Fuel pump controller from a light dimmer
07-12-2009, 01:02 PM
#16
Nordschleife Master
Thread Starter
Looking to save weight? If you are, a single big pump with adjustable speed is probably lighter than two smaller pumps. Also, flatheading the fuel system with a microcontroller and an adjustable speed pump would save you a (proverbial, not metric) ton of weight.
Tuomo,
I may consider running a boost switch to override the switches and turn on the second pump. However I plan on being able to manually activate it, along with the higher boost setting which will need the fuel.
I will not be running a stock EFI system, nor will I have any stock wiring in the car. This car will have the absolute minimum on it.
I may consider running a boost switch to override the switches and turn on the second pump. However I plan on being able to manually activate it, along with the higher boost setting which will need the fuel.
I will not be running a stock EFI system, nor will I have any stock wiring in the car. This car will have the absolute minimum on it.
09-26-2009, 09:20 PM
#17
1st Gear
Join Date: Sep 2009
Location: NorCal
Posts: 1
Likes: 0
Received 0 Likes
on
0 Posts
Any update on this project? I also have a large Fuelab pump and wondered if you ever got this to work. I am considering using my Aeromotive pump speed controlled on the Fuelabs pump but wonder how it will respond to the sub-12V voltage reduction as a method of pump speed control. The PWM method seems much cleaner and I was excited to see your idea, as I was toying with using my TPS input in a similar way that you used your MAF.
Jim
Jim
09-27-2009, 12:15 PM
#18
Nordschleife Master
Thread Starter
Here's the status.
The car is running at constant 80% pump speed, which means it's pumping about 140 gph now. The fuel lines are 5/8" and -10AN from the tank to the engine compartment. The pump is attached to the tank in a way that the end fittings are wedged against the tank.
As a result of running the pump at 140 gph and the fittings touching the tank, the pump is as loud as an out-board boat engine. Once the car makes some actual power, I'll calibrate the speed and reroute one of line to eliminate the contact with the tank, so I am hoping it'll be better. It'll still be loud.
I didn't put the light dimmer in at this point. It's still a possibility, but I've also worked on another solution. My father made me the following controller as a birthday present and we programmed it together. It's Arduino plus a home made extension board. The controller takes analog signals from one or two MAFs, computes the total mass air flow from them in kg/h, computes a PWM signal to the fuel pump based on assumed air-fuel ratio, and then emulates the analog signal from any single MAF. Thus, in addition to running the pump, it would also allow one to one or two "strange" MAFs and then emulate any arbitrary MAF, including that of the 928. Not that I would, but I could!
(The program is based on a MAF function in a table and an inverse MAF function in another table, where the result from the first is the index to the second. The computation is therefore instantaneously fast for all practical purposes.)
The car is running at constant 80% pump speed, which means it's pumping about 140 gph now. The fuel lines are 5/8" and -10AN from the tank to the engine compartment. The pump is attached to the tank in a way that the end fittings are wedged against the tank.
As a result of running the pump at 140 gph and the fittings touching the tank, the pump is as loud as an out-board boat engine. Once the car makes some actual power, I'll calibrate the speed and reroute one of line to eliminate the contact with the tank, so I am hoping it'll be better. It'll still be loud.
I didn't put the light dimmer in at this point. It's still a possibility, but I've also worked on another solution. My father made me the following controller as a birthday present and we programmed it together. It's Arduino plus a home made extension board. The controller takes analog signals from one or two MAFs, computes the total mass air flow from them in kg/h, computes a PWM signal to the fuel pump based on assumed air-fuel ratio, and then emulates the analog signal from any single MAF. Thus, in addition to running the pump, it would also allow one to one or two "strange" MAFs and then emulate any arbitrary MAF, including that of the 928. Not that I would, but I could!
(The program is based on a MAF function in a table and an inverse MAF function in another table, where the result from the first is the index to the second. The computation is therefore instantaneously fast for all practical purposes.)
Any update on this project? I also have a large Fuelab pump and wondered if you ever got this to work. I am considering using my Aeromotive pump speed controlled on the Fuelabs pump but wonder how it will respond to the sub-12V voltage reduction as a method of pump speed control. The PWM method seems much cleaner and I was excited to see your idea, as I was toying with using my TPS input in a similar way that you used your MAF.
Jim
Jim
