When you switch a return fuel system, as I have, you know that heat is the enemy. Get your return fuel too hot and your pump can cavitate, losing pressure and damaging components. This gets worse and worse the more power you make and the bigger pump you use. You're still only using a few hp at idle, but your pump is always pushing whatever peak hp you built it for, meaning a whole lot is going back to the tank with heat that it picked up from the engine bay.
There's a few ways to mitigate this. The first is a deadhead setup, which returns fuel before the rails. This means the fuel doesn't get a chance to travel through those aluminum heat sink rails before returning to the tank. It's a start. The next is pump staging (with multiple pumps or those that support the feature.) You run it on a "base" fuel level most of the time, but as soon as it goes into boost, you turn on the other pumps, or turn the pump on full. It's a viable solution. The last, most efficient, and most expensive, is a fuel pump controller. Lots of pump manufacturers sell controllers to the tune of $300+ that modulates the pump based on inputs like manifold pressure or throttle position. Nice, but expensive.
I wanted a controller of my own so I could turn my 1300hp capable pump down to a minimum while at idle or driving around town (and keep the noise to a minimum.) The pump I have is a Fuelab Prodigy, which has a unique feature- it's a brushless motor with it's own controller that has an extra input for controlling duty cycle. If you're so inclined, they have a regulator that does this. All you have to do is run a wire from the regulator back to the pump. Of course you pay a $200+ premium for this feature. Now I could just use a Hobbs switch to go from 50% duty cycle to 100% when in boost. That would be $30. But I have a controller project that might turn out cheaper...
Lucky for us, the 99-04 Mustangs already have a controller feature built into the car as a legacy from the returnless system. After all, that's how it functioned, controlling pump speed to achieve desired fuel pressure. But it isn't quite that easy. The ECM outputs a PWM command from 10-50% at 100Hz (half the desired duty cycle), and the FPDM basically opens and closes the circuit at twice the ECM's command at 20+kHz! Quiet and efficient, but too fast for our needs. The pump wants its command from 10-100% at 500-1500Hz. And the ECM needs an "OK" status at 50% and 1Hz from the FPDM. We need a way to achieve that.
Ok, some people are probably asking "what the hell is PWM?" PWM stands for "Pulse Width Modulation." It's a digital way of achieving an analog output. It gives an output by pulsing the signal to the "high" status for that fraction of the time period, then "low" for the rest. For 20% it will go high for 20% of the time period, then low, then repeat it. If your frequency is 1Hz (1 cycle per second) it would go high for 0.2s, then low for 0.8s.
Thankfully microcontrollers have gotten stupid cheap lately. I picked up an Arduino Nano for about $10. Using that microcontroller it's going to poll the ECM command, and then send the desired duty cycle to the pump at the correct frequency.
Now there's a couple wrinkles that make it a little more challenging.
The Arduino output is regulated to 5V. The pump and most car hardware takes battery voltage. So we need a way to step it up. The answer is using a couple transistors, a diode and pull up resistor. That way we can control a 12V signal using the 5V output. The most common way to do this produces an inverted signal (i.e. your 20% becomes 80%), but we can fix that in the code because hey, code is free.
The Arduino takes a 7-21V input, but you'll fry the integrated voltage regulator in short order if you spend time above 12V. We all know your car is going to be in the 13-15V range while running so you need a way to fix that. A 9V regulator will help with that, and for only a little more than $1.
There's the basis for the project. I'll update with circuit diagrams, code, components, and pictures as it progresses.
Right now the only thing I'm wondering about is how to make a clean finished project. I've got a breadboard to test it with, but I'm not sure I'm ambitious enough to make a dedicated PCB for it.
And I'm sure I'll end up needing some help from you electrical nerds on here.
For now all told with the components and chip I'm still slightly cheaper than a Hobbs switch.
There's a few ways to mitigate this. The first is a deadhead setup, which returns fuel before the rails. This means the fuel doesn't get a chance to travel through those aluminum heat sink rails before returning to the tank. It's a start. The next is pump staging (with multiple pumps or those that support the feature.) You run it on a "base" fuel level most of the time, but as soon as it goes into boost, you turn on the other pumps, or turn the pump on full. It's a viable solution. The last, most efficient, and most expensive, is a fuel pump controller. Lots of pump manufacturers sell controllers to the tune of $300+ that modulates the pump based on inputs like manifold pressure or throttle position. Nice, but expensive.
I wanted a controller of my own so I could turn my 1300hp capable pump down to a minimum while at idle or driving around town (and keep the noise to a minimum.) The pump I have is a Fuelab Prodigy, which has a unique feature- it's a brushless motor with it's own controller that has an extra input for controlling duty cycle. If you're so inclined, they have a regulator that does this. All you have to do is run a wire from the regulator back to the pump. Of course you pay a $200+ premium for this feature. Now I could just use a Hobbs switch to go from 50% duty cycle to 100% when in boost. That would be $30. But I have a controller project that might turn out cheaper...
Lucky for us, the 99-04 Mustangs already have a controller feature built into the car as a legacy from the returnless system. After all, that's how it functioned, controlling pump speed to achieve desired fuel pressure. But it isn't quite that easy. The ECM outputs a PWM command from 10-50% at 100Hz (half the desired duty cycle), and the FPDM basically opens and closes the circuit at twice the ECM's command at 20+kHz! Quiet and efficient, but too fast for our needs. The pump wants its command from 10-100% at 500-1500Hz. And the ECM needs an "OK" status at 50% and 1Hz from the FPDM. We need a way to achieve that.
Ok, some people are probably asking "what the hell is PWM?" PWM stands for "Pulse Width Modulation." It's a digital way of achieving an analog output. It gives an output by pulsing the signal to the "high" status for that fraction of the time period, then "low" for the rest. For 20% it will go high for 20% of the time period, then low, then repeat it. If your frequency is 1Hz (1 cycle per second) it would go high for 0.2s, then low for 0.8s.
Thankfully microcontrollers have gotten stupid cheap lately. I picked up an Arduino Nano for about $10. Using that microcontroller it's going to poll the ECM command, and then send the desired duty cycle to the pump at the correct frequency.
Now there's a couple wrinkles that make it a little more challenging.
The Arduino output is regulated to 5V. The pump and most car hardware takes battery voltage. So we need a way to step it up. The answer is using a couple transistors, a diode and pull up resistor. That way we can control a 12V signal using the 5V output. The most common way to do this produces an inverted signal (i.e. your 20% becomes 80%), but we can fix that in the code because hey, code is free.
The Arduino takes a 7-21V input, but you'll fry the integrated voltage regulator in short order if you spend time above 12V. We all know your car is going to be in the 13-15V range while running so you need a way to fix that. A 9V regulator will help with that, and for only a little more than $1.
There's the basis for the project. I'll update with circuit diagrams, code, components, and pictures as it progresses.
Right now the only thing I'm wondering about is how to make a clean finished project. I've got a breadboard to test it with, but I'm not sure I'm ambitious enough to make a dedicated PCB for it.
And I'm sure I'll end up needing some help from you electrical nerds on here.
For now all told with the components and chip I'm still slightly cheaper than a Hobbs switch.
