The way I am viewing it, it shouldn't really matter which way it's plumbed. I also didn't see any reference or caution note to a dead-head setup vs a thru-flow setup in the Edelbrock instruction; it's Edelbrock 1729, FYI.

If the FPR is before the fuel rails and is a dead-head design, then fuel is entering the FPR thru the input port, passing by the valve and exiting the output port until the pressure in the fuel rails AKA output port reaches the set spring pressure on the valve, then the valve opens and fuel bypasses out the return port.

If the FPR is after the rails, and the output port is blocked, then pressurized fuel is leaving the rails and entering the input port, hitting the blocked output port, and then if the pressure is equal to the adjusted spring pressure the valve opens and bypasses... it's a simple system that I just can't picture working for setup A but not for setup B.

 

As a side note, make sure your fuel pressure sender is not the problem. I chased my *** last week thinking something was wrong with my fuel system, when it fact it was my sender that was dying a slow death.

Andrew

 

BTW, the car is beautiful. The pressure drop in your chart is ugly.

Change one thing at a time to mail down the culprit.

That's a pretty drastic drop. I once had a case where the engine twisting under torque would short a wire.

 

I was going to check out your build, then I saw the thread was 117 pages long. Holy S batman! Nice car though.

 

I've used a bicycle tire pump to fake the pressure into the regulator to see if the fuel system can maintain pressure under boost with the engine not running. Won't check if you have enough volume but it's easy information.

 

I went through this same issue. I tore apart my fuel system, torn apart my regulator, had a fuel system specialist come by (a fellow engineer) and we found nothing wrong. For the hell of it, I ordered and installed a walbro 400LPH in tank pump just to see if that would help....not much really. So, all back together and started watching more logged data.

What it came down to was voltage loss. I logged my voltage and found that my system volts were dropping as RPM increased. So, got a new alternator and that helped a little. At this point, the light bulb started to get "brighter". Next, I looked into my wiring. The 12 gauge dedicated wires with 30 amp relays were NOT ENOUGH for the pumps!!! I changed all of the wiring to 8 gauge wire and added 50 amp relays. BOOM! Solid fuel pressure!

So, moral of the story, these pumps need wires, big ones. The pump graphs showing flow vs volts can be eye opening. I saw my system voltage drop down to 11 volts a few times. You gotta figure, when the R's get high, the injectors are firing fast, the ECM is chugging info, your pumps are drawing power to keep up, 8 individual coils are pulling power and you have electric fans that might be on? All of this amp draw is taxing your system.

Log your voltage and up the wiring. see how it goes. cant hurt.

Also, I have my battery in the trunk. So what I did was put the new relays under my trunk latch. I then ran dedicated group and power off the batter directly to the relays. Then, from the relays down to the pumps in the tank. What I did do to turn on the new BIGGER relays was just use the existing relays that were powering the pumps (they are under my dash) to just energize the new bigger relays. Overkill but simple and effective.

 

I did something similar, to make a quick & easy pressure test setup I changed the Holley fuel pump prime time when IGN is turned on to be like 120seconds, then I would turn the key to Start and would have the pumps running continuously without having to hotwire anything. I adjusted the FPR up to 90psi and had it running like that for a 5 minutes while I inspected every hose/fitting I had touched.. the pumps were audibly working harder but never once had a strange "dying whine" or any other concerning noise beyond the fact that their tone changed due to the increased load. Adding an outside influence to that like a control pressure source is a great idea.


This is a great point and very valuable input, I appreciate it. Every log that I looked at, my system was at a constant 13.9V.. and I've never noticed any occasion while driving where the dash voltmeter dipped or dropped weirdly.

I definitely want to inspect all the pump power connections, the battery connection for the pump solenoid power, and check all grounds to see if one is corroded or loose.

This is a breakdown of my current fuel pump wiring setup:

0) two in-tank TRE Performance 300lph fuel pumps
1) each of the two pumps has its own "stock" (~14ga) wiring directly off of the pump plug, about 3-4" long
2) pump wiring then crimped to 12ga wiring while still inside the tank
3) each pump has its own rubber compression-wedge-fit bulkhead fitting thru the fuel tank access panel
4) both pumps' wiring (2 x red power wires, 2 X black ground wires) connect to a Deutsch 4-pin DTP connectors (25A per pin, largest Deutsche connector I could find)
5) both pumps' power wiring are run to the output terminal on a cheapo 3-pole 80A solenoid
6) supply voltage to the solenoid is fused and hooks directly up to the in-trunk battery side terminal using if I recall correctly 10ga wiring

I would really like to make an extension harness on my 4-pin Deutsche connector to bring the solenoid into the trunk.. this will be easier to troubleshoot in the future, but will also make it easier when I change to having one pump a full time primary pump and making the secondary pump boost-actuated, when I revisit my fuel system in the future.

 

Sounds like a lot of your wiring is already substantial. Might just want to double and triple check some areas.

Also, grounds are key so give those a once-over.

When I was trouble shooting, what I did was just hard-wire the pumps on ALL THE TIME directly to the battery. It was my down an dirty way to diagnose that night. It also eliminated any variables. If you feel like to want to take the time to hardwire the pumps on, it will just eliminate ALL the variables. Just a temporary way to trouble shoot until the job is done right.

 

That's a very good idea, and I had thought of that.. I just haven't had time to jack the car up and crawl under it to inspect or hotwire. I'm still also pretty skeptical about the fuel filter I have in it.

 

I also have a rear battery and added an extra relay/fuse box in the back for the fuel pumps and oil and trans cooler fans. For the primary fuel pump, I run 2 standard sized, but 60 amp rated relays in parallel. I need to rig some higher amp fuse however. It's not good practice to run fuses in parallel.

I think you're on track to nail this problem soon.

 

I'm not speaking from experience on these solenoids, but are they designed for constant use or just short duration operation?

 

take that mega solenoid and just jump it from big terminal to big terminal. or, just tied the 2 eyelets on one post. eliminate that solenoid as a variable.

Also, I am surprised the ECM sees 13.9v all the time....should drop at some point.

 

I haven't looked into the internal design of that regulator. If it is a simple direct-acting poppet-style regulator then I agree it wouldn't matter which way you plumbed it, but if the regulator is pilot operated on the outlet side then you could have a problem. I suggest just asking to be sure. The diagrams in their instructions all indicate the outlet would be connected in a dead-head arrangement and not capped; I wouldn't assume it can run both ways.

 

That's a good point.. just hotwire across the solenoid. I have the alternator controlled by a single lead wire from the Holley ECU.. whether the Holley is programmed to cycle it on or off depending on system draw and voltage like an OEM setup, I highly doubt.. I presumed it just tells the alternator to turn on; I never thought much into load-varying charging since whenever the car is on, it's guaranteed to have two fuel pumps and two fans running so I definitely want the alternator charging fully even at idle.

That's an interesting point, and totally possible. I took it apart previous to getting the car back on the road, because I was curious.. but I didn't pay enough attention to remember if it's a poppet-style valve or pilot actuated. I may look into that first, and give Edelbrock a call as well.

 

Make sure you setup some a/f safeties in your Holley before you hurt something. How are the injectors? Have they been flowed or cleaned lately? Especially if on E85. Had a similar issue that took me awhile to figure out, good thing my safeties in the Holley shut it down.

 

I had set some fuel safeties in the beginning but I don't recall their settings.. definitely going to have to review those.

The injectors are new Fuel Injection Connection 95 lb/hr injectors, running pump gas. The injector duty cycle and calculated flow rate starts spiking after the fuel pressure starts decreasing, so I don't think I can attribute the pressure drop to one or multiple injectors getting stuck open or being commanded to spike in duty cycle.

 

Yes, you're correct there would have to be a leak somewhere for pressure to drop. I was trying to remember what mine was doing, fp was staying constant but dc was going up: https://ls1tech.com/forums/forced-in...-out-fuel.html

If the regulator is too small pressure would be higher. I always run the reg after the rails w/ a -6 return. 95 lb/hr seems small for this setup?

Nice build hope you get it figured out soon. For the safties, I would set one up if fp gets below xx value, and afr below xxx, oil pressure would be a good one to have too.

 

Thanks for that extra info 408GT... I'll dig deeper into the logs this weekend to see if I can find any bigger correlation between duty cycle, fuel correction, AFR, fuel pressure, and the resulting timing and order of these things changing.

From my reading and calculations it seems 1000cc injectors should suffice, but that's with assumed VE's in my calculations sheet.. I'll just have to wait and see how the duty cycles look when I fix this fueling issue and start increasing the boost.. I got the injectors for a great deal so that's why I went with them specifically.. we shall see if I end up needing bigger.

 

So I'm looking at the graph again. What is the "Fuel Flow" calculated from? It would seem logical duty cycle must be in there?
Also, why is the AFR staying correct when your fuel pressure dropped so much? Maybe it's wiggly there because of fuel correction, but I would think it should be much leaner with that much loss of fuel pressure.

Waiting to see what additional logs say.

 

Here are screencaps of that same above log & time window that LSswap quoted, but I've included injector pulse width, injector duty cycle, and fuel learn % for reference. I'm not expecting these to help get me any obvious solutions, but just to help you guys get the same visibility of what I'm seeing. It really looks like the fuel pressure is dropping off first and then everything is compensating [with a bit of lag] from there... to me it seems to point to a mechanical fuel flow issue and not a tuning one.

This is timestamp 190 seconds, the instant fuel pressure stops remaining constant due to constant MAP aka throttle position...



This is timestamp 194 seconds, when fuel flow ends its slow decline and afterwards starts to drop off FAST...



This is timestamp 197 seconds, when fuel pressure is at its lowest, compensation/learn is highest, and AFR is highest before I back off the throttle...