You know, the first two days were like two wrestlers circling the ring, me and this car. Then they lunge for each other and the battle is on and although the tide of the battle may shift back and forth, only one will emerge the victor. In the beginning, I was determined to win.
The lumber yards around here are getting pretty good about square ends on their 2x4s. The time spent checking each one adds up to way more time than assuming they are all square and then dealing with that one out 200 that isn't.
At this point, the best way to proceed, given that the problem won't hold still but comes and goes without any discernible pattern as of yet, is to let her drive the car and see if a pattern emerges of when the sooty exhaust condition is present.
(two hours later)
I made sure codes were cleared from the computer.
I disconnected the o2 sensor and the MAP sensor and drove the car.
I pulled codes and got code 22--'MAP sensor input out of test range'.
I reconnected the MAP and o2 sensors.
Cleared codes and disconnected only the o2 sensor and drove the car.
Pulled codes and got code 11--'System pass', still with wet, sooty exhaust.
When running the KOEO test, the computer does not detect the disconnected o2 sensor.
(a half hour later) The computer does not report a fault with the o2 sensor unplugged and the engine off, but sets code 41 'No EGO/HEGO switching detected, system lean' in memory after driving a stretch in city traffic.
(another two hours later)
I pulled the plug on the computer to make it relearn everything and then drove it for twelve miles on the freeway. When I got back I ran a KOEO and a KOER and checked for fault codes in the memory.
Everything came up '11', 'system passed'.
There is no black smoke or water from the exhaust and it runs like it used to before all the problems started. I haven't pulled a plug yet. but will after I eat something. I'm certain it's a toasty tan color and nit black with soot.
My friend wants me to hang onto the car for a few days. She does not want to be the guinea pig in case it stalls. I don't blame her.
It was for a dear friend and not for money and therein lies the rub. I like a good broken-car challenge now and then, especially one that doesn't involve grease and skinned knuckles.
NOT SOLD!
My friend won't sell the car to someone if it's not working right and if it's working right she wants to keep it. If it's not working right and there is no hope for it to ever work right, she wants to junk it. She doesn't want the karma of passing on a headache to someone, even if that person knows what's up with the car.
Confirmed that the fuel pump was working by taken the cover off the fuel pump regulator and grounding the FP wire in the test plug. Gas shot out the appropriate orifice of the FP regulator. No way to test pressure except with finger. I could feel it pulse against my finger when held tight enough to block the flow and grounding and ungrounding the test wire.
Next I confirmed that the injector solenoid was working by giving it 12 volts from the battery. I could hear it click.
Tried three times without success to get the injector to shoot gas and then discovered the adjusting screw in the pressure regulator cover. First I screwed in all the way up from inside the cover (it was capped on top), put everything back together and no gas. Then I realized that I needed to *increase* pressure and screwed it back down to where it was originally. Still no gas. The I screwed it down another 1/8 of an inch or so (that's the smallest unit some carpenters will work to) to increase the pressure and put it back together, grounded the fuel pump for a few seconds and then fire the injector off the battery and got gas. Put everything back together and started the engine, which was pouring black smoke out the exhaust like it does when it's bad, but it was good enough to get it off the street and into the driveway because tomorrow is street sweeping day where I live, and they take your firstborn if you are parked on the street.
So now I am getting curious about the fuel pressure in the system and I am beginning to think about how to rig it up so I can see the fuel pressure while driving. The only thing in the fuel pressure regulator that looks like it might need changing is the diaphragm.
I still don't understand this flip-flopping back and forth between too much gas and no gas. It could very well be that two different things are going on here.
At this point I am open to any suggestions you may have about how to proceed. If it's within my ability, I will do it and report back.
When there was no fuel being injected did it set any codes ? When there was no fuel being injected did you test for injector pulse ? When there was no fuel being injected did you check for power to the fuel injector ?
The second link on that google search leads to a list of the codes and
95 sounds a lot like what aarcuda69062 mentioned about grounding issues:
"One last item; these vintages of Ford products were lousy for EEC ground problems. Since the EEC self tests by design activate certain EEC components, the self test is a good time to voltage drop the ground circuit between the ECM and battery negative."
If the fuel pump never failed to run I would ignore the code 95 for now. You can check for injector pulse with with a simple test light. I would suggest a digital VOM be used to check for the proper voltage to the injector.
I used the duty cycle function on the DVOM to check the signal to the injector and got a 2.0 reading at idle that increase to between 3.0 and
4.0 with a slight jab to the accelerator. Fuel pump is whirring when it should, when key is turned to the on position and when the running engine is turned off.
The problem as it is being manifested at present is an overly rich mix at idle that causes rough idle and black smoke and water out the tailpipe. Adjusting the pressure regulator does not visually affect the flow of gas from the injector. Also, from a newly cleared computer, with about 20 to 30 seconds of idle and one or two revs and then shut down, I got code 32, 'EGR valve not seated'. I don't know what condition sets this code. There are no wires to the EGR valve so I don't know how the computer sense that the valve is not seated.
So I am getting this alternation between too much gas, no gas and just right. It's like Goldilocks and the Three Bears.
Just to be sure, I am going to pull the EGR valve right now and check it. But I am getting increasingly suspicious of the injector as being the component that, if malfunctioning, might cause these three conditions of too much, too little and just right fuel flow.
BTW... how do you test pulse width with a test light? Do you mean a 'noid'? I don't have one, but I have the duty cycle function on the DVOM, which I bought to check the fuel mixture solenoid... oh, that's where 'noid' comes from. (light goes on).
I just hooked up a #194 bulb to the injector wiring, like aarcuda suggested, and found that, at present, there is no grounding of the circuit. 12 volts is present at the hot, but there is no path to ground. The system is back in the 'no gas at all' mode.
This narrows the problem somewhat, but I still can't imagine what would cause an all or nothing condition as far as the grounding of the signal wire to the injector goes. Just this morning gas was pouring out of the injector at idle. It's as if the signal wire from the injector to the computer is open and sometimes it gets grounded out. Except that sometimes, on rare occasions the car seems normal.
So the problem *has* to lie either with the wiring or the computer. Unfortunately, the computer was designed into the most inaccessible place imaginable, the upper driver's side, behind the dash.
Wiring checks out okay from injector plug to computer plug for injector signal wire.
If there is no other condition that would cause the computer *not* to fire the injector at startup, then I can arrive at no other conclusion except that the computer is on the fritz.
Just looked in the Chilton manual and the procedure they have for troubleshooting the injectoris as follows:
If the injector is not spraying fuel, unplug the electrical connector from the injector.
Attach a noid light to the injector electrical harness and have an assistant crank the engine. (I used a #193 bulb as per aarcuda)
If the noid light operates, replace the injector. (it didn't)
If the noid light does not operate, check the injector wiring harness for damage and repair as necessary. If the noid light still does not operate, Check the Throttle Position sensor and it's circuit.
So I guess I need to check the TPS before jumping to the conclusion that the puter is bad. I checked the TPS earlier, but I didn't check the wiring. Time to get back at it.
Yes! of course! what was I thinking (smacks forehead)
On second thought, my question was based on incorrect thinking on my part. You'll have to excuse me on that one. I was thinking that if I used a volt meter to back probe a wire going into the computer, one that is supposed to be grounded by the computer, and attached the other probe to battery negative, and got 12 volts, then, providing that the ground wire from the computer harness is intact, the computer could be at fault, in other words, the computer would not be performing its grounding function.
But that's not what you wrote.
I did a continuity test between the ground wire at the disconnected computer harness and ground at the battery. It's good.
Don't get pissed but I didn't do the voltage drop test yet. It was a struggle just to get to the computer, which is now sitting in front of me. I almost broke something in my back getting it out.
Earlier, when the injector was not spraying gas while cranking, I hooked up a #194 bulb to the wiring harness for the injector like you said and cranked. Nothing. Then I hooked up the injector directly to the battery like you said and it sprayed gas. Then I checked the hot wire at the injector harness and it had 12 volts when the key was in the run position. It remained hot for a few seconds after the key was turned to the off position. Then I checked the signal wire from the injector harness for continuity to the computer harness and it was continuous.
The only conclusion that I could come to was that *at that moment*, while cranking the engine, the computer was not grounding the injector circuit like it should.
Then I started thinking, "What if the computer is good? What else might cause the computer to act this way?" I started searching again and found this. It's dealing with EFI on Chrysler cars.
Let's go to the fundamentals before troubleshooting. On EVERY EFI engine, the computer expects to see pickup pulses (from crankshaft position sensor if DIS system or distributor if not) if it is to activate the fuel pump beyond the fuel system priming stage and to provide spark. Chrysler is not different. Here's what I'd do:
1). Pull fault codes and see what's causing the problems. You may not have any codes stored...
2). Disconnect battery for about 30 seconds and then crank the engine.
3). Pull codes; is code 12 (working from memory here...) present? This code is an "informational" code which states that the engine was not cranked since the battery was last disconnected. No distributor pickup pulses, no cranking evident to the processor ==> code 12.
4). Alternative procedure to see whether distributor pickup signals are present: listen for fuel injector cycling while cranking. *Lack of**pickup signals will result in no injector activity*.
It's this last item that got my attention. I can get an exact replacement used engine computer for as cheap as $20 plus shipping from a wrecking yard, but I don't want to start throwing parts at this car again. An ignition module and a fuel filter already cost $55 and the problem still exists. I don't want to throw a computer at it just yet.
5). Suppose your diagnosis suggests that the ECM is not seeing pickup signals. In *my* case, I'd stick an oscilloscope across the pickup and see whether I have a pulse which is not riding on a DC pedestal; i.e., it is switching cleanly between *power* ground and the voltage which is supplied to the Hall Effect pickup in the distributor. If it's riding on a pedestal, I'd locate the poor ground connection and fix it.
6). If a signal is not present, then I'd look at the pickup supply (my recollection is that is 8 volts...could be 5) WITH the pickup disconnected, I'd do a resistance test between the signal ground and power ground. If pickup power and ground are ok, then either the pickup has failed (most likely) or there's a bad electrical connection in the pickup's connector (less likely).
These last two items suggest a question to me. If the wiring is okay and the distributor is failing to send a signal to the computer, could it still be sending the signal to fire the coil? In other words, if I get spark at the coil wire and barring any wiring problems, would I be getting the proper pickup signal to the computer, also?
Then I started wondering what else might cause the computer to fail to fire the injector at startup.
And then I started wondering what else might cause the computer to fail to fire the injector at startup, and at other times cause the computer to fire the injector continuously.
For $20 plus shipping, maybe it's worth it to throw another computer at it.
"Continuity test" can mean a whole lot of different things. None of them are adequate for testing the integrity of this circuit. If the ground circuit from the ECM were broken to the point that ALL but one strand of wire was broken, you'd still see continuity as defined but that circuit could in no way flow enough current to allow proper operation.
If I was going to get pissed because you failed to do a test, I'd have been pissed a week ago because you've skipped many tests recommended. You still haven't measured fuel pressure.
Again, a continuity test is worthless. One strand of wire will show continuity, but it will not flow the more then 1 amp of current it takes to open the injector. (remember, this is a peak/hold injector)
Smart man, his presence on usenet is sorely missed.
No. If there is spark, there is a reference signal available at the ECM to trigger the injector.
Yes, the proper signal is there.
I already listed those things way back. TPS voltage too high, coolant temperature sensor voltage too low...
Not until you properly test the power feeds and grounds.
That should have been one of the first tests you performed when it wouldn't start.
The computer fires the injector by supplying a ground.
Well, the computer needs some sort of signal to tell it when to fire the injector, right ? So it could also be a bad cam or crank sensor. When you lose injector pulse do you lose spark also ?
You don't. You only need to see if the computer is suppling a ground and if you have power on the other wire. If both conditions are met the test light flashes.
I don't have the equipment to measure fuel pressure at this time. However, the fuel pump has been working flawlessly in this vehicle for about three years. And while a fuel pump that was providing twice the pressure needed would not cause the injector not to fire.
And wouldn't the pressure regulator, do a degree, take care of the excess pressure caused by a fuel pump that put out too much pressure?
Besides, the other fuel pump that goes on this vehicle, the one for multi-port injection is mounted outside the tank, not inside, like the one for single point injection, so it could not be possible to install the other FP for this vehicle.
So all these factors combined, but mostly the one about not having the equipment, have forestalled a measurement of the fuel pressure. I am not adverse to any procedure. In fact, I would love to be able to have an O-scope and other devices to play with, but I am just a humble shadetree mechanic making due with paper clips and the cheapest DVOM money can buy.
Good. I appreciate your confidence on this point. I checked and the coil is firing. But the wiring carrying the reference signal from the distributor to the computer could be bad, so I can't yet rule this out.
Let me see if I understand this correctly. If the TPS voltage is too high, the computer responds with a clear flood command and ceases to ground the injector circuit. I can understand this.
If I understand the operation of the CTS, high voltage indicates low temperature. How could this cause the computer to *completely* cease grounding of the injector?
Okay... backprobing the ground wire at the computer with the red lead of DVOM and with the black lead at battery negative, run the KOEO test. Voltage should be less than .1 volts. I'm working from memory for purpose of responding. Will check your post again before doing the actual test.
I will need to review all your posts again to see if I can determine
*exactly* what you mean by "properly test the power feeds and grounds".
I can use the procedures in the Chilton manual for testing the various sensors and their related circuitry. I will get back here with the results later in the day.
The test light didn't flash and the I have power to the red wire at the injector harness. I am convinced that the computer is not grounding the signal wire. The only question that remains is 'why not?'
Possible reasons:
Faulty computer
Some other faulty component(s) that is causing the computer not to ground the injector circuit.
Faulty wiring in the injector circuit.
I just ruled out number 3 by directly connecting the injector signal wire to the ground wire at the disconnected computer harness. With the injector harness disconnected I turned the key to 'on' and then reconnected the injector harness. The solenoid inside the injector clicked with each connect and disconnect. So using the actual wires in the injector circuit, I can get the injector to fire.
As far as number 2 goes, I believe I can rule out the TPS by simply disconnecting it. If the TPS or it's circuitry is faulty, the computer should step in to supply the missing value and the injector should fire at startup. If the injector fires at startup with the TPS disconnected and doesn't fire at startup with the TPS connected, then the TPS or its circuitry could be faulty.
Staying with number 2, the wiring that supplies the computer with pickup signals from the distributor may be faulty. The coil fires when cranking, so that part is okay.
I'm sure there are huge gaps in my knowledge of computer controlled fuel injection. I can't think of any other input to the computer that would cause it to completely cease firing the injector at startup, but those are the ones I want to concentrate on.
No, I check last night and spark was strong and bright. If the problem is with the distributor pickup signal not reaching the computer, it has to be with the wiring. This is a single injector in a throttle body so the signal comes from the pickup on the distributor. The Chilton manual I have shows four wires going from the distributor into the computer. All I can do is check them all for continuity. If they all check out, I will assume that the computer is getting the signal it needs. I have to work with the tools I have and make the most of them.
MotorsForum website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.