3.2 Can't Hold idle sounds like a cylinder missing

Remeasured everything again for greater certainty...

Test 17 show .001 volts for the key pins 35 to 9. I do not have any 9,20 or 22 pins on the connector, they are blanks. I then reseated the connector, did the procedure again, and this time got the 5 volts!

The relay 204 definitely clicks when the fuse is inserted.

The voltage from the black on the diagnostic connector to the engine block is .002 volts, so very low, but not zero, tried various engine and chassis points and got the same readings.

Key off, the Resistance from terminal 31 of 204 relay to the middle black wire on 186 is 0.7 ohms, is this low enough?

Hard to believe I finally got the 5 volts on the 35/9 pins after measuring so many times, removing and replacing the connector a number of times, but it is what it is.

The mystery remaining seems to be the 204 relay clicking when the fuse is inserted while the key is in the engine on position.

 

Those are both OK.

Agree. One thing you might try here is measuring the voltage between terminal 31 of the relay to the engine block with: 1) the key "on" and no fuse in 203, and 2) the key "on" and with the fuse in 203 to see if there is any difference. I'll be surprised if they are the same voltage, but the relay actuates in one case and not the other.

 

Under test 1) 0.11 volts
Test 2) 0.007 volts

 

Not great for test 1 -- should be lower (but still surprised that that difference causes a change in state of the relay). Would still be suspicious of the wire swaging and/or connector contact in the large connector where the black wire passes thru. You can make the same measurements there -- black wire terminal on one side to engine block (key "on" and no fuse in 203) and black wire terminal on other side to engine block (key "on" and no fuse in 203). If both of those still measure ~0.1V, then the suspicion shifts to where the large N wire bolts to the chassis.

 

For what it is worth, I think the manual test procedure needs to be followed exactly. You have the ecu connector off, with no power, do the various tests for resistance. Then you put the key in neutral, then put the connector back on. Then you put the key in the on position, then you measure for the voltage. If you just have the connector off, and put it on with the key still off, and then turn the key on, you don't get the voltage. Having the connector off first with the key off, and then put it on with the key in neutral before moving to the on position for further voltage seems essential. Not sure why this procedure would work like this, but that is how my situation has unfolded. One would think that you could run the voltage tests without removing the connector and replacing it in the sequence required. Not so.

I suppose it is possible of there simply being a bad or intermittent contact on pin 9. The connectors show no signs of damage or corrosion.

 

Hmmm. Let me test again.

 

Test 1) .10 volt
Test 2) .009 volt

Pretty much as before.

I would have thought a suspect relay, but that would not explain the fuse activation of the relay issue, that would point to some internal path within the ecu being the culprit?

 

Just saw your earlier response post, will do the ground checks.

 

There are two black wires on the connector, pins 35 (two black wires to the one pin), and pin 8 with one black wire. 0.013V and 0.015V respectively. (key to engine on, testing pin to engine block).

Where is the large N wire connection to the chassis? I know there is the big engine to chassis ground strap I have cleaned in the past. From another thread, is this the one you refer to?



(taking break so forgive any further delays in responding til very late evening. Checking the big ground will be a job for tomorrow)

 

Don't know where that 180 bolted ground connection shown on the page 60 schematic is located -- perhaps you can follow one of the five wires that go there.

 

I just did the measurements on my car. There is 5 Volt at pin 9, no fuse in 203 and if I insert a fuse the relay is not clicking.

 

Another valuable test could be to measure the voltage at pin 10. Pin 10 is connected to relay 204 (base of the transistor) and probably can trigger relay 204. Measure pin 10 in case of no fuse in 203 and in case of a fuse in 203. It might be the control unit is activating relay 204.

 

Pin 35 to Pin 10 Key engine run position

No fuse 10.5 Volts
Fuse in 0.00 Volts

Not sure how to interpret this? Relay does click every time I put the fuse in for these tests.

 

In case of no fuse the control unit activates relay 204 (secundary air injection system). This could be normal behavior but I am not sure. The air injection system is only switched on between a temperature of 15 and 31 degrees Celcius during start. In my case it was only 12 degrees hence relay 204 is not activated.
Image Unavailable, Please Login

Besides the lambda probe, the control unit only has three inputs (Engine coolant temperature, Sensor plate position and Throttle opening) and one output (Pressure actuator) which can change the mixture.
Image Unavailable, Please Login

Be sure the sensor plate position test (test step 16), Engine coolant temperature (step 4) and Pressure actuator (step 5), Throttle opening (step 7) are passed. There could be a mechanical problem with the Pressure actuator though. The throttle opening switch should be open at idle speeds, if not the control unit will enrich the mixture.

 

I think what would help is if you could get a current reading to your pressure actuator. Cold, warm, full throttle, part throttle.

 

All the test passed, but I had some inconsistency on the step 7. I will redo all the tests for greater certainty.

It looks like new pressure actuators are available, based on the Bosch number on mine. Is this unit replaceable with a simple bolt off bolt on after relieving the system pressure?

Do recall my original observation, that it appeared the car would run properly briefly when removing the main fuse box fuel pump fuse. Does this pont to any particular fuel system component as being a culprit?

 

It is hard for me o get the car to start and run properly long enough to do this type of test. The resistance test was on spec.

 

If the current to the electromagnetic pressure actuator is too high, the mixture will be too rich. If you measure the resistance of the electrical disconnected actuator, then connect it and measure the voltage. Now you can calculate the current (I = U/R). The current range is -65 mA to +138 mA.

Another test is to disconnect the electrical connector of the actuator. In that case you know the current is zero and the mixture should be around it's basic adjustment. The motor should run fine (if warmed up). If not, I assume the problem is not in the control unit.

If the system pressure is too high, the mixture will be too rich (the pressure should be around 6.3 Bar). Measuring the system pressure is a good idea.

And there is still the cold start injector which could be mechanically leaking (mine was leaking with similar problems).

In all cases removing the main fuse box fuel pump fuse will reduce the pressure leading to a leaner mixture.

 

Is your motor running good ? What happens when you disconnect your actuator.

 

When connected my motor runs excellent. When disconnected it runs pretty bad when cold, low idle and some misfiring as if the mixture is to lean (like a carburated car without a shoke). When warmed up it runs excellent again.

 

Thank you all for the further thoughts. I am awaiting some parts, and have not found a pressure test kit that will ship to Canada any time soon, so may be delayed in further diagnostics on that front. This limits some of the diagnostic tests that address the fuel pump, pressure regulator and leaking within the fuel distributor itself. I have not done the voltage tests on the mircroplex and fuel ecu's with the car running, as it is hard to get the car running without intermittent massive fuel fouling to be able to do such tests.

To summarize where this is at from a diagostic perspective:

• My ignition system does not seem to the the problem, all typical coil/wire/cap issues ruled out. The fuel pump fuse being pulled causing a brief proper running condition is further evidence that it is not a spark issue. Something is not controlling fuel correctly. Even though the full rough running and stall condition is intermittent, even when running "well" for some time the mixture appears too rich and it is not running "perfectly" as it does breifly I can tell when the fuel pump fuse is pulled.
  
• I replaced the tach sensor, and now have a new crank sensor but have not replaced as it looks pretty challenging, and I have a couple of tools on order (magnetic socket, very low profile ratchet) to be able to get access to it. The sensor in there tests fine for resistance. Unlikely problem, but will replace and rule out.
• The fuse board is not expected to be the problem, in that the problem is too much fuel. I suppose it is possible as one suggested that the fuse board overheats and cuts the pump out, which the computer then responds to the lack of pressure by overcompensating and making the mixture way too rich? Other than replacing the fuse board, or doing the pressure testing assuming I can get the car running long enough to display such a pressure reading whule the car is stumbling, not really any way to confirm this hypothesis, although as all seem to think the fuse board issues would cause a no start condition, and not actuate the fuel pump, rather than my situation where the mixture extreme richness seems to the be the problem, I view this as an unlikely source of the problem.
• The various tests that can be performed with a volt meter on the fuel ecu show all in spec, other than the tests that require the car to be running. I was getting some lack of 5 volts measures on the fuel ecu internals test, but then got the voltage after repeated testing, suggesting the connector contact was an issue. It is possible that the connector contact issue is causing the symptoms of a computer malfunction, and would also explain the imtermittent aspects. I will be carefully cleaning with deoxit the various connector pins for greater certainty.
• The possibility of a bad ground has to be investigated further, at the big red connector of the main engine harness in the right wheel well area near the parking brake mechanism - will check and clean, but that seems a long shot source of a problem, my car is garage kept, not rain driven,so I would not expect the main ground connector to be exposed to elements that would cause a bad ground.
• The mystery of the 204 relay clicking when the o2 sensor block fuse is inserted is unsolved. A new relay on order for greater certainty, and other than the bad ground suggestion or internal ecu failure this seems to be hard to pin down a cause of this anomoly.
• The fuel pressure regulator could be a problem than causes the overfueling. Absent a pressure test, a new one on order. Same with a new fuel pump. New accumulator was replaced a couple of years ago.
• The electromagnetic actuator has also been suggested as a cause. Some further investigation to be done here. The part is available.
• The possibility that the cold start injector is permanently open, even though it test properly and the solenoid actuator is functioning, is not ruled out. My cold start injector is on the front of the plennum, and may require me to remove the plennum to access replacing it. Maybe I can get at from inside the car, it is hard to get the connector on and off for volt meter tests let alone replacing the item itself. I have a new injector on order, and I expect to replace to rule out this. Since I am getting fouling accross all cylinders, I am ruling out any one specific injector failing, and do not expect to remove or replace the main fuel injectors themselves.

My goal is diagnosing to to avoid just replacing parts, although I am not averse to do this to rule things out as the cars age makes such replacements not out of order anyways.

Two questions:

1) There are suggestions that the return line to the gas tank from the pressure regulator can be clogged, and cause problems with fuel pressure. How can one check this? Can I remove the return line at the fitting, and with the gas cap off simply blow through it? How can one determine if it is clogged? Is there are gasket or crush washer required when undoing/replacing this braided flexible line?
2) If I wanted to just replace the eletromagnetic pressure actuator, is this just a bolt on item, I presume after the relieving the system fuel presure?

Thank you, Andy

 

A simple test is to disconnect the electromagnetic actuator connector. I believe it is mounted on the side of the fuel distributor. According to Jeka, the car starts rough, and then runs good after warmup. Sort of the exact opposite of what you have now.

This is not a test for your actuator, but for your injection ECU. If the motor now runs well with this part disconnected, that would point to a faulty ECU. If the motor continues to misfire and stall out, the problem is elsewhere.

On testing for clogged return lines, I would attach a temporary tube to that port, and run it into a jar to collect the fuel. If the problem goes away, the line could be removed for further testing.

The spark plugs must be clean and dry before testing.

 

+1 -- On KE3-Jet, one of the improvements they made (versus KE-Jet) is that the nominal warm-running EHA current is 0 mA. Consequently, if the system was/is properly adjusted, and the electrical side loses electrical power, warm-running is virtually unaffected. (Of course, cold start-up and cold-running will be poor, if not impossible, with the EHA disconnected.)

 

The cold start is doable with disconnected actuator. It takes a few seconds until it starts and then you have to keep it running with the accelerator. But after a minute or two it's already running good enough to start driving. After 5 minutes it runs and drives pretty normal.

 

Before I try to start the engine after changing out the fouled plugs again, I am redoing the various volt meter tests. I seem to be getting almost no voltage on test 18, which calls for 1.5 volts. This is the test on "current voltage on pressure adjuster in prestart phase". Is this a useful clue?