Breaking this out from the other fuel pump thread. Here is my setup on the bench.



Here is the circuit diagram for the relay



And my chicken scratch circuit



Lets keep this thread open for help setting it up, and sharing tuning ideas. I hope to have my setup in the car in the next few weeks.

 

FWIW I believe this should only be used with Turbine pumps. DW200/300. Walbro450.

 

Are you going to run this with a regulator, or totally closed-loop without a regulator? If closed loop, it seems like anything below would cause a target pressure below regulator pressure would just mean nothing goes through the return. If set above, then the duty cycle will climb to 100% because it never reaches target. So open loop seems the way to go with a regulator, that way you can tune in to always provide enough duty cycle to keep the pressure just above regulator pressure.

I was considering this route, but my current plans are to put my flex fuel sensor on the return, in which it would rarely get fuel. However now I'm considering running the flex fuel sensor on the feed line, and PWM'ing it open loop. I'll be running this on an E85 LS1 swap with 630cc EV14's.

 

Definitely not closed loop. Tuning would have to be way too perfect.

Plan is to run lower DC at vacuum, ramp up to higher DC in boost. I have a big AFPR so im not worried about overpowering the FPR.

The return line should always get fuel, you would have to perfectly tune DC to keep any from flowing.

 

My expectation is that there will be X%DC which is the minimum duty cycle required to maintain adequate fuel pressure, and there will be Y%DC which is the duty cycle that starts to overpower the factory regulator. My expectation is that there will be a pretty wide range of "correct" duty cycles which will give adequate pressure without blowing the FPR open. Not an issue for you, but I see this as a great option for people who need a big pump but don't necessarily need the adjustability of an aftermarket FPR.

 

I'm really hoping those values are the same for each pump. I can't test it. But you should get a fuel pressure sensor and log where it overwhelms, and where it doesnt.

I think this value will also change with load/RPM.

 

It should change in lockstep with actual injector PW. I expect the "Y" number to hit 100% pretty early on, too - way before max boost, probably even before you get into boost at all.

 

It will change with PW (well, PW + RPM, so duty cycle) and injector size, it will also change with a ton of other things. Model of the pump, voltage at the pump, health of the pump, temperature, pressure in the tank, manifold pressure, resistance to flow in the rest of the fuel system, etc. Hell, it probably varies depending on the temperature and viscosity of the fuel.

Looking at the data I took before I installed a regulator, with the stock NB regulator and a Walbro on my 99 it took 4600 RPM and 220 kpa MAP (at a DC of 40% on my id1000s) to bring the fuel pressure down to the nominal regulator set point.



--Ian

 

You just opened my eyes to more than 4 traces per graph.

 

It works

 

Alright just tried out PWM control with the Walbro450. Fuel pressure is not stable at all if I am below the FPR setting. jumps back and forth 2-3psi. I'll get data this week.

This wouldn't work without an FPR. But with an FPR it works great. Right now I have it set to be 50% at idle and low rpm vacuum. 75% in high rpm vaccum. And 100% in boost for now.

 

Sorry for my ignorance, is megasquirt just pulsing the relay on and off? Is this relay special or just whatever 30/40A 4 pin

The diagram I'm used to seeing is

 

Yes.


It's a solid-state relay (essentially a FET), and it's being PWMed as a rough approximation of a variable-voltage drive.

 

@_@


okay, thanks

 

 

Are you using any kind of flyback diode with this?

 

I'm not. But now I'm going to add one.

A 1N4001 should be fine I think.

 

Cool. I've got a 1N5819 planned for my fuel pump. Do you happen to know of a diode that's rated for over 30A peak? I'm trying to find one for a cooling fan that maxes out in the 35A range.

 

Did you ever see pats post where he called spal and they said not to use pwm below 1khz?

 

My CAN-EGT can PWM up to 20kHz