PWM Fuel Pump Speed Control

[VL400]

So has anyone played with fuel pump speed control using PWM? I have upgraded the fuel pump to allow for E85 and found the fuel pressure at idle has increased significantly above what the stock regulator is meant to be. The change is going to a tank from a VS V8 using a single in tank 340LPH pump. Previously was in tank lift pump and external pressure pump, stock VL pumps and VN tank.

I dont have a log with fuel pressure from before the upgrade but its running far richer now than it was before with no cal changes (although there are a few changes to the rest of the combo its the same delivered fuel PW at idle, just rich now), so barring any regulator issues am suspecting the pump is supplying more than the reg can return (pressure on the return side is next to nothing so not restricted there).

So with the '424 PCM capable of PWM fuel pump control had a little play with the code, factory code allows only 100% duty cycle so a regular relay can be used. Added adjustable frequency and duty cycle control (mapped to engine load), plus mode 4 control of duty cycle for testing and knocked up a dodgy MOSFET circuit to replace the relay. Have some Infineon BTN7960B on order to make some more permanent.

The fuel reg is meant to be 43.5psi, so not too far off at reduced duty cycles. It does look like at 100% DC and normal running voltage (~14V) the pressure increases, this is with the engine off so a worst case where all fuel is returned it may be at the limit of the regulator.

Attached are some logs and excel data if anyone is interested in taking a look or has any experience in what just adding a larger pump does to pressure and AFR. Its been mentioned a few times on here where removing the dual speed pump control on S/C V6 has caused an AFR shift but not sure if anyone has actually done any testing to see by how much pressure changes?

[The1]

it was something i wanted to look at with the L67 PCM's on VS-VY but it was an extra line that was only in those PCM's

[Holden202T]

i didn't have much luck with pwm control of fuel pumps so shelved the burnt out PCB!

[vlad01]

wouldn't it be better to just use a better aftermarket reg to solve the issue?

[VL400]

i didn't have much luck with pwm control of fuel pumps so shelved the burnt out PCB!

Yeah that didnt work out too well Have since found that the SSR you tried switches pretty slow and just gets hot. The other driver you tried no idea what went wrong with that.

wouldn't it be better to just use a better aftermarket reg to solve the issue?

Yep a bigger reg would fix the problem, but reducing the pump flow at low load areas means not heating the fuel as much. I can monitor fuel temp with the flex sensor so will see how much of a difference it makes. Also lowers noise a fair bit at idle. The latest trend with return-less systems and PCM controlled fuel pressure are BLDC fuel pumps which can be almost silent at light loads.

[vlad01]

I like pump noise though

Is there such thing as a self regulating pump? one that works on demand based on back pressure?

[VL400]

I havnt seen a pump on its own that is a standalone self regulating unit, can get the entire returnless in tank module that are pressure regulated. Plenty are now using PCM controlled fuel pressure via a CAN PWM speed control module.

And further to the first post, tried a reg off a spare engine to rule out that. It performs identically except all readings were 1.1PSI lower. Played around with frequency and 400Hz seems like the sweet spot for this pump, reduced noise and also efficiency of pressure for a given duty cycle.

[VX L67 Getrag]

NICE WORK!

Sounds like a good option to have to save on idle & get every bit out of top end!

[Jayme]

in an ideal world, once you calibrate the load table to your liking, you would have some sort of fuel pressure sensor input to the PCM, and if the pressure ever dropped below a certain threshold while the engine was running, it would kick the pulse width of the fuel pump up a notch (eg multiply the output of your load table by 5%) . after its at 100% pulse width and pressure still drops, log a fault.

[VL400]

Finally got around to testing this out some more on a now running engine. The conclusion was as expected, the new pump is exceeding the return limits of the factory fuel reg. So the choices are, new large aftermarket reg or slow the pump down when the supply is not need. I chose the later, its more complex and adds another failure point in an already stupid setup so fits the theme better

The fuel pump speed control code I added has a 10 second (configurable) delay from starting to applying speed control. So can see at startup the fuel pressure is 51.5 PSI (dark blue bottom trace) until the fuel pump duty cycle drops from 100% to 50% (red trace). Fuel pressure immediately drops to 35 PSI. The two dotted traces are calculated figures, green is from testing and the base fuel pressure so should be the delivered pressure with 0 vacuum. Orange is desired fuel pressure, or fuel pressure that is corrected for manifold vacuum. This is calculated from the light blue trace of manifold PSI and base pressure - so 44.5 PSI minus 10 PSI = 34.5 PSI.

In this I have removed the vac line to the reg at 1:00min and pressure jumps to base pressure. Sucked on the tube and can now see pressure dropping like it should. Middle peak at 1:06 added some pressure and it goes above base, all good. With the pump running flat out no amount of vacumm on the reg caused the fuel pressure to drop, it was just pegged at 52 PSI constantly.