Hello,

in the process of getting a new to me 1972 914 1.7L with D-Jet I replaced all the rubber fuel lines (but not the plastic ones), the fuel pump (three port, with a rebuilt one) and the fuel filter. All the plumbing was done copying the original.

In the beginning I had trouble getting the fuel pump to prime, but after I left it overnight for one night the fuel started flowing well. The pressure was holding perfectly steady at 30psi +/- 1psi. (I hooked up a fuel pressure gauge to the test port on the driver side rail, and hooked up a 12 volt power source to the pump.)

All seemed well, and I started driving the car. I probably drove it for about 100 miles without problems, except that a few times it seemed to stumble on acceleration around 3000 rpm. Some days it would not do it, some days it would do it a couple of times. Once it was a bit worse, and I worried I would not make it home, but in the end it worked.

Then a couple of days ago I was driving it to the remote garage where I keep it, and it started stumbling worse than usual. I made it to the garage and parked it, thinking it will go away again. Yesterday I tried driving it again, and it was much worse, I managed to drive it out of the garage but then got stuck. It would stall, at idle and almost any other rpm; when starting it again it would run at idle for maybe 30 secs, then stall again. Trying to rev it up would not help.

So today I went again with a fuel pressure gauge, and what I saw is the following. When I just start it, the pressure goes to 30psi and stays steady for 15-30 seconds. Then it starts dropping towards 15psi, and the engine starts to die, but suddenly the pressure goes back up to 30psi and the engine runs good again. This cycle repeats itself for a few times, but the pressure drops more and more, until eventually the engine dies.

Any idea what could cause this? The fuel pump was a refurbished one bought from a respected member here (and it is not leaking, the common cause for failure). Could it be some trash in the fuel system? The fuel tank was cleaned and resealed professionally not more than a year ago. Could it be one of the lines under the fuel tank got kinked? Could the pressure regulator be going bad?

I was thinking of starting to debug this by hooking up a 12v power source to the pump directly (I was not yet able to do this at the place where the car is garaged, and I don't dare to try to drive it home), in order to see if the problem repeats itself with the engine not running. Anything else short of starting to pull off the fuel tank and the fuel pump? I'd hate to have to do this, this garage is a public one (think underground garage under big public building) and I don't feel comfortable working on the car there.

Thanks for any suggestions.

Andrei.

Who rebuilds 3 Port pumps?

Gotta go with trash in fuel. We are finding out that ethanol can eat the sealer out of the tanks. Also check pressure regulator, they rarely fail but they can stop up. What about the fuel injector rails, might want to check them too

I have a guy that rebuilds them. He has been doing them for 25+ years.

How would you check if it is trash in fuel? It is true that after I used only ethanol free fuel for the first couple of fill-ups, the last one I used regular (with ethanol) fuel.

If it is trash in the fuel, would the filter be the one that is clogged? Would it be enough to take out the fuel filter, drain what's left in the tank, put a new fuel filter, and get going? Or is there something else to do?

Also, if the fuel rails or the regulator were stopped up, would the pressure not go higher than it should be, rather than lower? My understanding of the fuel system is that the fuel pressure goes down only if something is stopped up **before** the fuel pump, is that not right? Unless there is some way in which some trash keeps the regulator open when it should not open, but to me it seems unlikely that something so big would travel past the fuel filter; but I may be wrong.

And one last question: if the whole system is clogged up from trash, do I need to also change the "sock" inside the fuel tank? If I know that I have to remove the tank anyway, I guess I'd drain it another way...

Definately check the fuel tank "sock"
Most likely that's the problem with your pressure...
It runs fine until you drive around and stur up all the trash in the tank.
You can't believe how gummed up one of them can get and how much crap can float around inside the tank. Especially if you have never cleaned and or replaced them.
Be sure to blow out both fuel lines to the tank...

In the beginning I had trouble getting the fuel pump to prime, but after I left it overnight for one night the fuel started flowing well. The pressure was holding perfectly steady at 30psi +/- 1psi. (I hooked up a fuel pressure gauge to the test port on the driver side rail, and hooked up a 12 volt power source to the pump.)


Few things to check before you start taking things apart.

a) Check for kinked hoses at the pump and filter. With the engine running wiggle the fuel pump relay around a bit to see if you get a rise/ fall in idle or feel any clicking in the relay body itself.

b) Since you already have a gauge that can be hooked up to the rail what is the pressure with the engine running? Is it constant with only small fluctuations with throttle input?

With the gauge hooked up to the rail do a simple fuel leak down test. Run the engine and check the pressure at idle. Then shut the car off and see if you get a large fast drop in pressure. It should hold pretty close to running pressure. A 2Psi drop every 5 seconds is excessive. It is normal for the pressure to bleed down over time but the 2Psi/5sec or more is the rule of thumb for bad regulator or leaking fuel injectors.

Next: This will take two people.

You already hooked up a 12 volt power source and got 30Psi. Now you need to check the voltage in the 914 wiring system.

c) Check for voltage drop at the pump ( mounted to the car using the vehicles wiring ) Similar to testing for continuity, but rather will tell you how many volts from the battery fail to arrive at the fuel pump and how many fail to make it back. Place the voltmeter dial on the volts setting. You will test both the positive and negative leads running to the pump.

If your pump is mounted in the front get a piece of wire 8-9 feet (16 or 18 gauge is fine) and two alligator clips as a jumper since the battery is in the rear. You can also perform this using a single lead of a pair of jumper cables.

Hook the jumper to the positive battery terminal. Place the positive probe of the voltmeter on the jumper that's connected to the positive battery terminal and the negative probe on the positive terminal of the pump. Start the engine for two to three seconds, read the drop on the voltmeter and turn the engine off.

Now switch the jumper to the negative battery terminal. Test the negative lead the same way, only reverse where you place the probes of the voltmeter. Touch the negative voltmeter probe on the negative jumper lead and touch the positive voltmeter probe to the negative terminal on the pump. Start the engine, read the voltmeter and shut down the engine.

Anything more than a 0.05 loss in voltage on either side of the circuit means you have a problem in the cars wiring between the battery and the fuel pump. 914 wiring is old and have seen this many times.


d) If the above checks out start at the beginning with the sock.

I think I got this sorted out. I towed the car to my house, hooked up the 12v source, and the pressure did exactly as before (started at 30psi, and dropped while the pump was running to about 10-15psi after 20-30 seconds, oscillating between 10 and 30psi). The pump was making weird noises, as if it was struggling to pump.

So I decided to pull out the tank. With the tank out of the car (but still hooked up), the pump runs fine, makes no unusual noises, and the pressure holds steady at 30psi. I let the pump run for 2-3 minutes and all was well. So I am now reasonably convinced that the supply line was kinked under the tank. (When I put the tank in I had lots of trouble not kinking it, so probably it was almost kinked, and then it must have moved and completely kinked after a week or so.)

I will now try to replace the lines from the tank with more rigid lines (high pressure fuel injection ones, I read they have less of a tendency to kink), and make them much shorter, 22in seemed to be the recommendation.

However, as I posted in a new thread, I don't seem to have a fuel sock in the tank. I am now debating whether to take the tank apart completely and install a sock, or go as is. I will probably end up ordering and installing a sock.

Now regarding the pressure drop with no power source to the pump. With the pump running I get a steady 30psi in the system. When I disconnect the power, the pressure drops almost immediately to about 20psi, and then it takes 7-8 seconds to drop to 10psi, from where it drops more slowly. This sounds to me like something is leaking in the fuel system. Which one should I suspect, injectors, cold start valve, or the regulator? How worried should I be?

I assume that if the injectors or the CSV leak, this would make the engine run very rich with bad consequences for the life of the engine and for the fuel economy. How can I test if the injectors leak? Do I take them out and put them in plastic cups, and run the pump? The CSV is harder to take out...

I add about 2’ of hose so the tank install is easier not to kink. You can also pull off the steering rack cover and look up under the tank after it’s installed.

And order the stainless lines from Tangerine. Very easy install with the tank out. Save the plastic lines for the museum.

If the injectors are bleeding out you will get smoke out of the exhaust on start up.
Replace the regulator as a precaution since it may be an issue. If you do not have a sock in the tank should get one in there. Cheap insurance for everything down the line in the system.

I'll get a sock right away, before I reinstall the tank. (See my other question about installing the sock.) I have originally made my lines 4', and that must have been too long -- I tried many times to get the tank in without kinking the lines, and did not succeed.

I don't get any smoke when starting up; but I was planning on sending the injectors out to have them cleaned up and balanced, so I might as well do that now since the fuel system is empty anyway.

If the regulator holds the correct pressure in the system well, why should it matter if it leaks when the pump is shut down? The fuel would be going back into the tank, wouldn't it? I had read that these regulator never fail...