Really REALLY Off-the-Wall Question

[Eagle]

Let's suppose that I have an '86 XJ 4-cyl with throttle body injection. Let's further suppose that for reasons not easily determined, the injection doesn't work.

Does anyone know if there is a feedback to the ECU from the injectors? Would it be possible to remove the injection, install an intake manifold from an '84 or '85 4-cyl, and run the engine on a carburetor ... leaving the ECU to control timing and ignition, but not fuel delivery?

It seems like if there is no feedback to the ECU from the injection, the ECU would not know if the injector is doing anything or even if it is missing entirely. If the ECU doesn't miss the injector, it should happily go on sending data to the distributoir and ignition even without an injector being present.

It hasn't quite reached that point, but I'm just curious if this might be possible and I wondered if anyone out there knows enough about the TBI system to tell me.

 

[Lincoln]

My sister's had a '86 MJ and it had one of those crappy computer controlled carbs. A sensor went out on it and it ran like crap. I would think if it had some sort of feedback the TB would also.

Have you check continuity between the plugs at the firewall, TB, etc? They are very prone to problems.

If I was going to pull one part I would pull them all and start from scratch. Try calling Howell Fuel Injection (might not be the exact name) they shoudl be able to give you some idea's. I believe those are the same TB that Chevy uses on their 2.8's.

Lincoln

 

[Ghost]

Eagle......

Do you have the FSM? If not I have a bastard FSM that has 84-85 stuff and notes on 86 and 87 stuff. Hell it even has info on the AW-4 that I did not realize was in there. Also were you aware that the check engine light can flash codes if you jump a wire between terminals 6 and 7? Let me know If I can look anything up or help you? As for the question about the carb swap, yes it is posiable but you have to bypas the computer I believe. But do not hold me to that info. What determined the injection does not work?

 

[Lou]

(not an expert on FI stuff, but I do know electrical)

The feedback mechanisms would be current through the injector, or voltage at the injector. (this comes from some info on minimopar)

You should be able to do this, if the driver(transistor), and the wiring to the injector are working properly.

Generally, you should be able to leave the injector connected(just not use it for fuel delivery), or connect a resistor of suitable value in its place.

The problem is if the driver, and wiring are bad, the ECU will detect a fault. How it behaves at that point is the question. Does it just record the fault, or does it affect operation?
(Minimopar seems to indicate that this would not cause a limp mode.)

 

[Eagle]

Any other ideas, folks? So far this is a hypothetical question, but may become a real question in the near future. If the problem is the injection, a replacement injector costs about as much as a new carburetor. That's what prompted this line of thought.

 

[Ghost]

Well if it is the injector......

Let me know I may have a more cost effective solution for you.

 

[BlackXJ]

Feedback to the ECU from the injectors wont be your problem, but info from the TPS and IAC on the t-body will be your problem. Still wont run.

 

[Ghost]

Now that makes sence....

The ECU knows what is going on per the TPS and the IAC. So what are the symptoms?

 

[CheapXJ]

just from having seen this XJ...

I'd go through all the wiring with a fine toothed comb, and check EVERY SINGLE WIRE that goes from the ECU to wherever for continuity and shorts.

pay the most attention to the wires that go around the back of the block, by the head, that are on the firewall.


have you tried hooking a test light to the injector to see if it's getting power?

 

[sternbal]

The only real feedback I can think of on this system is the O2 sensor.

Most of the timing info comes from the CPS.

I have a copy of the TBI manual for that system. If you need any other info, just ask. The system is relatively simple.

 

[HossHoffer]

Other way

Eagle, I went the other way, from carb to tbi. The main problem to overcome that I could see is that the ecu from the tbi adjusts the timing with readouts from the mat sensor, the cps, knock sensor, O2 sensor and the throttle position sensor. Of these the carb models only had the O2 sensor. There is no feedback from the injector itself per se only that the ecu knows how long the fuel injector is open by how long the pulse lasts. It is a fairly simple system as far as tbis go and should be easy to troubleshoot. My feelings are that it would be better and more effective to put on the Holley tbi setup rather than go back to the carb if paying for a new injector is the issue.
If the problem is the injector itself then maybe this link will help:
Rebuilt Injectors
Mine has been working well.
Just my 2 cents.

 

[PaulJ]

Generally, the engine computer looks at the O2 sensor to know if the injectors are open for too long.

 

[CheapXJ]

I'll hafta look and see if it's still layin' in the back of my parts XJ, but I do have the old throttle body from the MJ (complete w/ the injector) and it did work.

the longblock, and all the attached sensors are already gone though. If there's anything else you think you need, let me know, before it all gets trashed