PWM for variable pump airflow + EMI question

[Brad]

I am working on an application using an air pump where I need to
change the rate of inflation. What I was doing in the past was
decreasing the inflation rate by lowering the the voltage to the
pump. With this approach I am operating the pump out of spec and
sometimes the voltage is too low for a higher load and the pump stops!

I was thinking of changing the duty cycle so that I could account for
variable flow rates. My microcontroller is connected to a MOSFET that
toggles the power to my 3.3V powered pump. Are there any potential
"gotchas" with this this setup?

From a EMI/RFI prospective should I be putting any type of filtering
on the line that is going from the microcontroller to the gate of the
MOSFET? What about some in-line resistance for ringing? Any caps on
the line?

Do pumps in general have any issues being used this way?

Thanks in advance for your thoughts!

-Brad

 

[[email protected]]

I am working on an application using an air pump where I need to
change the rate of inflation.  What I was doing in the past was
decreasing the inflation rate by lowering the the voltage to  the
pump.  With this approach I am operating the pump out of spec and
sometimes the voltage is too low for a higher load and the pump stops!

I was thinking of changing the duty cycle so that I could account for
variable flow rates. My microcontroller is connected to a MOSFET that
toggles the power to my 3.3V powered pump.  Are there any potential
"gotchas" with this this setup?

From a EMI/RFI prospective should I be putting any type of filtering
on the line that is going from the microcontroller to the gate of the
MOSFET?  What about some in-line resistance for ringing? Any caps on
the line?

Do pumps in general have any issues being used this way?

Thanks in advance for your thoughts!

-Brad

You will be definitely generating some noise. Try putting an am radio
on the same circuit as the pump, on the line side, not the pump side
of course, and tune to a weak station. Turn on the new pump controller
and set it for some midpoint pump value and see how bad the radio
noise gets.

Capacitors from each side of the power line to a good ground should
reduce the noise somewhat. Ferrite beads clamped around the three
leads bunched together would also help. You could also buy a premade
filter, there are many manufacturers out there.

First I would try to get the circuit to control the pump the way you
want it to, then worry about the nose it creates.