Second fuel pump question

[Rickster]

This past weekend, about 40 miles from the house, I experienced vapor lock like symptoms in my 1985 CJ7 with the Mopar MPI conversion, on about 1/2 tank of gas on 95 degree day. The CJ7 would sputter, backfire through the throttle body and die at idle speeds--i.e. red lights. I would let the Jeep cool and it would start and allow me to drive it normally until the next stop, then the above would be repeated. I am going through the usual suspects--running of gas line, fuel filter, fuel pump, fuel regulator, etc. Plan on replacing fuel filter and pump. Not yet tried testing the regulator or manifold pressure as I have the battery unhooked and the fuel filter removed.

My question to you guys, and I apologize if it has been answered before but I could not find it with various searches, is that I would like to put a low psi Carter centrifugal fuel pump before the high pressure pump (mounted external) to pull fuel out of the gas tank, and I am not certain how to wire this up. Should I put a separate 30 amp relay for the Carter pump triggered off of the PCM, or get a 60 amp relay and power both pumps from the same relay in parallel? Or put a toggle on the dash and run the Carter pump continuously while driving?

I do not want to fry the PCM or damage the new high pressure pump (purchased through HESCO).

Thanks in advance.

 

[Torxhead]

My guess is that I don't think the additional fuel pump will solve your problem. You should not be having vapor lock with a fuel injected set up, i'm thinking you are not getting enough gas. Externally mounted fuel pump setups can be quite tricky, I think Ford did them only once. Have you been watching your fuel pressure by any chance?

 

[Rickster]

Thanks Torxhead.

No, I have not been checking the pressure. I just bought the CJ7 in April and really have not put many miles on it yet. Just purchased a new pressure gauge as a result of this incident and plan to monitor and check the pressure in the future. Unfortunately I did not have the gauge when the vapor lock like symptoms started.

 

[Torxhead]

You should also be getting a "Check engine light" that will contain an error code. It would help if you had the manual to go with that mpi.

 

[Rickster]

The kit is the HESCO kit (Mopar MPI). I have downloaded what information is available. Most of the information concerns installation with little to no information on the wiring diagram of the harness. HESCO offers a check engine light that you can add to your dash or you can put one in from another source. There is a wire on the harness labeled for that use. That is my next project. Then you have to cycle your ignition switch and read the flashes for the code. I have the codes that go with the OBD-I so all I need to do is hook the light up.

The dual pump idea came from someone on another board that stated several people have had success with the vapor lock like symptoms on the Mopar MPI conversion using two pumps. I think that the pump gets to hot and the fuel vaporizes. I did not have a gauge with me when I had my symptoms, but what I heard and what I saw was going on it fit with what has been described with this system in other people's CJs (my first thought was this is vapor lock like in the old carburetors of the 70s--but thought ,as you did, that it cannot be as it is EFI).

Of course, the best thing would be to have the pump in tank, but I am not ready to tackle that yet. The theory of the dual pump is that the high pressure pump for the Mopar MPI is meant to push gas, not pull. So it goes that if you put a low pressure pump in front that can pull the gas up to the high pressure pump it may take the load off of the high pressure pump and keep it from getting too hot and vaporizing the gas or perhaps decrease chance of cavitation.

Just a shot in the dark, but the Carter pump is not that expensive and thought I would give it a try. After I get it back together the plan is to take it back out on a similar outing (do not think I have to worry about the weather being any cooler around here ) and carry the gauge with me in case the symptoms happen again.

Also thinking about wiring the Carter pump 30 amp relay up to the ignition and not the PCM trigger.

Thanks for your input.

 

[Rickster]

Torxhead, also meant to say in the previous post that I know where you are coming from that the problem may not be the fuel pump and related to other issues like crank sensor, O2 sensor, vacuum leak, etc. I had similar thoughts which is why I had already researched getting a check light, finding the appropriate wire in the MPI harness and getting the OBD flash codes, just in case I did not find anything after going over the fuel line connections, proximity to exhaust, fuel filter, regulator, etc.

The thing is, after reading the post about the dual pump setup, the idea has gotten into my head and now it is an itch I've gotta scratch before I can move on. Not rational I know, but what can I say. That is why I had asked in the original post about help in wiring the second pump correctly and for safety, knowing I was going to put the second pump in not matter what (however, any and all input about the driveability issue is greatly appreciated).

As an aside, a friend at a local Jeep dealership sent me an e-mail asking if I knew what the word JEEP stood for. I thought I was about to get a lecture on the history of the Jeep, but he followed it up with this:

Just Empty Every Pocket because a Jeep is never completed, it is always under construction.

Had not heard that before, but I can relate.

 

[Torxhead]

I would think you might be able to install it wired with the other pump, if the relay can take it and the wires are big enough to carry the current. It will also be an added drain on the charging system. I think the base oem alternator was a 40 amp, but there were larger options for a/c, and a 10 gauge wire and fusible link carried the current. On my tbi injection the fuel pump is located next to the Transfer Case on the frame which puts it mid-way and at the lowest point in the system. With the pump intake line disconnected the fuel can siphon out on its own, indicating a free flowing system. A functioning fuel tank vent system is a must for the pump to operate correctly. Let us know how it works out.

 

[Sasquach]

Hesco has a forum if you are not aware and a section that discusses the MPI kit.

 

[Rickster]

Thanks Sasquach, I did not realize they had a forum. I will definitely give it a look.

 

[Rickster]

Followup on "vapor lock" issue. Bought the Carter pump but elected not to put it in yet. Installed the check engine light. Then traced the fuel line put in by the PO. Found PO ran fuel line along the exhaust on the driver's side. Also PO used worm screw clamps on the high pressure hose. Poked around the lines on the top of the tank that I could see and found dry rot. So I took down all of the lines the PO had run. Re-ran the fuel line under the radiator to passenger side and along frame. Dropped the tank and replaced sender unit, roll over valves and all lines--including fill hose and vent hose. Ran new line from charcoal canister to tank vent. Also moved the pump and regulator mid frame near Transfer Case as you suggested you did with your TBI setup (HESCO recommends cross member at top of tank but this area collects a lot of heat). Used high pressure hose clamps and double clamped where I could. Wrapped lines in engine compartment with thermal protective wrapping.

Took the CJ out this weekend. One of the hottest days we have had around here (over 100 degrees). Used the air conditioner (wife went with me) and managed to ride all afternoon with only one vapor lock like symptom a couple of miles from home on our return trip. I am pleased, as before we only drove a third of the distance and had 4 episodes of the vapor lock like symptoms.

I plan to put on a fuel line cooler on the fuel tank return line from the regulator to try and cool down the gas some before it goes back into the tank. Found a small unit that you can mount to the frame. By the way the check light did not give me any specific code after this last event.

Once again thanks for the input.

 

[Torxhead]

Thanks for the update, although the "Fuel line cooler" does sound to be a bit extreme though, if it works go with it. Not sure how your lines are but "Under the radiator" does sound like a long way around. Any vapor lock at all needs to be addressed.

 

[Rickster]

View attachment 10211View attachment 10212

Two small fluid/fuel coolers that I found. Thought I would give one a try. The Derale is smaller and has a plastic holder that protects the fins. Also, looks like it will mount easier to the frame.

Also, what pre-pump fuel filter are you using?

 

[Torxhead]

Those fuel coolers look pretty cool, no pun intended. I use a Fram hi-flow unit for the efi for my Howell system, it is quite a bit different than yours. Mine requires only around 10-12 p.s.i. fuel pressure, while yours is in the 50 or so range. I would think that kind of pressure would make for a cooler fuel system. Getting enough fuel to the pump is a critical factor, I have heard that using a 3/8" fuel pressure line from the tank to pump can help things out over the oem 5/16" size. Other than the relativly new fuel pick-up sock that I have there is no other filter in the system. I have heard of some guys going to a 3/8" return line to help out, not sure if it does, the oem size is 1/4". I would also think that if you are experiencing vapor lock, the computer would send a rich fuel mixture signal.

 

[Rickster]

I would also think that if you are experiencing vapor lock, the computer would send a rich fuel mixture signal.

Checked the check engine light again, in case I misread the light blinks (as I did not clear the codes since this incident). The check engine light did not come on during the "vapor lock" episode. Checking the codes stored in the PCM showed two codes: 12-battery disconnect and 37-Torque converter. Then I get the 55 at the end code.

 

[Rickster]

Torxhead, I think we are talking the same language but not dancing the same dance. I agree with you that in the in it is a problem of fuel starvation, but for different reasons. I do not think it is just a line diameter issue alone, as the CJ runs well up until a point, specifically in hot weather, then the engine stalls and backfires at lower rpms. It will not restart and run normally until waiting 20-30 minutes for everything to cool down. And it does not seem to happen until less than 3/4 of a tank.

Based on posting from HESCO and searching the Internet it seems to be a combination of our current state of fuel in this country and heat, specifically heated fuel going back into the tank from the regulator return line that vaporizes in the tank, or lines. That is why I want to try the fuel cooler on the return line. It is interesting to see on a search the variety of cars and years of manufacture involved that have this problem. It is not just the Mopar MPI from HESCO. But most cars are from the late 1980s early 1990s. And not just limted to multiport systems.

I am just a shade tree mechanic and do not have your expertise or experience, but one of the reasons for this post was to bring discussion and awareness to this problem. My goal is to decrease heat in the engine compartment, even if I have to resort to louvers in the hood, and if I have to, add a second lift pump to decease the work and heat generated on the external high pressure pump.

Maybe I am barking up the wrong tree, but working on th CJ sure beats healthcare! As always I really appreciate your input.

 

[Torxhead]

If your think you have a fuel delivery issue, a fuel pressure check could verify that. I think the best place to install a gauge would be as close to the throttle body as you can, I paid around $25. for mine. You seem have done some work getting the fuel cooler by rerouting the line. Another check you can do would be to see if any vacumn has developed in your fuel tank by removing the fuel filler cap and listening for a hissing noise as air enters the tank. That would indicate a poor fuel tank venting system which is critical for getting gas to the fuel pump. Sometimes it takes awhile to solve a problem doesn't it? and hopefully it will not cost too much.