AHU - Air in Fuel System

[kurtzl]

Apologies in advance if my search skills are just failing to find the information I need (I can find plenty about the ALH) but I'm trying to diagnose my no-start and at the moment I'm suspicious of air in the fuel system.

Brief background: Yesterday after I drove home I pulled the timing belt top cover just for a quick visual inspection. All looked good, put it back on and didn't bother starting the car again. This morning I started the car, it ran for less than 2 seconds then died and hasn't started since.

My first thought was that in pulling the fuel lines to the side (gently) to get the TB cover out that I'd somehow caused the system to lose pressure and fuel had drained back to the tank overnight. I was about due for a new filter anyway so I swapped the filter, primed it with a MightyVac...still no start.

So then I tried priming the injection pump by disconnecting the return line that goes to injector #4 and connecting the MightyVac to the nipple on the pump. I can get fuel to come through, but only with a LOT of pumping (it doesn't hold vacuum really at all) and the air bubbles are constant.

Should I eventually see a steady stream of fuel with virtually no bubbles, or are bubbles typical? If I shouldn't be seeing bubbles, then I guess I have to do some more troubleshooting to figure out where the air is coming in from.

 

[Mongler98]

Your 109 relay is loss or bad. Wiggle it or replace it. Some air bubbles are ok. But not much. It's from the t fitting on the filter.

 

[Tdi Junkie]

Bubbles in the fuel line would definitly cause it not to start. I know its counter intuitive since you have a mity vac (I went through this same thing) but I finally got mine to start by cracking and the injectors turning it over until diesel comes out. Might want to cover them with a rag or something as it will spray everywhere.I started with them all loose then tightened all but the higher one. Recranked then tightened the last one.

Sent from my SAMSUNG-SM-G891A using Tapatalk

 

[Tdi Junkie]

I also had to replace my t fitting as well.

Sent from my SAMSUNG-SM-G891A using Tapatalk

 

[kurtzl]

Thanks for the suggestions - I considered the T fitting but I'm actually more inclined to blame the clear input line that goes to the pump. The short section of black fuel line that connects to the filter output is dry rotting and looks like it's getting damp with diesel when I try to start. I also happened to get out of the car quickly enough to hear fuel draining back to the tank after trying to start it the last time, so I'm really suspicious of an air leak now.

I actually have to start over again with troubleshooting, since I realized that my catch container wasn't holding vacuum. Just bought another one, so hopefully I'll get some more conclusive results now.

Regarding relay 109, is it responsible for anything besides the N109 solenoid? I confirmed the solenoid was getting power (~12.5V) and clicking, even tried removing the plunger as suggested by KLXD in a different thread.

 

[ToddA1]

109 also powers the ecu. If you’re seeing CEL, GP, etc with key-on bulb check, it’s good.

Clear .25 ID vinyl will work for a replacement line. It’s not diesel rated, but it’ll last around a year.

-Todd

 

[kurtzl]

Well I'm more confused now than I was before.

I do have all the lights at key-on, including GP (it comes on briefly and goes off again like I'd expect it to). I don't think it's either relay 109 or the N109 solenoid.

I managed to get proper vacuum at the injection pump return nipple and fuel comes through pretty consistently. Interestingly I seem to get more air bubbles as I increase vacuum, so that seems to suggest an air leak, but I cracked the lines at the injectors again and I'm definitely getting fuel delivered to all of them. I also haven't heard the fuel draining back to the tank again so if the air leak is still there I don't think it's nearly as bad as it was.

Now I'm getting a strange sound intermittently when I crank. Not sure what's causing it but I took a video. Unfortunately it doesn't quite come through in the video the way I hear it - in the video it sounds like a bit of a misfire (which it may be) but in person it sounds more like a slapping or snapping sound. Could it be the engine almost firing?

https://youtu.be/s86A7tTQhNc

 

[kurtzl]

BTW, thanks for the vinyl suggestion. I bought some 5/16 ID fuel line when I was at the parts place earlier (they have no metric sizes) but I'm reluctant to actually swap that factory molded line until I'm sure it's the issue. If I do end up having to pull it off it would be nice to replace it with clear so I can see what's happening.

 

[ToddA1]

I can’t hear anything in the video.

All you did was bend the IP hoses away, pull the belt cover and look at the belt? Was it running fine, before?

Are you getting smoke at the tailpipe when cranking?

-Todd

 

[Mongler98]

109 also powers the ecu. If you’re seeing CEL, GP, etc with key-on bulb check, it’s good.

Clear .25 ID vinyl will work for a replacement line. It’s not diesel rated, but it’ll last around a year.

-Todd

WHAT? no, only if the relay is TOAST, you can still have a faulty relay connection and the CEL and all that works fine. Trust me, i am to the point now that i have to wiggle it to start it every time.
The connections can get weak and only if the right ones go bad, does this symptom you speak of happen.

 

[ToddA1]

If your car relies on you wiggling the relay to start, I’d imagine it also gets flaky when running. Why would it only be a poor connection when the car is attempting to start....?

I own 3 TDIs, and the only time I experienced a no start condition and the key-on bulb check didn’t occur, was when the 109 was bad. Replacment fixed the issue and the issue hasn’t returned. A poor connection is possible, but if his N109 is getting power, I’d bet there’s nothing wrong with the relay or connection.

-Todd

 

[Mongler98]

no idea why but it runs like a champ once started, i suspect is due to ONLY ONE of the relays positions pins. been that way for years. i dont drive it much though now but when i do, have to wiggle it. i think its not a loss connection as it is more of corrosion and once started the voltage is slightly higher and has no issue with it unless trying to start. Just my hunch but the CEL does come on, thats why it took me FOREVER to figure it out, i did all the things OP did, trying to prime the fuel and such.
literally the exact same issue and things as described.

 

[kurtzl]

That's essentially all I did, but I'll back up a bit more and give the full history in case there's something there that I don't realize is relevant.

I've spent the last 4 years abroad, just moved back from Australia and got the car going again. It sat in a garage the whole time I was gone, but I did drive it for about a month when I was back for a visit about 2.5 years ago. Anyway, when I moved back I did a visual inspection, pulled the timing belt cover and rotated the engine a couple full rotations by hand, then hooked up the battery and it fired right up. Cranked for only about 1 second. Ran rough for a couple minutes but smoothed out and it's been running fine ever since - always fires up immediately.

I did standard maintenance like oil and air filter changes but didn't immediately change the fuel filter - I didn't do that until just yesterday. Did some suspension related work and brakes, but nothing to the engine.

About a month ago I noticed I was leaking coolant, and discovered that my IP was leaking diesel and had rotted the 3-way coolant hose. I swapped that hose out with a silicone replacement from Cascade German so I could continue driving while I looked for the source of the leak. Back in 2006 I had replaced the top cover and QA seals and neither of them appeared to be leaking, but I figured with sitting for so long some of the other pump seals had started to leak.

After replacing coolant hoses and cleaning everything up I tried to identify the source of the leak, but didn't have any success. It looked like it might have been leaking somewhere on the back of the pump (the side facing the cylinder head) but even with a mirror it was hard to tell. The leak was very slight, and I wasn't even convinced it was still leaking (maybe the seals swelled up again after running the pump for a while?). Fast forward a couple weeks - life gets busy, I'm working from home and hardly driving so it takes me a while to get back to it.

Anyway, I still wanted to check and make sure the IP shaft seal wasn't leaking diesel on the timing belt, so Friday when I got home from work (car ran fine, drove about 70 miles that day) I pulled the timing belt cover to have a look. Had to remove the airbox cover and hose, and had to pull the supply and return lines out of the way to get the cover off. Confirmed there was no trace of diesel and the belt appeared in good shape, so I put the TB cover back on, put the airbox cover back on, re-routed a vacuum line that was rubbing on the airbox hose and left it at that.

Yesterday morning I got in my car to leave and it started right up like it always does, but then it died immediately. Didn't run for even 2 seconds. I tried a couple times to start it again but it wouldn't go so I popped the hood. I immediately suspected fuel since I had to pull those lines to the side, and I noticed 2 things: (1) the hose clamps on the supply and tank lines were only finger tight (2) there was diesel on the fender liner that had run down onto the A/C hoses that route through there.

At this point I was certain it was lost prime in the fuel system, so I decided to go ahead and swap the filter while I was at it. I also wanted to clean up the diesel from the A/C hoses so they wouldn't get destroyed, so I pushed the car out of the garage so I could give the lines a bit of a soap and rinse. Swapped the filter, filled it from the pump with my MightyVac, cleaned and reconnected everything - no start. No problem, just need to prime the pump too. Pulled the return line, hooked up the MightyVac and applied vacuum until I was getting fuel, then hooked that back up too. Still nothing.

And that's where I am now - I've spent probably 15 hours over the past 2 days trying to figure it out with no success. And my driveway goes up a little hill into the garage, so my car is also stuck outside

 

[Abacus]

Hook up a jug full of fuel hanging from the hood, so it bypasses the fuel filter. You can run the return to it or back to the tank directly, bypassing the filter (so you don’t get covered in diesel). This will eliminate the fuel filter and allow self priming of the fuel system to rule it out, and it’ll also show if you have a crack in the clear lines, which I doubt since they’re so robust. I do this for a Diesel Purge.

Any white smoke coming out the exhaust?

 

[kurtzl]

Sorry, I neglected to answer the smoke question.

Yes, I'm getting smoke at the exhaust, though I would have described it as bluish. Also, I'm not sure how else to describe the intermittent sound I tried to capture in the video - maybe a knock, or a cough? It's irregular, so I don't think it's mechanical (valves or pistons).

Bypassing the filter sounds like a good troubleshooting step, but let me make sure I understand the setup you have in mind.

The clear IP input line is still connected to the pump with the factory crimp clamp, so I'd run a hose from the fuel jug to the filter end of that clear IP line. The return line is also factory crimped, so I'd disconnect it from the tee and run another line back to the jug. Is that right? I'll make sure my jug is clean and filled with clean fuel (via MightyVac through the filter) but how much should I plan to have in the jug in order for this to work? Will a pint of diesel be enough, or will I need more than that?

 

[ToddA1]

The more fuel in the jug, the better. You’re relying on the weight of the fuel to gravity feed the IP.

Jug to the IP, bypassing the filter. Disconnect the return at the filter and install a coupler, returning fuel to the tank.

If it doesn’t start, I’d suggest scanning the car.

-Todd

 

[kurtzl]

Here are the troubleshooting steps I've taken so far:

1. Replaced and primed fuel filter
2. Primed IP via vacuum pump at the return line nipple
3. Attempted to prime via pressure with a squeeze bottle full of diesel (could only get fuel to come out at the return line, not at the injectors)
4. Confirmed a steady stream of diesel through the the filter (via MightyVac on the output side of the filter)
5. Confirmed GP light on dash at key-on (troubleshooting relay 109)
6. Pulled codes via VAG, only N109 mechanical failure. (I suspect this is from a couple weeks ago when I was getting the engine race at startup for a day or 2 - it stopped on its own and I didn't pursue it.)
7. Confirmed N109 fuel shutoff solenoid is clicking at key-on and getting power (~12.5V)
8. Removed spring and plunger from N109 solenoid
9. Cracked lines at injectors and confirmed fuel sprays from all of them
10. Checked operation of EGR valve (opens with about 18in.Hg vacuum)

I feel like there's more I'm forgetting, but I'm a bit tired right now I'll add to the list if I can think of anything else I've tried.

 

[kurtzl]

Thanks for clarifying - I'll see if I can get at least a couple quarts in there.

Depending on how much trouble I have getting one side of the tee off, I may run the return back to the jug instead of the tank - any reason to prefer the tank over the jug (perhaps gravity again)?

 

[ToddA1]

No difference, other than ease and less tubing.

-Todd

 

[kurtzl]

Ok thanks.

Unrelated: I clicked on the link in your signature...an hour ago lol. That's an incredible project!

 

[ToddA1]

Thanks, unfortunately that car is being neglected. I haven’t touched it in months, since picking up another Rabbit GTI mid-December. I guess it’s become another project... I’m in the middle of an engine refresh.

I’m calling it quits after these are finished.

-Todd

 

[kurtzl]

Understandable - looks like an enormous amount of work. Good luck finding the motivation to finish it off. It would be a shame to put that much work into it and then abandon it. Plus it's an awesome looking Rabbit

 

[kurtzl]

So I got back to it this evening after work but unfortunately have no good news to report.

I rigged up a fuel bottle and ran the return directly back to the tank:

I cranked for maybe 8 or 10 seconds the first time, with no discernible difference in behaviour. When I got out I could see a bunch of bubbles released back up through the line into the supply bottle, and a few back to the tank.

I had my housemate crank the engine for me 2 more times, each time about 6-8 seconds. The second time there were fewer bubbles released, the 3rd time I didn't notice any.

Tried cranking a 4th time for about 3 seconds with injector line #3 cracked and confirmed there's fuel squirting out.

Is it safe to conclude fuel isn't the problem?

I'm really confused now. Is it possible that relay 109 could be bad in spite of the fact that the N109 solenoid seems to operating properly? Mongler98, did I correctly understand that you have no CEL symptoms (glow plug light comes on normally at key-on) even though your relay is causing problems?

 

[kurtzl]

My VAG skills are pretty limited - I can pull codes and read a few obvious measuring blocks, but beyond that I don't know what I'm looking at. Any advice on what I can/should look at while cranking to hopefully diagnose the issue?

I think timing could produce symptoms like this, but I can't come up with any explanation of how timing might have changed overnight.

 

[kurtzl]

I think I may have stumbled upon the root of the problem.

I left everything hanging and hooked up after trying to start from the suspended bottle. Went inside to eat and write up the results of my testing. I went back outside about 45 minutes later and noticed the fuel level in the bottle was down significantly farther than I remembered, but figured I hadn't paid attention.

I poked around under the dash for a couple minutes, checking fuses (and grasping at straws). Reseated relay 109 and figured I'd try starting one more time before giving up for the night. Still no start, but when I got out to look at the fuel bottle, the level had barely changed. I thought maybe my bottle might be leaking at the cap, but it was dry. So I pulled out the drip tray I picked up this afternoon, and there was a LOT of diesel collected in the tray.

Grabbed my mirror and light and started investigating, and immediately saw the drip. Probably a drop every 3-4 seconds. Looks like it's leaking on the driver's side of the pump (opposite side to the timing belt). I don't know the correct terminology, but it looks like it may be leaking at the joint between the pump and the housing where all the injector lines connect.

I feel a bit silly for missing this earlier, but without the weight of the gravity feed the drip slows down to about one drop every minute. I'd noticed the diesel on the ground, but had chalked it up to my various testing procedures. It wasn't until I put the drip tray under there and saw the level change in the bottle that I realized how bad the leak is.

Huge thanks to Abacus and Todd for fuel bottle suggestion - that really produced the aha moment.

Guess I have to do some research to find out how hard that seal is to replace. Anyone done that one before?

 

[ToddA1]

Well, that explains the stream of bubbles in your supply line.

That distributor head has been discussed it many times. Many people change the O-ring while the IP is still in the car. I’ve done it that way once. If you don’t have enough preload on the plunger, the innards can fall apart. I prefer to have the pump off off the car and reseal the entire pump at once. Logic being if one seal is bad, others will soon follow.

Take your time, and work clean. I think there’s a diy in the sticky.

-Todd

 

[ToddA1]

https://cdn.shopify.com/s/files/1/2597/0906/files/TDI_Injection_Pump_Head_Seal_Replacement.pdf

I’m sure YouTube will have something for you.

-Todd

 

[kurtzl]

Thanks Todd! That procedure looks fairly straightforward, so I'm pretty tempted to change the o-ring without removing the pump. Pardon my ignorance, but what do you mean by "enough preload on the plunger"? Is there more to it than making sure the head doesn't come too far out of the pump body?

On the other hand...even though the TB only has 16K on it, it's going on 5 years old now. It looks ok, but would a prudent person replace a 5-year old belt regardless of mileage? If yes, then I might as well pull everything apart.

I think it was just an unfortunate coincidence that the seal completely failed right after I did the TB inspection. Not sure I would have found the problem any faster, but it definitely had me looking for answers in the wrong place (sort of).

 

[ToddA1]

There are spring loaded parts inside. If enough preload is not kept on the internals, the internals falling apart is a possibility. Reinstalling the IP with these parts out of place can break parts.

If you choose the insitu route, rotate the IP so the roller is at the top of the camplate ramp, and only loosen the head as far as you need to expose the seal. The head will be very wobbly and feel like it’ll fall apart, but it’ll be fine.

5 years or 50k is the rule. I need to do a tb swap for to age and only 20K.

-Todd

 

[KLXD]

Obviously the seal is a problem but I wonder if it is THE problem. Even with the head seal leaking it should have run off the bottle.

I'm afraid you'll get the seal changed and it won't start. Will it be because of the QA position, insufficient bleeding or the original problem?