I put a Velleman K-8004 Pulse Width Modulator on my Walbro and can't seem to get reliable settings. I have the MIN voltage all the way CCW and can control the pump with the MAX voltage at halfway or above. The FREQ is what is apparently controlling how hard the motor works. If I bump the starter and set the FREQ to put out 15psi, when I start the truck it pops to max psi and stays there. I think the voltage spike when starting the truck interferes with the circuit. If the pump goes to max psi I am unable to get it back below max with the FREQ, but have to stall the pump out with the MAX voltage turning it CCW. That resets everything. Any ideas?....

 

Velleman K-8004 PWM circuit is prone to oscillation and will cause the lift pump to run at full speed if it oscillates. I played around with this circuit and found that the power connection to 12 volts needs to be decoupled using a 10 to 20 uF tantalum capacitor.

I am assuming that you have connected the “IN” terminal to the “V ref” terminal to allow using the pot to control max pressure. If this is true than try more power supply decoupling to kill off the oscillation.

When you get this working to regulate the pressure at say 15 psi, the pressure will drop off as engine load and RPM increase. So keep an eye on your pressure gauge.

Someday I'll find the time to write up a circuit for a proper PWM lift pump circuit, but it may be awhile.

 

Right now I have the control voltage to the Vref and a 1K resistor to the negative control voltage. Do you have the Vref connected to the control voltage through a 10 to 10uf tantalum capacitor (as opposed to the flux core capacitor I have on my time machine....)? Does Radio Snack have those for cheap? Tantalum sounds expensive.... Right now I only feel comfortable with my Swagelok bypass giving me 20psi to the VP.

 

WARNING, WARNING, WARNING, Geek Speak to follow. There is nothing to see here, please move along.

Sorry for my cryptic post, no flux capacitors are needed. I was disappointed with the velleman kit because the SG3525 is now obsolete and the circuit is unstable. I also had the problem of the lift pump randomly running at full speed because of the instability of the circuit. Here is what worked for me before I abandoned the idea of using this kit. First C1 and C7 need help in decoupling the chip from oscillation. A tantalum capacitor in parallel with C1 helps, just watch the polarity. I just found a Radio Shack number, 272-1436, which is a 10-uF 16V capacitor, might work and it is less than $2.

I found that when setting the frequency at maximum it was more prone to instability so I backed it down to near mid-range. If you are using the frequency control to vary the pump speed there is probably instability. It is not clear to me if you are using an external potentiometer to control the lift pump pressure. If so, this may be another point of instability and another 10-uF cap between “IN” and “GND” will help. What I was writing about was to simply add a wire from “VREF” to “IN”. Now adjust RV2 to set pressure, no external pot needed.

 

I agree with you on the instability. I was using freq to control pump speed. Adjusting voltage at RV2 did little to move the pump either way as I was looking for a solid 15psi and the freq had to be adjusted to about 75% to get that at any voltage. I think I'll use the Velleman to control something less important than my LP. My Swagelok popit valve works like a champ anyway and seems bulletproof.

I love your disclaimer! It sounds like you could design and market your own PWM LP controller since I didn't even know the chip was junk when I bought it. Is there a newer IC out there that doesn't have the oscillation problem?