Honda BF50D H/P fuel pump not staying entergized

I have a 2012 BF50D S#BBEJ 1012938 that will entergize the H/P fuel pump with the key on & I can achieve 40 psi. When I start the engine, it will run until the fuel pressure drops to about 20psi. I replaced the H/P fuel pump due to stuck in the VST & cleaned. I bought a factory manual to find what components control it. With a test light & DC volt meter. the pos. side is good to the pump. The neg. side gets a 2 sec. completion to the pump. It drops when the engine is started. Cleaned all connections with electronic cleaner & applied Dielectric grease. Hooked up the motor to a separate battery & cables to insure a strong 12v. The manual states that the ECU controls the neg. (blk/white) wire with inputs from the (2) crank & (1) cam sensors. There are no alarms active via a Sierra Stats Dia. tool. Also checked by jumping the data link conn. Replaced the ECU with a new unit. same results. replaced the Crank & cam sensors. Same results. I intersected the F/P (blk/white) with a test conn. (male/female).(The VST is buried behind the intake) I can put a jumper to the motor block to the blk/white (separated from from the ECU) & the engine will run & maintain 40 psi F/P. A volt meter shows varying dc volts.(from the ECU) Some times 13.5 down to 4v. Strange that while the motor is off or running, the Sierra Stat showed the fuel pump was turned off. I pulled the (2) crank,(1) cam,& heat sensors 1 at a time. they would set a alarm & I could reset the individual alarms via the Sierra Stat.
Any suggestions as I am stumped!!
Iceman

Here's a way to test the integrity of the negative return to the ecu. Using jump wires,to run the fuel pump so that you can start and run the engine. Using a 12V test lamp with a 5W bulb (not LED ), connect the one end to the black/white in the fuel pump connector and the other to the positive battery cable and check to see if the light remains on and bright. I am guessing it will not glow or if it does, very dim. Then using the test light again, do the same test on the neg wires to the ecu plugs with the plugs disconnected and check you have a good solid glowing light. This should verify that the neg wires are good. People make the mistake of using a multi meter to carry out these tests which does not load the circuit and can show 12 volts when in fact there is very high resistance. Try this and let me know how you go, I believe this will reveal a bad ground, either an individual wire or poor earth to engine block connection.

iang6766
Tks 4 the reply
I will do as you suggested in the am. also I was able to free up the original fuel pump. I'll place that one on the neg. output of the ECU & the one in the VST I'll wire to ground & start the engine. This will allow me to check the voltage(voltmeter) & amp draw (DC ampmeter) to compare both pumps in a running situation.
Question- Is there any other sensor or component that controls the H/P fuel pump other than the (2) crank, (1)cam sensors & ECU?
As stated in my 1st post, I went to a seperate

battery & cables. The new pump runs perfect when the blk/white wire is grounded to the block.
Your input is leading me to the harness &/or plug @the ECU.
Iceman

sorry 4 the split posts as i'm a newbe as this

I would not be using anything other than a test light as I described, you will risk damaging the ecu if the old pump draws too much current. There are no other sensors involved, you have been very thourough in working out how it all works and I commend you for this. However, I belive in keeping things simple. you need to thoroughly check your ground circuits as that is the only thing you haven't done

iang6766
Tks 4 the additional info.
As soon as I can tomorrow, I'll follow your suggestions & update you.
Much obliged.
Iceman

iang6766
Success!!
I installed the 5w bulb as you suggested. Nice & bright. A volt meter showed 13.5 volts with the motor running (vst pump straight wired). Then i got hard headed - I reinstalled the old ECU & wired the old (factory) pump it. then I place it in a bucket of fuel to allow it to load up. It ran fine. The amp draw was 1 amp. when I placed my finger over the outlet, it went up to 1.2 amps. The new pump in the vst was showing 5.0 amps (yikes). So I reinstalled the old pump back in the vst. ( I failed to mention that the new pump was a after market unit guaranteed to match factory specs.) The motor runs fine with the new ECU & new pump laying on the bench. I then ordered a new factory pump today

FYI
The old pump pulls 1.7 amps installed in the VST. Apparently The ECU does like that high amp draw & lock out the fuel pump.
Iceman

Interesting, I have not come across that before, one has to wonder then why they use a 10amp fuse.