FI- strange fuel pressure behavior

[Belgian1979]

I was testing with a ford fuel pump control unit (pump speed controller) to regulate the speed of an Aeromotive A1000 pump.

The system originally uses a fuel pump pressure regulator that is attached to manifold vacuum. I left it in place as it has a certain amount of leakdown which might be beneficial to deaerate the fuel lines and to prevent vapor lock. The strange part is that when turning down the pump speed (can actually hear the pump spin less fast) I'm only able to reduce the fuel pressure by about 4 psi in total which seems strange. I went down to about 20% of the normal speed, but this is about the max.

My system consists out of a fuel accumulator with a high and low pressure side (high pressure to the EFI circuit, low pressure from the tank to the accumulator and back to the tank to fill the tank.

When running the pumps and turning them off, I see the pressure bleed off fairly quick, so this doesn't seem to be a line restriction or something.

Any thoughts ?

Edit : to add : when running the pump at full speed and basically as a non regulated system (except for the FPR) everything works fine with fuel pressure moving up and down with the intake vacuum.

[peejay]

Are you checking at idle or under load?

It takes very little volume flow to make significant pressure at idle. If your pump is a positive-displacement pump rather than a turbine style pump, it should easily make enough volume at only 20%.

[Belgian1979]

yes this is at idle.

I'm not sure if it's positive displacement or not. Cannot seem to find any info on it, except that pressures could run up to 100 psi if used with a non return style system.

[peejay]

I'm unfamiliar with the A1000 and also could not find any exploded views. But the aftermarket-type Aeromotives that I have taken apart were positive displacement.

Trying to PWM a positive displacement pump as a method of regulating fuel pressure would be difficult, at best. Your PWM would need to follow your injector duty cycle, and it wouldn't necessarily be linear.

OEM pumps that are PWM-controlled for pressure are all turbine style, that I'm aware of.

[Belgian1979]

Actually the intention was a PID closed loop setup.This would automatically adjust.

I'm also thinking this is a positive displacement pump. My thinking is that the FPR would actually be a sort of throttling valve in that case. This could explain the large and fast bleed down rate that the FPR exhibited so far.

I would think that in this case I would need to set the minimum speed I want to run the pump with lowest fuel demand, which is during idling. Then adjust the FPR so that it reaches my setpoint and then have the PID control do the rest. But I'm guessing here.
It's not so much fuel pressure I'm trying to control with pwm but I want to reduce pump speed as a way to reduce it's noise during low speed/cruising.

[mk e]

2 thoughts

1) the A1000 is a HUGE pump so you'll need to slow it down a lot at idle where you need very little fuel. I have the same pump hooked up to let the ECU PID the pump to control pressure and had just started playing with it a couple months ago....but I had it at 50% PWM without seeing any pressure change at idle.

2) If you are going to try to PID the pump you need to have the regulator set ...at least 5psi above whatever pressure you want the control to work at. The reason is if it hits the external regulator then the feedback to the control is wrong abd it it do all kinds of wrong things as a result.

[Dan Timberlake]

Are any of the pumps, accumulator, and piping hard mounted to the chassis or body panels?

[user-23911]

You really want to set it up using a bench test.
The regulator is probably too small to handle full pump flow at idle.

[Belgian1979]

Are any of the pumps, accumulator, and piping hard mounted to the chassis or body panels?

Accumulator and pump are hard mounted. Fuel lines are braided rubber.

[Belgian1979]

You really want to set it up using a bench test.
The regulator is probably too small to handle full pump flow at idle.

The regulator is the Aeromotive one destined to be used for this pump.

Bench testing is a bit of a problem as MS only allows the have the pump running in test mode at full speed or when the engine runs (when it sees rpm) at the programmed pwm

[Belgian1979]

2 thoughts

1) the A1000 is a HUGE pump so you'll need to slow it down a lot at idle where you need very little fuel. I have the same pump hooked up to let the ECU PID the pump to control pressure and had just started playing with it a couple months ago....but I had it at 50% PWM without seeing any pressure change at idle.

2) If you are going to try to PID the pump you need to have the regulator set ...at least 5psi above whatever pressure you want the control to work at. The reason is if it hits the external regulator then the feedback to the control is wrong abd it it do all kinds of wrong things as a result.

Could you explain 2 a bit please

[mk e]

2 thoughts

1) the A1000 is a HUGE pump so you'll need to slow it down a lot at idle where you need very little fuel. I have the same pump hooked up to let the ECU PID the pump to control pressure and had just started playing with it a couple months ago....but I had it at 50% PWM without seeing any pressure change at idle.

2) If you are going to try to PID the pump you need to have the regulator set ...at least 5psi above whatever pressure you want the control to work at. The reason is if it hits the external regulator then the feedback to the control is wrong abd it it do all kinds of wrong things as a result.

Could you explain 2 a bit please

A PID controller is looking at the difference between desired pressure and actual pressure and even in a perfectly tuned setup there will be over shoot. Attached is a graph of my throttle control....you can see momentary mis-match even on small moves and significant overshot on step changes. This is a very normal behavior you see on any PID setup because the controller needs to see those error to react properly,

For the throttle that means taking care to be sure the throttle did not/could not hit a stop and for you it means not hitting the regulator hit acts the same as a stop. Maybe the ford controller has some code in it to let it recognize a limit and deal with it?

[Belgian1979]

This is what I found on the Aeromotive website :
"Many Aeromotive pumps are of the “non pressure limited” type, including the A1000 for example. This type of pump cannot be used with a static (non bypass) regulator, because to stop the flow coming from the pump completely would drive fuel pressure to 100-PSI or higher,"

It seems to imply that the pump is indeed a postive displacement pump.
This type of pump puts out a volume that is directly related to the rpm of the pump. So normally at least reducing the speed would have the flow reduced as well.

However I think that the Aeromotive FPR kinda acts like a throttling valve. Actually never closing, but just bringing the ball closer to the seat so that flow is inhibited and pressure goes up. This is probably what I'm seeing : When reducing pump speed, the flow is reduced and the ball again moves closer to the seat (but not touching) untill a certain balance is reached.

If I use a higher pressure set point for the FPR, I only adjust the setpoint upwards, but probably would not be able to get more than the previous 5 psi adjustability.

[mk e]

This is what I found on the Aeromotive website :
"Many Aeromotive pumps are of the “non pressure limited” type, including the A1000 for example. This type of pump cannot be used with a static (non bypass) regulator, because to stop the flow coming from the pump completely would drive fuel pressure to 100-PSI or higher,"

It seems to imply that the pump is indeed a postive displacement pump.

No, it means it's not safe to use the pump without an external blow-off valve of some type and that normally takes the form of a by-pass type regulator. You need to keep the regulator in the system, but if you want to PID you should set the regulator above the control pressure so it's only acting as a safety device.

[Belgian1979]

Hm, I was thinking of the other way around, choose the minimum speed I want to pump to turn at idle, reset the FPR so that it reaches the desired pressure and then let PID get it right from that point on. The controller should actually have to increase the pump speed in order to maintain the desired pressure if I'm not mistaken.