cool, damp mornings, but was fine after sitting in the sun for several hours. But not always. Sometimes baking in the sun didn't help. Either way, sounded like electrical.

Checked the plugs, injectors (via stethoscope), coil wires, temp sensor II, air flow sensor, all resistance readings at the "brain", cleaned all the ground connections, cleaned every other connection I could find, and checked the O2 sensor. The air flow sensor readings were measured both uninstalled and at the big connector near the electrical panel. Fuel pressure is about 30.

All good, except the air flow sensor temp sensor readings. The resistance reading is low, but I seem to remember that it was always low, ~ 1.8 K ohms.

However, the fuel mixture will not adjust. The voltage off the O2 sensor will not go below .84 volts. It will go up to just over 1 volt and higher with the full throttle switch engaged. Just won't go down. Seems to be okay when connected to the computer, varies as expected.

Right now it's running fine (it's sunny today) , but that mixture thing has me stumped.

What else is there to check?

Maleficio

Did you have a good look at your distributor?

Seems like ambient temp would not affect too much considering engine heat will have more impact on engine bay components than outside air or moisture.

Bad gas?

Rich9928p

The base air/fuel mixture is set by a screw on the airflow sensor [outlined in Workshop manual 1A 24-6c and 24-24], the CO% is 0.4 - 0.8 % (measured in front of catalytic converter with oxygen sensor plug disconnected). The idle is 650 RPM +/- 50 RPM.

If the proper mixture cannot be obtained at idle there are many possible causes
- improper fuel pressure
- fuel injectors not sealing
- leaking fuel pressure regulator
- failing air flow meter
- vacuum leak in and around intake manifold
- failed temperature sensor for L-Jet fuel injection module
- failed L-Jet fuel injection module

Phil Silano

Rich,

Thanks for the reply. Here is the data for each item,

- improper fuel pressure - showing 30 psi at idle, 36 immediately after shut down. The gauge is starting to leak and has liquid spots inside the face, so not sure of the accuracy any more.

- fuel injectors not sealing - no evidence of leaks, injectors were cleaned and tested three years ago, new seals installed then.

- leaking fuel pressure regulator - No evidence yet

- failing air flow meter - resistance checks good, both hot and cold (except temp sensor, need to know if this is critical or not. I could splice a resistor in the line to check it out but would prefer some input on actual measured values first)

- vacuum leak in and around intake manifold - pulling 22 inches of vacuum and steady

- failed temperature sensor for L-Jet fuel injection module - resistance checks good

- failed L-Jet fuel injection module - Not sure what this is?

I should note that the O2 sensor reading can be driven way down by creating a vacuum leak. Pulled the connection at the 4 way near the brake booster. Voltage went below .1 volt.

jpitman2

Fuel pressure for L jet looks OK to me - 30 at idle, ~36 with vaccum disconnected - same as engine off. Temp sensor in AFM is critical to enable unit calculate mass from flow. All I can find is 2k Ohm at 'room temp', but I would be surprised if it was different from the other temp sensors used . ie, 7.5-12k at 14F, 2-3k at 68f, 250-400 ohm at 176F. To check for vacuum leaks, spray anything moderately flammable around hoses and intake fittings (including injectors), and see if the idle changes.
jp 83 Euro S AT 52k

Phil Silano

Today it ran absolutely terrible for about 15 minutes. Then, after sitting for between 10 - 20 minutes it ran almost perfect. Still missing slightly with minor vibration at 1,600 - 1,800 RPM.

Will check the distributor tomorrow.

Maleficio

I wonder if your ECU is dying?

control/alt/delete by restarting the car later, it runs good. Start it later, runs like crap.

I wonder if you're having ground issues? Every ground I clean or replace, the car runs better and more consistent. Kind of sounds like your injectors are not firing correctly, and could be caused by the ground strap on the pax head.

Reading out a ground strap with a multi-meter will tell you nothing. Just because a meter's nine volts can find it's way through a cable doesn't mean that cable can carry a load.

Have a look at the wiring repair I did a couple of days ago. My car's running tons better after this fix.

http://forums.pelicanparts.com/porsc...oday-pics.html

Phil Silano

All of the grounds have just been cleaned as part of trouble shooting this problem. Also some of the positive connections, like those at the post. No difference noted.

Wondering, how do you test a distributor cap?

jpitman2

Difficult - clean thoroughly inside and out, and the rotor. Look especially for signs of moisture, as your issue sounds a bit like that. I have had issues in the past with distributor triggers being sensitive to gap when cold (misfiring), but that was a bar magnet trigger, different from these. When its running rough, can you try to short or disconnect (CAREFULLY!) one plug at a time - the one whose disconnection does NOT make it run worse is the cylinder that is in trouble.
jp 83 Euro S AT 52k

Phil Silano

Tried disconnecting one plug at a time early in the process. No difference between any cylinder. However the spark from the coil to the distributor is yellow. I read somewhere, long ago, that a good spark should be white/blue. Is that right?

jpitman2

Generally I would expect a strong spark to be blue/white, but the colour might vary a bit due to the actual metal you were earthing it to. I think I would look for size , or 'fatness' of the spark first. Coil top is very clean, no moisture inside lead boot at either end? No signs of moisture in the plug caps or in that cavity in the head? Correct plugs fitted? Leads been tested ? Remove the leads (one at a time), unscrew the plug cap, check it (3k ohm), then with a meter on the lead, flick it around, twirl, whatever to flex it, and see if the resistance changes at all - should stay at pretty much zero.
With LH injection , I have seen one test of ECU is to disconnect AFM , and if it runs better, the ecu is dying....
Have you been able to check afm temp sensor readings ? Best I can find is pins 6 & 27 should read this - 2k ohms at 'room temp'.
jp

tailpipe

green wire going bad?