I've been trying to chase down a few nagging issues with my efi setup. I could never get the idle silky smooth and it seems it only gets worse with heat soak and/or hot weather.

When a friend heard my fuel pump running he pointed out how it sounds like there's air in the line. I agreed... seems air is being sucked in somewhere. My pump (walbro gsl392, inline style) is strapped to the subframe just before the arch over the rear end starts. It sucks fuel from the tank through the stock sending unit. I put another barb on the sending unit for the return.

To eliminate anything in the tank, we got a 2 gallon gas can. To fill it, I disconnected the regulator from the fuel rail and took the rail directly to the can. As the can filled up we still heard the gurgling, so we know the issue is from the tank to the rail, not the return side. We hooked the regulator/return circuit back up. Took a hose from the can directly to the pump. Ran the pump, gurgling still there. There are no evident leaks.

I'm scratching my head here... where's the air coming from? I used clear vinyl tubing so I could see the bubbles when they were being pumped out of the rail. Using the same clear tube between the can and pump, no bubbles. Could the pump housing itself be leaking? The barbs have crush washers on them.

I would try your clear tubing on a temporary basis in all the sections of the line if you can get some clear tubing that will hold the pressure.

Sam

_________________

That's not a bad idea - would require swapping some AN fittings for hose barbs but doable.

I got a new pump... I let it draw from a catch can, through the cars plumbing, fuel filter, rail, then back into another can. Bubble free. Hooked rail to regulator, then back to can - bubbles. I have no idea how a regulator can cause this. The regulator wasn't leaking fuel anywhere externally or out the vacuum fitting. And its completely under pressure, not on the "suction" side of the pump.... so it shouldn't be drawing air into the system. I slathered grease around the fittings and hose in an effort to seal anything tiny, still bubbly.

I took the regulator off and apart - nothing obvious. No holes in the diaphragm. I gave up at this point and ordered a new regulator. I got tired of the one I had anyways... the ports were just in a bad spot and small.

I wonder if you're too focused on these "bubbles". If you don't have bubbles in the pressure side of the rail, then what happens downstream of the regulator (return line) doesn't matter. If I had to venture an explanation, I'd say that you're seeing bubbles formed when high pressure liquid "sprays" into the low pressure side of the regulator. Pressure drop = vapor. I usually hear some kind of "gurgling" noise from rail mounted pressure regulators in operation.

_________________
63 Valiant Wagon
225 - 4 bbl

With the new pump drawing from a can, through the rail, and back into a can there are no bubbles. With the old pump drawing from the in car tank, through the rail, to a can, there were bubbles. It's not just the regulator that was causing issues. While I'm into it I think it would be a good idea to remove all sources.

I'm not familiar with this pressure drop = vapor principle, I'll have to look it up. There must be some qualifiers to it though... I've never heard of any sort of gurgling from a modern cars regulator.

I can see that pressure drop equals vapor. I did not know it always happens, but if your cooling system suddenly gets a leak, it boils, when it was not boiling before the leak. The pressure keeps those excited molecules from breaking free into vapor.

Sam

_________________

That radiator analogy makes sense. In that case the temperature of the water made the vapor pressure significantly higher then atmospheric. I wonder what the boiling temp of gasoline is... must be low since it evaporates on pavement on any given day. Wait though... if that's what's at play here, what happens at the pump itself? The difference in pressure before and after the pump is big as well, no?

The pressure goes after the pump. It drops at the regulator. The bubbles make no sense to me. It had to be drawn in on the suction side of the pump.????

Sam

_________________

A mainly water related situation that causes bubbles, is cavitation. You see it in a glass walled water basin behind propeller blades, when the speed of the blades causes the low pressure areas that make water evaporate and make bubbles.

Depending upon the pump design, it may not be totally unlikely, but it must be very unwanted, in a fuel pump!

Olaf

_________________
Aspenized

So in this case the bubbles are fuel vapor and not air. It seems then this would be a result of design of the pump. As I understand it, the problem is now gone, and we are simply exploring concepts at this point. Do I have that right Pierre? Is the new pump a different maker, or design?

Sam

_________________

The old pump was a walbro gsl392 - new one is a gsl393. Same everything except lower HP support (less flow volume).

I think I was / am dealing with multiple problems. I suspect the old pump was worn/damaged. The new pump is much quieter.

With the new pump going through the rail only it had no bubbles. Going through the regulator, bubbles and noise. I replaced the regulator - now I get only bubbles, no more noise. This lends practical example credence to what WagonsRcool said.

While the new pump and/or regulator may not have been the "ultimate fix" I think it helped me find the issue. After getting the new pump and regulator buttoned up I noticed something. For the first few seconds everything seemed normal. I set the pressure, gauge relatively stable, maybe +/- 0.5 psi (total gauge flutter 1psi). After that the tone noticeably changes and gauge flutter grows, maybe to 4-5 psi (+/- 2.5 psi) very rapidly.

I think my return is too close to the suction line. The bubbles induced in the regulator are getting sucked back up. Since it takes the fuel a few seconds to go "Full circle" I'm assuming that's what causes the delay. I can stop and start the pump, same thing happens. The only way I can get it to stop fluttering for the few seconds again is to wait long enough for the pressure to dissipate after turning off the pump.

Sam, you still liking your Rick's tank? I'm tempted... I think they sell the sump/pump unit separately, that would make for a good welding project. Otherwise I need to figure out how to get another return line in the tank. Current setup is a short piece of steel fuel line welded to the base plate of the pickup. Don't think it's long enough to bend. Maybe I could attach a short 90deg bend of tube using some of that submersible rubber hose.

Well, there goes that theory. I took the sender out... I didn't think the return was close to the suction. But since it was out, I made a piece of tubing that directed the flow away and attached it to the existing return stub with some submersible rubber hose (Great googly moogly batman $30 a foot!). As predicted, same issue.

So... came down to something with the return line. I took the return line off the tank and let it return into a can. With everything else hooked up proper, I still got the fluttering pressure. This eliminates the tank, boils down to something in the return line itself.

I took the line apart - found a bit of rust and crud... spent quite some time cleaning it out, ran air through it, brake cleaner, etc... put it back together, same thing.

So either the line is too small, or the bends in it too restrictive. There are quite a few 90's in it. Maybe there still is some rust in it I missed? Now I'm planning for a new return line.

I have been quite happy with the in-tank pump. It was the wrong fix for the problem I was having with fluctuating fuel pressure, but I have not been sorry I spent the money when I did. Your bubbles still seem a mystery to me. You still have our attention.

Sam

_________________

Sam - I was putting some thought into that in tank setup when the problem seemed to be inside the tank but since that was eliminated I put that idea on the shelf. Maybe in the future, but not now.

Did round 2 with the tube cleaning. I cut out a short piece that looked like the factory bent it too sharp and collapsed a bit, replacing it with rubber. A bit better but issue still there.

I've been using that clear vinyl tubing to do my experimenting. Its cheap as dirt... $0.35/ft for 5/16" ID. I got a 15ft piece, ran it from the regulator under the hood, over the fender, along the car, through the wheel well to the tank through the sending unit. It was quite comical. But it proved once and for all the car line is just too restrictive. Ran it like this for several minutes and with starts and stops to (hopefully) eliminate the "Just got lucky" first time around possibility.

Hard tubing is measured in OD. The vinyl tubing is 5/16 ID. To get 5/16" ID in tube I'd have to step up to 3/8 OD. What I still don't grasp is how so many OEM's can get away with 5/16 OD return lines... thought I even saw 1/4" once. Maybe weaker pumps I guess.

Alright... so I ran a new 3/8 hardline. The only part thats 5/16 still is stub that goes into the sender. That was the same with the vinyl tube though. The problem's still there. Grrrr... so a 5/16" ID vinyl tube = ok, 3/8" od hardline = not.

At this point I don't think its a restriction, but perhaps the bends in the hardline causing some sort of resonance. At least its very small now - less then 1 psi flutter on the gauge. You can feel pulses in the rubber parts of the return line if you squeeze it, but not the pressure line.

Some OEM's use pulse dampers. I'm thinking of building my own - tee in a vertical hose and attach maybe a fuel filter or longer length of rubber line and cap off the top - the trapped air acting like cushion. Think like a water hammer arrester you'd put in house plumbing. The OEM ones use a pressurized diaphragm but I suspect those are tuned to dampen higher frequency pulses from injectors not just resonances from the lines.