89 Escort LX 1.9L CFI -- Fuel delivery is out of wack

The continuing saga of a dear friend's 89 Escort LX 1.9L with Central Fuel Injection (it has a throttle body with a single injector)...

I have ruled out spark as the problem. Spark is good.

I caught the fuel injector in the act of not injecting fuel when it should, and all the while the coil is firing a healthy spark to all the plugs.

I replaced the fuel filter on the off chance that I was having a good day and that would solve the problem.

The gods would have none of it. I wasn't having a good day... not as far as reviving my dear friend's Escort goes. Many other things went surprisingly well. Little things, but very much appreciated.

Wiping the previous slate clean, I offer the following symptoms:

Sometimes the injector will spray gas and sometimes it won't. When the injector will not spray gas, there is fuel and pressure in the fuel pressure regulator. I confirmed this by unscrewing the fuel pressure regulator cover and watching as fuel sprayed out with surprised force and duration. Therefore I have taken it into my head that the problem must lie with either the injector, a sensor, faulty wiring or the computer. I can hear the electric fuel pump doing its thing and the fuel pressure regulator seems to be full of pressurized fuel. Whether the pressure is up to specs, I don't know, but I was mightily impressed with the force with which is sprayed out when I loosened the screws of the fuel pressure regulator cover.

Another symptom is that the engine tries to stall immediately after starting. I can keep it alive by constantly pumping the gas pedal. No one position on the gas pedal will keep the engine running, but pumping it will. But that is no way to go through life. At present, the engine seems to start consistently but will not run.

I don't have a noid light to check the signal to the injector wiring, but since the injector sometimes works I have to assume that the wiring sometimes works as well. The Chilton manual suggests that the Throttle Position Sensor plays a role in the correct operation of the fuel injection system, as does the PCM (Powertrain Control Module, or engine computer). I'll give that last one a great big 'DUH'.

The Chilton manual spells out some diagnostic procedures which I will endeavor to undertake in the coming days. However, my vast arsenal of tools does not include noid lights to check injectors as I am new to fuel injection. I may buy some... or I may not. It all depends on whether or not I solve the problem without them while I waffle on their purchase.

God, I hope to hell, for the sake of future generations, the world transits soon to electric vehicles.

Reply to
Simpson
Loading thread data ...

Noid lights are handy but they're not absolutely necessary. You could rig a #194 bulb to work the same.

Possibilities; Either the injector is being commanded to NOT allow fuel or there is an electrical problem with it. The return voltage from the TPS could be high enough that the ECM is interpreting it as a clear flood command. or The injector is shorted or open and is electrically incapable of operating.

Check the return (signal) TPS voltage, at idle position, it should be under 1.00 volts.

Measure the resistance of the injector windings, are they in spec?

Confirming fuel pressure would be a smart step also. 'squirts good when I crack a line' isn't in any manual I've ever read....

What happens if you pressurize the system up and then activate the injector with a couple of jumper leads? Does fuel spray?

What if the coolant sensor is shorted and the ECM thinks the engine is hotter than it actually is, what do you suppose the injector pulse would be, shorter or longer than needed? Do you get more fuel spray if you unplug it?

What if the MAP sensor was reading a much higher altitude than actual, shorter or longer injector pulse? Do you get more fuel spray if you unplug it?

Do you have 5 volt reference voltage at the CTS, TPS and MAP sensors?

You don't need a noid light, you need a clear path of deductive logic.

Reply to
aarcuda69062

I got the engine to the point where it blows black smoke at a rough idle and is perfectly driveable. It didn't stall in several miles of driving. I checked the TPS with a DVOM as per the Chilton manual (probably the same in the Ford service manual). It had good 5 volt Vref, but the manual said it should read 'approximately' zero at idle, but this read .76 volts. I disconnected the wiring harness from the TPS to get closer to 'approximately' zero, but this did not change the rough idle/black smoke condition. Disconnecting the CTS didn't change anything either. I didn't get around to checking the increasingly suspicious MAP sensor, but will do so in a day or two.

The single fuel injector in the throttle body was looking good today, but just too much gas at idle. Even I, who am new to EFI could see that.

More as I get to it.

And thanks...

Reply to
Simpson

As a diagnostic procedure for ruling out sensors as a cause of rough idle and black smoke at the tailpipe, how does this sound:

Disconnect the O2 sensor or the MAP sensor or the TPS, forcing the computer into limp mode, and see if this changes the rough idle and black smoke at idle. I am away from the car at the moment or I would try it now. I was just doing some google and google groups searches on this problem of a TBI engine dumping too much fuel at idle, causing rough idle and black smoke out the tailpipe and the idea of forcing the computer into limp mode came up in one discussion of this problem.

If forcing the computer into default 'limp' mode does *not* change the problem, that would seem to indicate that the problem is *not* sensor based. Does that sound like correct logic? If the problem is not sensor based, it must be component based, i.e. the injector, components that maintain fuel pressure, engine computer, vacuum hoses, etc.

I have yet to test the MAP sensor and the fuel pressure. The first is easily within my reach but the second will require some nosing around to find the "fuel pressure diagnostic valve" mentioned in the Chilton manual, but not pictured. One is asked to "Locate the fuel diagnostic valve on the fuel rail". This may be fine for the MFI, Multi-Port Fuel Injection, but this car has the CFI, Central Fuel Injection, with a single throttle body injector.

What is really discouraging is that every single discussion that I have found dealing with this kind of problem on a vehicle with fuel injection ends without resolution. Not a single one ends with the OP coming back to post what fixed the problem. I think that most people would want to come back to the discussion and crow about finding the solution. I would.

Reply to
Simpson

And if it does change, then what ??

I am away from the car at the moment or I would try

See above. If you disconnect a sensor, it puts the computer in "limp mode" and the vehicle now runs better, what does it prove ? It proves nothing ! Every time you unplug a sensor ( ANY sensor ) it should go into limp mode. When in limp mode it may not use any sensor values and run on a pre programed map in the ECU. In that instance, ant sensor you unplug will have exactly the same effect. It all depends on how the ECU is programed to operate.

No. What if the problem was a broken wire in the harness going to a sensor ? If you unplug the sensor you still have an open connection and it will still run the same.

Why not start with basic diagnostics first ? Is the check engine light on ? Did you check for stored codes, even if the light ISN"T on ? Do you have a scan tool that is capable of reading sensor values ?

Another piece of info that may be helpful, DO NOT DRIVE THE CAR if it is running that rich as you will ruin the catalytic converter, it will overheat and melt from the excess fuel.

If the problem is not sensor

Reply to
Mike

That would indicate that the problem could be the result of a malfunctioning sensor.

Since limp mode does not depend on sensor input, if the engine runs

*better* in limp mode, then that would indicate that the actually components involved in delivering fuel are working okay and one or more sensors may be at fault.

If the engine runs *the same* in limp mode, that would indicate that the components involved in delivering fuel may be faulty since both limp mode and computer control mode depend on these components.

Sure it does. It proves that the fuel pump, fuel pressure regulator, and injector are capable of functioning better than they are at present with the sensors informing the computer. If I'm not mistaken, all that is removed from the system in limp mode is sensor input, which is replaced by a computer program. Obviously, the computer needs *some* input, such as rpm, but it can do without the O2, MA, ECT and knock sensors. I'm not sure about the TPS. Does any computer ignore the TPS in limp mode?

All I am trying to accomplish by putting the computer into limp mode is to determine whether the problem is sensor based or component based. If the computer is getting bad info from a sensor and is responding accordingly, then forcing the computer into limp mode should remove that bad input and the computer will, as you point out, replace it with a program. Right now, at idle, the engine is running so badly that, if the problem is with a faulty sensor or faulty sensor wiring, I would expect to see an improvement when the limp mode program takes over.

If the car runs just as poorly in limp mode, then the fault *must* lie with the components and not the sensors, which have been removed from the system.

I will report everything.

The check engine light hasn't come on while the engine was running since I started this project. It comes on when the key is in run position and the engine is not running, so I know the light works. I disconnected the battery several times since starting work to clear the computer to see if that affected the problem. It hasn't been disconnected for the last 5 or 6 starts and having been driven for about a mile, so if codes were set, they are still in there. I just looked in the manual and all that is needed to read the codes is an analog meter, which I have. I will definitely check for fault codes.

As for the sensor values, I can check them with a DVOM. The TPS checked out okay. I plan on checking the MAP sensor next. I may be wrong, but I don't think any other sensor malfunction would cause this heavy black smoke at idle.

As this is a single injector throttle body system, I don't believe there is an Idle Air Control Valve

Reply to
Simpson

Simpson snipped-for-privacy@epoxy.com wrote in news:KjxJj.136$ snipped-for-privacy@newssvr27.news.prodigy.net:

I'm curious why you haven't responded to aarcuda. He made some awfully good suggestions in his reply to you.

Reply to
Tegger

Mid to late 80s Fords do not illuminate the check engine light like GM and Chrysler cars do.

Warm the engine up, do a "Key On Engine Off" self test. Record the codes.

Then do a "Key On Engine Running" self test. Record the codes.

An open Engine Coolant Temperature sensor certainly would.

It probably has an Idle Speed Control motor, an external device that pushes against the throttle linkage much like the early GM CCC carbed cars used.

I will now re-boot into Windoze, log onto MOD and verify.

Reply to
aarcuda69062

Well, that was painful...

The 89 Escort came with two EFI systems that operated at drastically different fuel pressures. Knowing that, I'd really-really-really want to measure the fuel pressure since it stands a good chance that the wrong pump would cause the exact symptoms you're describing.

Forget looking for a pressure test port. There isn't one on CFI. You need an adaptor.

CFI fuel pressure spec is 14.5 PSI SFI fuel pressure spec is 35-40

Looking at the wiring diagrams for this system, you should not need a analog volt meter to retrieve trouble codes. The check engine light is wired in parallel with the STO circuit so it should flash the binary codes when STI is jumped to pin 2 of the EEC test connector. (looks like you can do it either way though...) This arrangement further proves WHY the check engine light doesn't come on as expected if there is a trouble code lurking.

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.

Reply to
aarcuda69062

I *did* respond to him with what I had at that point. The post is right there dated 4/3 at 4:46 pm. I'm surprised you didn't see it. I got started on his suggestions and reported what findings I got at that point.

He wrote,

"Possibilities; Either the injector is being commanded to NOT allow fuel or there is an electrical problem with it. The return voltage from the TPS could be high enough that the ECM is interpreting it as a clear flood command. or The injector is shorted or open and is electrically incapable of operating."

I opened with "I got the engine to the point where it blows black smoke at a rough idle and is perfectly driveable. It didn't stall in several miles of driving."

That indicates that the injector was now working. There was a stretch where it was not spraying gas when the engine was cranked, but that condition changed. Don't ask me how. One theory is that the car is haunted, the previous owner is dead and wants it back. But I'm not running with that one yet. Since the engine was now running, but blowing black smoke at idle, there seemed no need to check the injector. It was obviously injecting gas and doing a good job of it once the car was moving. It drove perfectly once over 5 or 10 mph. The problem may yet be the injector, but this is as far as I have gotten with it so far

He suggested, "Check the return (signal) TPS voltage, at idle position, it should be under 1.00 volts."

I reported back, "I checked the TPS with a DVOM as per the Chilton manual (probably the same in the Ford service manual). It had good 5 volt Vref, but the manual said it should read 'approximately' zero at idle, but this read .76 volts."

He wrote, "Confirming fuel pressure would be a smart step also."

I agree, but I don't have the capacity at present to do that.

He wrote, "What if the coolant sensor is shorted and the ECM thinks the engine is hotter than it actually is, what do you suppose the injector pulse would be, shorter or longer than needed? Do you get more fuel spray if you unplug it?"

I responded, "Disconnecting the CTS didn't change anything..."

He wrote, "What if the MAP sensor was reading a much higher altitude than actual, shorter or longer injector pulse? Do you get more fuel spray if you unplug it?

I responded, "I didn't get around to checking the increasingly suspicious MAP sensor, but will do so in a day or two."

He wrote, "You don't need a noid light, you need a clear path of deductive logic."

And I am trying to develop one while I am away from the vehicle so that when I go back I can use my time more efficiently. That's why I asked the question about forcing the computer into limp mode as a means of determining whether the fault was sensor-based or component-based. It makes sense to me. Mike shot it down, but the more I thought about it, the more convinced I became that it can be a valid diagnostic technique and I plan to try it, and then focus my attention on either the sensors or the components based on the results I get. It takes about ten seconds to pop the hood and disconnect the O2 sensor and anothe ten seconds to get in and start the engine.

If no black smoke, then the problem is probably sensor-based.

If black smoke, then the problem is probably component-based, and by that I mean the actual components involved in delivering the fuel to the intake manifold, not sensors that report to the computer. This includes fuel pump, fuel pressure regulator and injector. If that is the problem, then checking the fuel pressure becomes more important.

But if I get no black smoke in limp mode, then it makes more sense to check the sensors first.

In closing, you obviously did not see my response to aarcuda or you would not have tried to bust me for not responding to him or following his suggestions.

Other than mistakenly criticizing me, do you have anything of substance to add to the actual matter at hand?

Reply to
Simpson

Will do at the first opportunity tomorrow. The Chilton manual spells out the procedure for doing this with an analog meter. I've done this in the past so it should be a piece of cake.

Then I'll add that to the list of things to check. I had a malfunction of the CTS on my 87 Dakota but the carb is no longer the feedback variety and the only thing the computer controls is the engine timing, and that stayed retarded because the computer never saw the engine as warmed up. Either that or the computer was in open loop mode. Either way, the gas mileage sucked and it nearly doubled on the freeway when I fixed the problem. The plastic wiring connector plug was broken and the circuit was open.

That's exactly what it has:

1989 Ford Escort Idle Speed Control Actuator
formatting link
?2H1455.jpg

MOD?

Reply to
Simpson

The symptoms I am describing only started a few days ago. Before that engine ran perfectly through complete range of rpms. Plugs looked perfect. Based on that I am all but certain that the fuel pump os the right one.

I came to that same conclusion. If I get to the point where checking the fuel pressure becomes the best thing to try next, I will figure out how to do it, but there are other easier test to run that will help narrow the search.

One way or another I'll run both tests.

Sorry but you lost me on that last one :-)

Reply to
Simpson

Mitchell On Demand

Reply to
aarcuda69062

Was any service work done immediately prior to when it began running bad?

Voltage drop= the proper way to measure resistance in a circuit.

Positive meter lead back probed into the ground wire at the ECM, negative meter lead connected as directly to the battery negative as possible, run the self test which will flow as much current as possible (under reasonable circumstances) thru the ground circuits, watch volt meter reading, shouldn't go above .1 volt.

Reply to
aarcuda69062

Just before I was called in, someone else replaced the timing belt and water pump. He was called in because the car stalled twice. He said it was a broken timing belt, which got me very suspicious.

  1. The engine stalled twice while the car was being driven. You can't restart an engine with a busted timing belt.
  2. The car continued to stall after the timing belt was 'replaced'.

I will have to ask my friend if anything was done to the car by anyone just prior to the stalling problem.

There are two distinct problems.

  1. The engine stalls without warning.
  2. The engine blows black smoke from the exhaust at idle due to an excessive of fuel pouring into the throttle body from the injector.

Changing the notoriously flaky ignition module did not solve the stalling problem.

The stalling problem came on first. The plugs were all toasty tan colored when I checked them and the car idled smoothly and correctly.

The black smoke from the exhaust began after I became involved. The first thing I did was remove the air filter housing and plastic intake air ducting to familiarize myself with the layout. To do this, I removed two vacuum hoses, one to the PCV valve and one to the heated air inlet diaphragm on the plastic intake air duct. Both were plugged for any engine running tests that were done with all that baggage removed. Assembly was the reverse of removal, as they say.

I can imagine the 'aha!' moment this info might impart, and believe me, I, too, suspect that it must be something that I did to cause the excessively rich idle. The vacuum hoses will all be checked again today... for the umpteenth time. I plan to crack this nut today or tomorrow or die trying. I'll report back.

Got it!

Reply to
Simpson

This sounds like one problem not 2. Very rich idle would be likely to cause stalling.

Are you saying problem 2 was not there until after you worked on it? Or that you don't know if it was there before? Are the vaccum hoses you messed with in bad shape?

-jim

----== Posted via Pronews.Com - Unlimited-Unrestricted-Secure Usenet News==----

formatting link
The #1 Newsgroup Service in the World! >100,000Newsgroups

---= - Total Privacy via Encryption =---

Reply to
jim

Have you verified that the timing belt is installed correctly, that the cam timing is correct?

Is the ignition timing correct, have the plug wires been shuffled on the distributor cap, is the mark for the distributor hold down still where it has been for the last 20 years?

Reply to
aarcuda69062

I haven't yet, but I can check those two things.

I have the car at my place now, which makes it easier to work on.

I checked for stored codes and got the following:

11 O/R/C -- System pass

I ran a KOEO test with the engine fully warmed up and got the following:

32 R/C ---- EGR valve not seated 41 C ------ No EGO/HEGO switching detected, system lean 51 O/C ---- ECT sensor input exceeds test maximum 53 O/C ---- TP sensor input exceeds test maximum 63 O/C ---- TP sensor input below test minimum 87 O/R/C -- Fuel pump primary circuit failure

O - Key On, Engine Off R - Engine running C - Continuous memory

The KEOR test requires that the engine be revved to 3000 rpm and held for the duration of the test and I want to wait a bit before destroying the tranquility of the day.

I haven't yet tested the EEC ground. That will be next

The fact that the TP sensor both exceeds test maximum and falls below test minimum is interesting.

(three hours later)

I decided to clear the computer after pulling the codes to see if they were a current condition or no longer valid. The car seemed to be running well, so I took a chance and drove three miles through city traffic and got on the freeway with it. I drove it about seven miles on the freeway and drove it back going up and down hills with lots of stop and go. Before getting on the freeway I pull the codes from the computer to see if any had set. Nothing but '11'- 'system passed'. I pulled codes again after coming back from the freeway and the hills and again, the same thing - a clean bill of health.

Yesterday my friend, who owns this car, made mention of a woman neighbor of hers who has a bit of a mental condition and mentioned that a trusted man in the neighborhood told her to get a locking gas cap for her car because this woman was capable doing something wacky like pouring a none-gasoline substance down the gas tank of anyone from whom she perceived a slight. This woman is in the process of moving from this neighborhood and this trusted man of the neighborhood feared that this woman might decide to give certain individuals a going-away present.

I don't know what effect a quart of water in a gas tank would have, but I noticed early on that water was leaking from the weep hole in the muffler at idle. The more the engine was revved, the more water came out. It went from being a steady drip to being a steady trickle. I figured it might be a blown head gasket. Their was no coolant in evidence in this drip and the cooling system is half and half. I'm not sure if coolant would be evident coming from the exhaust of a system with a blown head gasket. But I put that problem aside to concentrate on the problem of stalling, which became supplanted by the problem of the smoky, sooty exhaust.

The problem with the sooty exhaust showed up the morning of the second day from when I began working on this car, and that correlated with the super heavy gas flow coming out of the injector.

As it is now, the car is running fine, there is no smoke from the exhaust, no water is dripping from the muffler, it hasn't stalled in over twelve miles of driving in all kinds of conditions, the plugs are back to normal, there are no fault codes in memory and the idle speed is comfortably within specs and I don't know what happened.

The fuel problem went from being grossly rich to grossly lean to the point of stalling and not restarting. When the injector wouldn't shoot gas, the fuel pressure regulator had pressure that shot gas out when I backed off a screw on the cover. The next day the engine started right up and ran fine, then started smoking again. Today the engine began running fine, but later developed a bit of sooty exhaust but no where near as bad as the first day of sooty exhaust. The tank was down to less than a quarter full and my friend put in enough to get up over a half a tank and now there is no soot in the exhaust.

I haven't checked the timing, but the distributor and the head had some tick marks that lined up and when I changed the ignition module I made another pair and lined them all back up when I was done.

I am totally stumped as to what happened. All those fault codes listed above happened after the last time I cleared the computer while working on the car. I may have ran the engine with the TPS disconnected, but not the o2 sensor or the ECT sensor and whatever caused the EGR code to be set is fine now, as is the fuel pump primary circuit failure, which has me most concerned, as the fuel pump was supposed to have been changed recently and it is inside the gas tank.

At this point, I'm thinking wiring problems or the computer, but at present, it ain't broke and if it ain't broke I can't fix it and, God, I hope it don't break again.

Reply to
Simpson

Given the prior history, I can't imagine why this wasn't one of the first things done.

Certainly could cause the symptoms described.

Shorted O2 sensor, would drive the mix rich.

I gotta think you set these during your 'testing.'

Interesting...

Held at 3000 RPM? According to who? There is a point during the engine running self test when the throttle needs to be snapped, but never have I seen it required that the engine speed be held at 3000 RPM. You -really- need to find better reference material.

If you held the RPM at 3000 during any portion of the self test, that's where it came from.

Water is a natural byproduct of combustion.

One set of tic marks is from the factory, if the ignition timing is correct (as checked with a timing light, spout connector un-plugged) then it's a safe bet the cam timing is correct. But, lining up tic marks is NOT a substitute for a timing light.

You will get an ECT sensor code if the self test is run without warming up the engine. Being as you're the third cook to fry this bacon, who knows where it came from.

"Primary" may refer to the primary winding of the fuel pump relay, it wouldn't necessarily have anything to do with the fuel pump itself.

Now, since we now know that the fuel pump was recently replaced, scrutiny of the repair order for that job to determine whether the correct pump part number was used is in order.

Reply to
aarcuda69062

Given that the problem was stalling and nothing else, would it make sense to check the timing first? The guy who changed the timing belt is a full-time mechanic. I'm a carpenter. I have to assume that a full time mechanic is as good at what he does as I am at what I do and I'm pretty damn good at what I do. I'm probably better than some full time mechanics. I expect you are, too.

The black smoke came back and the plugs are full of black soot again. What I am dealing with is an intermittent condition that just won't hold still. Now it becomes a question of just how worth it is it to continue. My friend only wants a car she can rely on. She turned to me as a last resort because I once resurrected a 96 Kia Sephia that she bought for 3 grand and it was dead a week later and no one could fix it. Turned out it was a shorted hot wire to something by the gas tank and it fried the computer. It took me two months and two computers from Car-Parts.com but I tracked it down and fixed it.

I just got news that my cousin's grandson needs a cheap car to commute to his telemarketing job selling double paned windows. I think I can get $300 for this doggie.

SOLD!

Reply to
Simpson

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.