Strange Fuel Pressure Issue

[shainiac]

Soooo, I can’t seem to shake this fuel issue. After a few minutes of run time, my main fuel pressure starts to drop off significantly.
The car will start and run cold with approx. 44psi base pressure, pressure rises and falls like it should with the FPR (goes down with vacuum, goes up if I pump shop air into the regulator).

See the diagram below-
I have a Walbro 450 (Pump A) in the main tank as a lift pump feeding the surge tank.
In the surge tank are two genuine AEM 380lph pumps.
One AEM (Pump B) is on when the engine is on.
The other AEM (Pump C) is turned on only under boost. Pump C is isolated by a check valve so it does not back flow when only Pump B is running.
Pumps B and C also tee together and feed the engine with 8AN PTFE line. All the line is PTFE braided and I haven’t ran any ethanol through the system. Pump A and Pump B are powered by Dorman 902-310 PWM fan relays whenever the engine is running/priming.  I’ve set duty cycle to 100% all the time. Initially it was set to 50% DC and upping to 100% seemed to extend how long it would run before crapping out.
Pump C is powered by a Bosch 30A relay.
Fuel pressure readings are taken at the Aeromotive FPR with a transducer.
I guess it’s worth mentioning that all the pumps prime and make pressure when bench testing, it only seems to be an issue after the engine has been running a while.

Between Pump A and the surge tank I have a very large spin-on 10u filter. Between the surge tank pumps and the engine I have another inline 10u filter.

What I’ve tried so far:
Replaced the in-tank Walbro because I thought the surge tank may be going dry.
Replaced the spin-on fuel filter
Swapped Pump B and Pump C positions (kept check valve correct to isolate inactive pump).
Replaced the relay wiring to all 10AWG w/ 30A fuses.

The problem still seems to be the same.






Here's the only pic I could find of the surge tank installed:

[digitalsolo]

FPR diaphragm?   That's a really solid fuel system and super similar to what I'm designing for my car, I can't see any reason it shouldn't work great.

[Supe]

Have you tried swapping the PWM relays for standard relays?

[Supe]

Also - tank vent has been checked as clear?  Are you seeing any engine symptoms that correlate to the pressure drop or checked with a mechanical gauge (i.e. confirmed the transducer isn't bad)?

[shainiac]

Swapping out the PWM is on my list. I actually bought a heavy duty Eaton fuse block/relay combo that Blake recommended. Unfortunately it won't be here till Friday and I'm leaving for vacation that morning. I can definitely rig up a couple Bosch relays in the meantime.

I pulled the vacuum line off of the FPR last night and it was dry. I've replaced the diaphragm maybe 3 years ago because it had began to leak into the spring housing and out the vac line. I guess it may be worth changing.

[shainiac]

Forgot to mention, I pulled the gas cap off when the pressure was about 10psi low at idle and it never went up, just kept going down until the car stalled.

Edit: The Dorman SSR's that are controlling Pumps A&B are the same units I'm using for my twin rad fans without issue. I'm still going to try bypassing them on my lunch break and see if that helps.

I'm also monitoring fuel temp via the flex fuel sensor and it's 80-90F cruising.

[shainiac]

Another update:

• Wired the pumps directly to 12V via a big 75A Tyco relay.
  
• Got a full tank of gas
• Changed my plumbing back so both pumps are in parallel with the check valve.
  
• I pulled all three fuses and could hear each pump turn on/off.
  Both AEM pumps made pressure at the regulator, but maybe a couple psi under what I had set it to.
  One of the AEMs definitely had a different tone that the other, not sure if it's because it's pushing back through the other pump with no check valve

With all 3 pumps running, fuel pressure still sucks. I can blip the throttle in neutral and see fuel pressure dip down into the high teens.
I have two theories I'd love input on -
1. The surge tank is getting aerated and the pumps can push enough bubbly fuel to keep up with engine demand.
     I have the surge tank mounted like the manufacturer shows in all of their literature. I would guess that as long as the inlets/outlets are above the pumps, it should self-bleed and stay purged of air.
     Manufacturer's drawing: https://pbie.s3.amazonaws.com/guides/IEFUU6_PLUMB_SCHEMATIC.PDF

2. I had a failed alternator regulator a week ago and my whole charging system saw 20V for a few minutes. Possibly fried the pumps? They're rated for 18V from AEM. Everything else electrical on the car was fine. The car was still on jackstands when the alt fried, so I had never put the pumps under any more of a load than just idling and free revving. And I never really paid any attention to fuel pressure other than initially setting up the regulator. [/list]

[cholmes]

Can you check pressures before each filter? Wonder if filters are partially plugged and pumps can put up enough pressure to overcome restriction when cold, but can't as they heat up.

[babillaracing]

when the pressure start dropping clamp the return lone and see if the pressure goes up. thats sound like fpr issue

[shainiac]

I can't really clamp the PTFE lined braided hose without kinking it
I'm leaving for vacation for a week in the morning. I ve already ordered 1 new pump to see if that fixes and it should be here before I get back. It wouldn't hurt to take the reg apart and make sure everything is okay. When I replaced the diaphragm last time, it was visibly defective, so hopefully it's obvious if that's the fault.

[jwvand02]

I can't really clamp the PTFE lined braided hose without kinking it
I'm leaving for vacation for a week in the morning. I ve already ordered 1 new pump to see if that fixes and it should be here before I get back. It wouldn't hurt to take the reg apart and make sure everything is okay. When I replaced the diaphragm last time, it was visibly defective, so hopefully it's obvious if that's the fault.

Can you throw a valve from home depot in temporarily?

[shainiac]

I installed new pumps in the surge tank and the issue seems the same if not worse. Only running one surge pump atm since that’s the configuration it’ll run in 99% of the time. It primes up to 43psi and then slowly drops the longer the surge pump and in tank pump run. After 5 minutes of  the surge and lift pump running and engine off, my line pressure was 18psi. You can hear the pitch of the surge pump change quickly as it continues to run. I’m kind of thinking that all of the flow from Walbro 450 lift and 380LPH surge pumps are aerating the small surge volume and the surge pump is sucking bubbles and not able to build pressure. Even though there is a check valve isolating the second surge pump, I capped off the lines to rule out that the fuel isn’t being pumped back though the off surge pump.

I’m going to take apart the FPR tonight and see if there’s anything fishy there.

I also emailed the surge tank manufacturer and haven’t received a reply after a week 

[cholmes]

I think you may be right about the surge pumps sucking bubbles. Do you have filter socks on the surge pump inlets? The typical Walbro filter socks act like a small Holley Hydramat, and will help prevent bubbles from being sucked into the pump. Sounds strange, I know, but in our testing (we build a surge tank as well) we found that to be the case.

[shainiac]

At lunch I pulled my FPR apart and it looked fine. There was a little bit of wear where the ball seats into the body of the regulator but didn’t look horrible. The diaphragm was in great shape. I was curious if maybe the lift pump never had a chance to fill the surge completely before I turned on the AEM pump and it instantly cavitated which is why the pressure seemed maybe worse than before when I tested it last night. Today I pulled the surge pump fuse and let the lift pump fill the surge for about a minute and hopefully get most of the air out. After doing that, the surge pump seemed to make the regulators set pressure okay. But as soon as I blip the throttle, pressure falls to >20 psi (at 0psi vacuum).

My surge tank has the pumps mounted horizontally. How my surge tank is constructed, the inlets of the pumps are about 1/3 up from the bottom. The feed from the lift, the return from the rails and the return to the tank are all about ¾ the way up from the bottom.

The fuel sock is an interesting idea. The AEM pumps I’m using have a “prefilter” built in on the inlet. It’s more of a rock catcher to keep junk out of the pump. Interestingly, the AEM pumps I removed yesterday didn’t have any debris in this filter. Thread-in filter socks are available, but I'm not sure if they'd physically fit since the pumps seal into the surge tank with a o-ring that's the exact OD of the pump body.

The only other piece of hardware I haven’t checked or replaced is the inline XPR 10u filter between the pumps and rails. I guess I know what I’m doing tonight lol. Really, there’s no way for junk from the tank to get through the Walbro pump strainer, the huge 10u canister filter, the AEM pump strainers and clog the XRP filter. I did blow out the new lines with brake clean and air before I assembled all of the fuel system. I guess if one of the old AEM pumps broke down, it could have pushed some junk into the XRP filter, it’s not very large.

[shainiac]

Well, it turns out it was the XRP filter. To my surprise the filter isn't pleated and is basically a 3"L, 3/4"OD fine mesh tube inside the filter body.

 The filter element was coated in what I'm guessing was Krytox and dirt. Krytox is a badass  PTFE-based o-ring lube that's basically impervious to everything and I used it to install the pumps into the surge tank. I guess it wasn't able to pass through the filter but also didn't dissolve into the gasoline.

Anyways, the car runs great now. Fuel pressure at idle was 5psi higher and doesn't drop if I blip the throttle. Also, the car RIPS! I took a quick spin and rolling into second gear annihilated the 285 rear tires on only 8psi of boost.