twentysevenlitres

Hi guy's,
looking for some advice on my fuel problem:

Long story, so here goes...
Fuel Pump was making noise, car was hard to start cold, and engine pinged under load.

Replaced several items (filter, pressure regulator, injectors) which didn't solve problem (though regulator improved hot starting).

Decided to replace pump. Bought VDO unit from World Car Parts in the UK as most DIY mechanics on here seem to do, and gasket set from Audi.
Followed YouTube videos on the subject and got what I'm reasonably confident is a correct install.

Now the car won't start!

Assuming it was the last thing I did (pump), I rechecked all my work, and still think the job is good.
I tested the pump using a direct hotwire from the battery and it operates well and I disconnected the high pressure line at the fuel rail and ran the pump for six secs and got around 400ml of fuel into a container (about 1/2 quart I reckon that would be).
I read somewhere the pump is rated to 370 litres per hour, which gives 75ml per sec - so 400ml looks about right (6x75 = 450 - my six secs was a guess and I didn't measure the fuel!)
I have no equipment to test pressure.

Anyway after a long chat with a mechanic friend of mine, we move away from the pump on the basis that I had not heard the pump prime at all during any of my attempted starts, plus I suspect a weak battery.

So I started checking circuits.
Upon turning ignition on I checked the fuel pump fuse circuit and got only 2.5 Volts, which remained on while ignition remained on , then switched off after around 10-15 secs after turning ign off (key still in barrel). There appeared to be a latent voltage of around 0.5 - 0.8 volts with ign off.
Checked Fuel pump relay. Operates correctly from a 12V source (triggered with test leads and resistance measured across switched terminals with multimeter) and the positive feed terminal on the relay panel has 12.4 Volts (same output as battery).

I know the fuel pump fuse voltage is notoriously hard to check as some sources say the pump will only prime for about 2 secs on "initial" ignition-on and subsequent "ignition-on's" may not trigger a voltage - though my 2.5 Volts seems odd.
I have not yet checked fuse voltage whilst cranking - may try tomorrow.


Battery - after I finished installing my pump, I initially attempted a test fire only to find the battery had drained (at that point the car had not been started in 9 days) - I think the car gave me some kind of warning chime, which I didn't realise at the time. Only when I attempted a start and got nothing did I pay attention to my dim dash!

Anyway, I gave the battery around a 12 hour charge, then disconnected it when I went to work. Got home mid afternoon (about 8 hours later - work close to home!) then tried to start the car - three (maybe four) longish cranks later and the battery was flat again. Me thinks this isn't right...
Tried again today - gave the battery a good 16 hours on the charger. Lasted a bit longer. Still no start, though the engine did stumble a couple of times.

So, anyone who's still reading:

Would a bad battery prevent the fuel pump from getting sufficient power?
Is the 2.5 volts at the fuel pump fuse during ignition-on (no crank, repeated attempts) something of concern?
How do I test the "initial" ignition-on which gives the fuel pump prime?
Should I check the fuel fuse voltage during crank? (Yes...)
Should I test with my A4's newish battery?
Is there a way to reset the ECU and is this worth a shot?

Oh yes, I also scanned the ECU with VCDS lite. Only recorded fault code was for a triggered alarm on what I think it said was the central locking module. No others.


I need to buy a Bentley manual!

Thanks Guys!

silverd2

I haven't diagnosed (probably can't) your problem, but I'll try to provide a few facts that might help.

Bentley (yes, get one) says fuel volume is measured at the fuel return hose under the hood...remove fuel pump fuse (#1 on row #4), ignition off, jump 12v to the lower socket of the empty fuse receptacle. You should get minimum of 350ml of fuel in 15 secs...anything above that, volume is good.

The fuel pump relay is the central power supply for a lot of things, including the fuel pump fuse, the injectors, 02 sensors etc... So the fuse does not get full battery voltage unless the fuel pump relay is energized.

The fuel pump relay is only energized (grounded by the ECU) while the engine is turning (running or cranking) and ONLY if the engine speed sensor ("crank position") is OK, reading off the TC drive plate. The only exception is the 1st 2 seconds the ignition is turned on AND ONLY the 1st time you try it. I've found that I have to remove the key and actually lock/unlock the car to get that initial 2 seconds to reset.

Checking fuse voltage while cranking WILL indeed tell you if the relay is being energized..if it is, then the ECU is getting a signal from the engine speed sensor and all is OK there. BUT >>>

...Every time the battery is disconnected (or run below a certain voltage, I believe), the ECU has to be reset (or no start): Turn the key to ignition only, WITHOUT foot on the throttle for at least 6 seconds before any attempt to crank starter. This is straight out of the owner's manual...has to be done every time battery is reconnected.

And, yes, start with a new (known good) battery and reset the ECU as described.

Hope this helps you track it down....Don't know what the original problem was...a little late to say, but you probably didn't need to change all those components(?). But 1st, you need to get it started again, to see if you cured your original complaints.

Please keep us updated.

Cheers

PS: Been to your area years back and love it down (up) there
Hope to return some day.

twentysevenlitres

Brilliant!

That's answered a couple of problems.

I reckon I probably did undervolt the car - the battery was totally flat.
Seems to be charging ok off the mains charger, so if I can do the ECU reset procedure I might actually get it started.

I keep going back to first principles - ie. what was the last thing I touched (the pump).
But every test I perform, and everything I inspect says I've done it right.

Hope the 6 second rule works - fingers crossed!

Flow rate - I got a pdf off one of the forums which described the Audi flow rate procedure. Seems I got it almost right! Need to test it with the return line rather than the feed line! Pretty certain the new pump is good!


I like it down here too! If ever you're in town, drop me a line and I'll shout you a beer or three!

Cheers

Marty

TerminA8r

That battery is DONE! Forget recharging it, get a new battery. Clear all fault codes.
These cars starts doing weird things when "undervolted".

twentysevenlitres

OK,
Here's an interesting turn of events...

I tried the ECU reset procedure - no dice.

I tested the voltage at Fuel fuse after I tried starting it - close to 12 Volts (11.6 or something pretty high) during crank so ECU is probably OK.

So I plumbed up the Audi fuel flow procedure, this time doing it correctly both from the fuel pump fuse and using the return line after the fuel pressure regulator into the waiting container.

Nil fuel! Nada, zip!

Pump worked ok, so wiring is fine - battery was showing 12.69 Volts after test.

I could hear the fuel flowing when the pump was turned on. It had this noise with the old pump and I always thought it was a little odd - i reckon it might be air bubbles in the line. Never completely goes away, and once you shut down the pump for a couple of minutes and start again, it has the same noise.
Best description is a similar to a garden hose when it's nearly expelled all its air (not the spitting at the nozzle bit, the water flow in the hose itself).

Anyway, assuming the car is not meant to make this noise in the lines, I'm thinking the pump might not be seated properly afterall.

It's odd though, as i get fuel flow on the high pressure side of things - or at least I did yesterday!
Also, the tank is just over half full and when I had the pump out, it was just covering the bottom pick-up (where the nut for the long bolt sits and the metal/rubber gasket seals up).

So just to sum up:

Yesterday I tested pump and had fuel flow on high pressure side of fuel rail.
Today I rig up the low pressure side of the rail (after the fuel pressure regulator) and get no fuel at all in my container.
So we could assume here that we're not bulding up enough pressure (the regulator is new)

Should I pull the pump and have another go - can't deal with this today though as I'm off to work in an hour and a half...

silverd2

Wow. Just a couple of thoughts...maybe no help.

Many suggest filling the tank completely after a pump change, but if you get lots of fuel at the high pressure line, this doesn't seem like an issue to me.

Do you have your old pressure regulator? I'd consider refitting it, on the outside chance that the new one is a dud. Just a wild guess(?).

twentysevenlitres

It all worked ok before I fitted the new pump.

I'm leaning towards bad fitment of pump - only thing that makes sense to me.
The car operated (albeit long start cold, small amount of knocking under load and the pump being noisy once warmed up) before I changed it.

Last thing I touched and now car wont go!

I pretty much destroyed the old regulator getting it out, and I think I binned it.

If I filled the tank, the pump would probably work regardless - until I got down half way... Getting it to work now will prove it's mounted right.


Just on the mounting thing - I wasn't totally satisfied with how the pump finally sat in the housing. It all went in square, but looked slightly high and a test fit of the cover plate for the boot (sorry, trunk!) floor didn't sit flush.
More ammo for the "pump's not in right" theory!

This thing is a stupid design - honestly what were they thinking!

silverd2

Agreed on that philosophy...if it worked before the pump replacement, things point to the pump.
I just don't get how you seem to have fuel at the high pressure line but not the return...but I've haven't had to replace a pump in my D2 (pretty simple in other cars)...yet...I'm at about 127K miles.
A little late now, but some of the best advice I've heard is to mark the exact clock position of the original pump housing, before breaking it loose...most have stated re-installation was a breeze with marks in place...slipped right in.

"Boot"? ...understood..I speak a little Aussie
Bonnet = Hood, etc...

The toughest thing for me to "learn" down there was RHD & left-lane driving in the rental car...REALLY tough at 1st. But I found, on later trips to NZ and GB, all I had to do was flick a switch in my brain and the Aussie-planted driving "cookies" kicked right in...like riding a bike, once you've actually done it.

Anxious to hear an update.

Cheers

twentysevenlitres

There's actually a couple of arrows that line up on the pump housing - on the rear right hand side of the pump flange, just under the return hose clamp, is a diamond which lines up with a triangle on the tank top.

The biggest problem is the reservoir housing the pump has to mount to. It's a floating design which is never in the same place twice! I've been having a good read on all the forums and the number of people who thought they had it right, or even had mechanics think they had it right, only to run out of fuel at half a tank is amazing!

How they got this right at a production line level has me lost - dry tank out of the car would make it much easier though!

I'm off to buy a telescopic mirror this morning. I tried to use my mates camera scope thingy originally (I forget what they're called - small camera on the end of a flexible tube...), but it's so hard to orientate and get it where you want it!
A mirror sent down the sender tube hole will hopefully provide a reasonable viewing angle - also not worried about submerging a mirror in fuel!

I hear you LHD vs RHD - I've only driven LHD once when I was in Italy. Most confusing thing to do when driving a car! It's like trying to learn to drive again. Hired a manual (stick) too - changing gear with my right hand was an odd feeling, though left hand indicator stalks (like in Audi's!) start making perfect sense!
Even crossing the road is confusing as I'm used to looking right first!

Cheers

Marty

BrianC72gt

...properly with the assembly inside the tank. Therefore the suction hoses at the bottom of the tank aren't getting any suction. The pump does two things simultaneously: 1) it provides suction to the hoses that constantly fill the cup in which the pump sits; and, 2) it sucks fuel through the pump from the bottom of the cup. Right now, you've got 2 but not 1. Prove it? Pour a pint into the tank. It primes the cup directly. The car will start, sputter then die.

I'm also guessing you are somewhere just below a half tank of gas. The pump sits in a cup. From full to about halfway, the cup is submerged and the pump sucks fuel directly into the bottom of the pump housing. Below a half tank, it relies on a series of suction hoses that reach to the lowest points on this saddlebag tank to keep the cup full.

The way I tested mine was as follows:

Remove the banjo bolt on top of the fuel pump. That is your high pressure feed out to the fuel rail. Find a straight hose barb that matches the thread of the banjo bolt. Screw that into the hole where the banjo bolt was. Stick a couple of feet of fuel line on it secured with a decent hose clamp, and put the other end into an empty five gallon fuel jug. It doesn't need to be high pressure hose, because with no restrictions, it won't be under pressure...just fuel rated hose that fits your hose barb.

Pop the cover off the fuel pump relay (under the right side floorboard under carpet in U.S. spec cars) and close the contact points with a folded matchbook cover or the like wedged between it and the relay next to it.
Now, with the relay contacts push closed, switch the key to run (don't need to crank it) and the pump will run continuously and pump out into the jug. A helper is handy to have to switch on and off the ignition as needed.

If nothing pumps out, switch off. Pour in a a generous pint of fuel. The fuel filler neck inside the tank dumps straight into the fuel pump cup and primes it. Try to pump out again...you will get your prime back....then nothing. Fuel pump misaligned. Once you get it right, you could pump out the entire contents of the tank. Really if you can pump out a couple of gallons from where you are right now, you will know it is correct.

It has been about 6 or 7 years since I've done one of these, and my first one was a bitch because the captive nut at the other end of the foot long securing bolt had broken loose. I had to siphon the tank, fill it with water, cut out the fuel pump assembly with a foot long hole saw I welded up, cut the securing bolt in half, fish out everything, shop-vac out the water, magnet out the metal shavings, vacuum out the broken plastic, towel out every last bit of water, epoxy the captive nut on the internal housing then reassemble..

I did the second one for a friend in about an hour. Piece of cake. To install the assembly, you have to lower it and rotate it clockwise simultaneously. Ideally you want to rotate it as much as possible before you lower it. The only analogy I can think of is, it's like thrusting before ensuring proper alignment. Make sure you are aimed at the right hole before driving it home. Otherwise painful -yet still partially productive-results ensue. If you push down before rotating fully, the male portion of the fuel pump assembly will be beside - not in - the corresponding female receiver inside the tank. You can bolt it up with this misalignment, but the bolt will not go easy.

If after multiple tries, you remain unsuccessful, remove the assembly and don some fuel proof gloves with long cuffs. Tape the cuff closed. Reach in and move the assembly in the tank slightly away from the center of the car outward toward the fender. Just the slightest move will make it easier. The internal assembly floats, it is not directly secured to anything. The internal hoses hold it in place. The corrugated suction hoses inside the tank are flexible and loosely clipped to the floor of the tank. It has a little give, just a little nudge can make all the difference.

Good luck,

Brian C.