Ok Saturday coming home on the highway the car just faded out on me. Revs started falling and hitting the gas did nothing and I coasted to a stop on the shoulder. Car would crank easily but wouldn't restart. It caught once, flared to 1400rpm and died. Unplugged the MAF and cranked - same result. Timing system still fully intact. No warning lights on the dash (such as when an alternator gives out), and after a vagcom scan the only code found was a MAF code intermittent from when I unplugged it and cranked the car.

Today, under the assumption that it's fuel-related, Bill and Keith stopped by and we started troubleshooting. Fuses are fine. Pulled the fuel pump relay and applied power/ground via the battery and it snapped closed no problem. Per the inscriptions on the relay case, all terminals were properly insulated from each other, or tied together, as they should be. Took off the access cover to the tank and when the key was turned there was no audible whine from the pump. Popped off the high-pressure fuel line, attached a spare length of fuel hose and ran it to a bucket, and cranked the key - not even a drip of fuel.

Since my manual shows a direct electrical pathway from the battery, to the relay, through the fuse, to the pump, that seems to say the pump is shot.

The kicker that is throwing a wrench into it all is that measuring at the fuses, the fuel pump fuse reads 4v. The same reading is at the pump harness connector, whether or not the harness is plugged in. All other fuses show 12v. The only way this makes any sense to me is if the pump loads the system and draws more heavily. If that load is shot in my fuel pump, it may be causing the strange reading. All three of us are scratching our heads over it. A new pump arrives tomorrow and I'm going to take side by side comparative readings on them and then plug in the new pump without immersing it to see if this changes (confirming the loading issue).

I wanted to try and get a little feedback here in case anyone has seen or dealt with this before I drop a new pump in the tank to find out something else is the issue. There aren't any other components in the electrical system for fuel but these cars can be weird and anyone else that's seen this (I'm hoping) may have some insight to offer. TIA for any experiences you can share.

 

While not having any real experience with a blown fuel pump itself, I do have a few things to add.
You can jump the relay socket with a piece of wire to see if you can force the pump to kick on. This will definitely rule out a blown pump.
Also recently, my friends cbr 600 would do the exact same thing as your describing, everything was normal, it would crank, but no fuel pump priming noise. Turned out the diode and wiring for the fuel pump was fried inside the ignition switch.

 

Thanks bro, I was thinking of jumping the relay socket but didn't since the relay closed when power was applied. I didn't check power continuity out of the switched legs though so that'll get done tomorrow. I will have to look but I think I have a spare ignition switch pulled from a car last year. I'll swap that in too and try it just for kicks but wouldn't it make sense in this car that if everything else worked with the key on, that it'd be signalling the pump too? I don't know the internal workings of the switch but I thought everything went out through one or two contacts. Worth a shot for sure since it's easy to pull.

 

Yea like i said, Im not too sure about it, but i know that was what was up with the bike. Good luck and keep us posted

 

I definitely will carry this thread through to its resolution with as much detail as I can supply. I'll post up all the readings I take on my new pump as well as the old one too - maybe someone will benefit from it - as well as a flowchart of the troubleshooting. I'm going to look for that spare switch tomorrow and try that as well as jump the relay and will post up those results. Thanks again for the tips - I'll follow up with those results.

 

daym bro. i hope the pump does it for you. you might as well get a filter from API while you're at it though. bill knows the whole drill and it's much easier than you think it is.

 

Yeah the filter will get done but probably not tomorrow. I have Keith's A4 right now until mine is fixed and I don't want to hold onto it longer than I have to since he's a little leery about the timing system on the A8 so he's trying to minimize the miles he puts on it until that's done. I'd like to just get it running again for now and get his car back to him and then mess with the filter later when there's more time to take on that project.

 

Update from today:

1 - did complete fuel circuit continuity checks and everything is apparently good. Terminal 30 from the ignition switch is connected to duel pump relay terminal 30 (incidentally the fuel pump relay in my 99 30v quattro is in position 4, not 6 as I've seen noted elsewhere - it's the #372 relay). Relay output pin 87f (87) is connected to the left terminal of fuse 28 and fuse 29. Voltage measured at pin 1 of the fuel pump harness was 4.27v, same as at the fuse (connectivity in the harness between fuse panel and pump is correct). Pins 1 and 4 on the pump connector are battery and ground for the pump (2 and 3 are for the sender), and they are isolated properly. Conclusion - no wire breakage in the system.

2 - tested the relay as described above. Applying proper power and ground clamps the relay on its own.

3 - forced power to the fuel pump by pulling the relay and putting a jumper across terminals 30 and 87. It did not spin the pump (pulled the high pressure line off and ran a hose off the pump to a bucket and no fuel came out). Did the same test but attached the new pump to the power harness via jumper wires, and it spun. Conclusion - pump is shot.

A couple new ideas surfaced today. We were able to spin the new pump by jumping the named terminals, and the relay itself latches, but that omitted a very important factor - the ECU. Pin T on the relay (the skinny one, labeled as pin 3 in a Haynes manual) ties to the computer, which has the ability to shut down the pump if certain conditions exist. One of those conditions is if the engine speed sensor is bad. Supposedly this will throw a code but my car did not. But if the engine speed sensor has failed, the car will crank and crank but won't start (as mine is doing). What may be happening is the ECU sees a bad speed sensor and overrides the fuel pump relay, shutting it down. When powering up the new pump, jumping the relay terminals (and thereby taking the ECU out of the loop)
allowed it to spin, but using the relay to control it (and thereby giving control of the relay back to the ECU) resulted in no pump activation.

I haven't yet installed the new pump, in the hopes that I might not need it, but after today it's looking like my pump probably is dead and will have to be replaced. It appears that something else also is dead and the combination is what took my car down. The engine speed sensor seems likely and I'm reading up on it. The ignition switch was removed and discovered to be open between pins 30 and S when it shouldn't be. My spare ignition switch read differently and will be tried out along with a new one I ordered today. We'll see if that'll make any difference but it seems more likely that the ESS is the culprit.

Hopefully good news to follow soon. Again, all comments are welcome and encouraged.

 

Man, you are a thinker, for sure. and your theories make sense, but I think maybe you are over-thinking it, maybe.. Just based on the following, I'd say you should feel confident installing the new pump.

 

Although thats a good point about the ecu being "skipped" when jumping the relay socket, you said you tried it with both pumps, and it only worked with the new one. I would find it very hard to believe that the pump and speed sensor went out simultaneously. And that would be the only explanation. If the old pump is shot, which I think we can agree it is, then thats what caused no fuel. Not the sensor failing.
But anyhow I use that to empty out bad gas/unwanted octane gas all the time. But it doubles as a fuel pump test obv.
Glad your figuring it out.