Think I'm sucking air from the gas tank.

[offalot]

Ok I have had a similar problem ( viewtopic.php?f=8&t=11796 ) So i am not sure if this is a continuation or totally different. Either way, my caddy stalls out every now and again and I found it was due to sucking air through the lines. It happened fairly seldomly, and I tried tracing it before, so my band-aid solution was to T off the line right after the filter and run a length of hose that would capture the air before it got to the pump and I could pull a vacuum on it at times to purge it. Clear as mud right?

Anyways, it was doing it so much I couldn't get a mile without it "airing out" again. I'd already replaced all the lines from the tank to the pump with nylon and bunyl. I then took off the filter housing and pressurized it and submerged it, no bubbles. I ran a short tube before the filter and stuck it in a little tank and ran the truck, no bubbles present, and finally I disconnected the nylon line from the rubber hose that connects it to the tank, put the nylon line into it's own tank and ran it, after the initial bit of bubbles present from the cangeover, I didn't see any. So is there any way the air is getting in through the tank? This baffles me. I don't think it is exclusive to a full or near empty tank.
Very sorry for the wordiness. Just want to be thorough.

[damac]

Has your injection pump been resealed?

How are you sure its getting pockets of air that is stalling it out?

If your tank vent is clear, you have new runs and a sealed working injection pump, I can't see how a tank that is 3/4 to full would introduce air specifically.

I have seen tanks with broken internals on gassers where when the level gets 1/2 or below the tank will act up all of a sudden at angles, long sweeping turns, etc. Had one in my rabbit and this was enough to shut the car off dead all of a sudden, but it was a fuel delivery problem.

Do you know the health of your pump? Is it pumping a good amount of fuel on chasis even when cranking?


I have had tons of problems before on multiple pumps/chasis thus far and have had the bubble thing before but never enough to kill my car dead in its tracks when running specifically dealing with air.

Maybe a mix of issues is leading to bad running. Like weak vanes, leaky mainshaft seal, etc.

[Fatmobile]

Air is usually caused by high vacuum so pressure won't find a leak, there isn't one.
A vacuum gauge on the fuel line will confirm.
More air when you rev it is common with high fuel line vacuum.

Can you give me a good pic of the top of the pump with special attention to the bolt that is supposed to say "OUT" on it.

[82vdub]

I was thinking the same thing as fatmobile. You may have a blockage and the high vacuum pulls the air out of suspension and then the fuel line turns to foam and the car dies. After a few seconds, the car will fire right back up. I observed this once years ago while trying to determine why the car would run for a while and then quit (had an inline electric pump that was apparently failing causing the issue).

[offalot]

That makes sense. I was wondering why I wouldn't see a leak while it sits, even for a long period of time.
I'm pretty sure the leak isn't through the IP because I can see the bubbles coming from the filter housing, and as you said fatmobile, more revs= more air.
I am pretty sure the obstruction isn't in the tank, as it had no problem soaking me with diesel last night when I pulled the line.
So from there, it is all new 3/8" nylon line up to the filter and I ran that submersion test I mentioned earlier with the hilter housing under a fair amount of PSI, so I can't imagine where it might be still
Here is the pic of the top of the IP as requested.

[Fatmobile]

Hard to tell if the OUT bolt actually say OUT on it? If not you will need an OUT bolt.
The OUT banjo bolt on the output of the pump isn't a normal banjo bolt.
If you take it out there will be no holes in it,.. unless you look very close, With my old eyes I have to look so close that I can actually smell the bolt, ha.

With no screen or restrictor in the output banjo bolt that pump will really move some fuel.
So much that even a clean filter will cause a high vacuum.

[damac]

On another note, unless I am blind it looks like your timing belt is shredding itself terribly on the outside.

You should get to the bottom of that. Make sure covers are sound and mounted correctly. I like to use the im shaft pulley with no lip as well and I think thats why they came that way on later cars?

If everything is as it should be, you probably need to adjust the injection pump bracket to get your belt to track straight.

My first vw I got though was an interesting case in that regard. I did not know it when I picked it up but it had half a timing belt when I got home! The lower timing belt cover was fiberglassed to the oil pan, I have no clue what they were thinking.

Car ran great but I soon found that I could not adjust this pump straight. It ended up having the loosest mainshaft injection pump bushings I have ever seen. I could wobble it back and forth with my hands. I ended up dumping it and haven't had the same problem on any car since.

Another interesting thing to me is I have changed pumps that I thought were doing the jobs and have the timing belt tracking change on the same engine. Wouldn't think its possible if it didn't happen, although most of the time I have found pump swapping went ok.

[offalot]

I do indeed have the OUT bolt with the smaller ports installed. Think I went through that whole deal last time when the IP was the culprit.
Damac, you're right, thanks for pointing that out. I didn't even notice. I'll have to get to the bottom of that. Probably is a misaligned IP.

[Fatmobile]

Soo we are back to it pulling a high vacuum due to a fuel line restriction.
Is there a clear line/air before the filter?

[TylerDurden]

The common diagnostic for fuel supply issues is "bottle feeding"... clear lines in and out of the IP to a clean container of fuel.

The filter can be added to that loop after the basic loop is verified leak free. (Pressure and vacuum leaks are different, BTW.)

If the problem persists after filter and IP are verified leak free, it is obviously lines to/from the tank or tank. Tanks are vented, so test system with tank cap on/off... if the vent is plugged, vacuum can build up and cause problems.

[offalot]

From the first post: " I ran a short tube before the filter and stuck it in a little tank and ran the truck, no bubbles present, and finally I disconnected the nylon line from the rubber hose that connects it to the tank, put the nylon line into it's own tank and ran it, after the initial bit of bubbles present from the changeover, I didn't see any"

The line going from the tank to the filter is semi transparent but when I rev the motor and bubbles go to the pump, I cant see bubbles coming up but hard to tell 100%.
I think later today, when I get the chance, I'll plug the line that connects to the tank and pull a vacuum at the line that connects to the pump and hopefully find something.
I am suspicious of the hard nylon line that goes into the back of the filter housing. I have it slid over a barbed banjo so it probably doesn't have the sealing power of a rubber type hose. I just looked past it since it wasn't actually leaking, but as you say, a pressure leak is different than a vac leak.

[Fatmobile]

Sounds like good progress.
Have you cleaned the intank screen and pickup area?
It sounds like it's restriction free all the way back to the tank.

[offalot]

I have not cleaned the screen. Do you get to that from the top? Seeings as it gushed out pretty good from the tank though, it seems the tank is pretty free.
I still haven't got to mess with it since the other day. A lot going on.
I wouldn't be surprised if it was one o those frustrating instances that pops up every now and again where some little flap of rubber or floating obstruction only acts up when you're not looking.

[offalot]

OK, here is where I'm at. I vacuum tested the fuel line from the pump to the tank, after I found a eak in the contraption I put on to catch the bubbles, the whole line was tight. So I put the vac gauge on the pump at the inlet and blocked the return. I was getting a vac loss, fairly quickly. Then I sucked diesel and ruined my vac pump diaphram . So can I put some pressure into the pump to try and find the leak without ruining anything?

[82vdub]

I wouldn't put much pressure into the IP. You may blow out seals.

You can get a Mity-Vac vacuum system for bleeding brakes to use for vacuum testing. That's what I use, however, when any fluid is involved, I always make sure the canister is in line so I don't suck fluids into the vacuum pump itself. Matter of fact, you can rig up any canister that can hold a vacuum in this line, if you need a larger canister then what the Mity-vac includes.