PWM Fuel Pump Control
01-18-22, 02:33 PM
#1
PWM Fuel Pump Control
I was going to stick this all in my build thread but figure maybe better here as I'd like to hear feedback and input on the topic and maybe it's worth building upon as its own thread.
So my primary pump has once again met its end. As has been consistent, during an extended use / drive, the pump began losing pressure. I say once again because although it was the first Walboro, it's the 3rd pump since I've put out to pasture since going single turbo (upgraded from the Walboro 255 that served the stock twins for the better part of 40k miles under stock fuel pump ECU control). Was it the heat? Was it counterfeit? Are Walboro's just destined to last 4,000 miles and 8 months at full 12-14 volts? I don't know.
I contacted the seller from Amazon and got no response. Then contacted Amazon and didn't get a warm fuzzy about the authenticity, but I did get my money back. Next stop was Summit to order a new pump and ensure authenticity. This time around I went with the 90000274 pump - which is known as the "450". It's also the same as my secondary. I figure with two I didn't need the big pump, and maybe the less amperage consumption would be a good thing for it.
The next issue was the possibility of the heated fuel causing the premature failure. I'd killed 2 340lph pumps before. Very similar failures - during extended use road-trips. In fact, during one road trip the pump was acting up, and I had a little under a half tank of fuel. I filled up and the behavior immediately resolved itself. Which makes me think heat is a key factor here. I am running a heat sink on the fuel filter under the car, but the fuel still gets very warm. The key difference between the way I'm running the pump and the way the OEMs run the pump is that they reduce the pump speed while the car is at idle and low throttle through pulse width modulation. I figured since I have everything except the solid state relay needed for the job, I should get off my lazy keister and see about setting this pump up via such a method in order to see if this would extend its life.
I did a ton of research because I had a general idea but not tactically how to pull this off. At first I was curious if I could utilize the factory fuel pump ECU to do the work on the pump. Maybe I could have. But there were too many unknowns for me to get comfortable. So I went ahead and picked up a 100 amp solid state relay. Should be way overkill for my ~ 20 amp maximum on a single pump. There are some threads out there online that talk about this approach. It's very common with cooling fans. Less common on fuel pumps. The general consensus I had heard was "just run them 100% they'll be fine." My experience has differed.
The concept of PWM is you're toggling the device on and off very quickly. I set mine at around 150 hertz.
The biggest area of concern for me at this point is I don't have 100% confidence that these Walboros like being controlled via PWM and further what the ideal frequency is. I called Ti Automotive and they said it's fine, just try to target no less than around 9 volts with the pump. Easy enough, but not confidence inspiring. 60% duty cycle will measure about 9.1 volts on the meter. Keep in mind, you're sending full voltage, it's just switched very quickly, so it gives an effective reading of less than the full beans.
The wiring was pretty straightforward and there's a nice diagram put forth by Haltech on their website as to how to wire up the SSR. But there are a few caveats. One - use a pullup (1k ohm) resistor between the SSR power and the ground signal coming from the ECU. And secondly, have a flyback diode between the pump + & - to prevent the inductive current from "flying back" toward the SSR and destroying it during switching. Thirdly, mount the SSR in a nice heat sink. These can get hot and die and that would be bad.
Given the foul weather at the moment, I haven't taken the car out of the garage. But I have idled it and set the duty cycle and adjusted the fuel pressure to compensate for it. That's one thing that I'm still a little iffy about, and I'll have to take the car for a drive to ensure the fuel pressure is behaving as hoped / expected. I may have to get into injector pressure differential to accommodate for the reduced duty cycle at low load - but I'll see if the sole mechanical adjustment I made is adequate once I can take it for a drive to confirm.
So my primary pump has once again met its end. As has been consistent, during an extended use / drive, the pump began losing pressure. I say once again because although it was the first Walboro, it's the 3rd pump since I've put out to pasture since going single turbo (upgraded from the Walboro 255 that served the stock twins for the better part of 40k miles under stock fuel pump ECU control). Was it the heat? Was it counterfeit? Are Walboro's just destined to last 4,000 miles and 8 months at full 12-14 volts? I don't know.
I contacted the seller from Amazon and got no response. Then contacted Amazon and didn't get a warm fuzzy about the authenticity, but I did get my money back. Next stop was Summit to order a new pump and ensure authenticity. This time around I went with the 90000274 pump - which is known as the "450". It's also the same as my secondary. I figure with two I didn't need the big pump, and maybe the less amperage consumption would be a good thing for it.
The next issue was the possibility of the heated fuel causing the premature failure. I'd killed 2 340lph pumps before. Very similar failures - during extended use road-trips. In fact, during one road trip the pump was acting up, and I had a little under a half tank of fuel. I filled up and the behavior immediately resolved itself. Which makes me think heat is a key factor here. I am running a heat sink on the fuel filter under the car, but the fuel still gets very warm. The key difference between the way I'm running the pump and the way the OEMs run the pump is that they reduce the pump speed while the car is at idle and low throttle through pulse width modulation. I figured since I have everything except the solid state relay needed for the job, I should get off my lazy keister and see about setting this pump up via such a method in order to see if this would extend its life.
I did a ton of research because I had a general idea but not tactically how to pull this off. At first I was curious if I could utilize the factory fuel pump ECU to do the work on the pump. Maybe I could have. But there were too many unknowns for me to get comfortable. So I went ahead and picked up a 100 amp solid state relay. Should be way overkill for my ~ 20 amp maximum on a single pump. There are some threads out there online that talk about this approach. It's very common with cooling fans. Less common on fuel pumps. The general consensus I had heard was "just run them 100% they'll be fine." My experience has differed.
The concept of PWM is you're toggling the device on and off very quickly. I set mine at around 150 hertz.
The biggest area of concern for me at this point is I don't have 100% confidence that these Walboros like being controlled via PWM and further what the ideal frequency is. I called Ti Automotive and they said it's fine, just try to target no less than around 9 volts with the pump. Easy enough, but not confidence inspiring. 60% duty cycle will measure about 9.1 volts on the meter. Keep in mind, you're sending full voltage, it's just switched very quickly, so it gives an effective reading of less than the full beans.
The wiring was pretty straightforward and there's a nice diagram put forth by Haltech on their website as to how to wire up the SSR. But there are a few caveats. One - use a pullup (1k ohm) resistor between the SSR power and the ground signal coming from the ECU. And secondly, have a flyback diode between the pump + & - to prevent the inductive current from "flying back" toward the SSR and destroying it during switching. Thirdly, mount the SSR in a nice heat sink. These can get hot and die and that would be bad.
Given the foul weather at the moment, I haven't taken the car out of the garage. But I have idled it and set the duty cycle and adjusted the fuel pressure to compensate for it. That's one thing that I'm still a little iffy about, and I'll have to take the car for a drive to ensure the fuel pressure is behaving as hoped / expected. I may have to get into injector pressure differential to accommodate for the reduced duty cycle at low load - but I'll see if the sole mechanical adjustment I made is adequate once I can take it for a drive to confirm.
The following 2 users liked this post by KyleH:
firelizard (01-18-22),
Frankn (01-22-22)
Thread
Thread Starter
Forum
Replies
Last Post