Help with air pump and valve setup


#1

I am working on a project and could use some wisdom. The goal is to have 5 variable flow outputs from an air pump. I could use 5 separate air pumps and control the output at the pump level but would rather have a single pump with a way to control the output at 5 variable control valves. I can’t just use a basic solenoid as I nee variable flow at each output.

The pump has a max output of 12psi and flows <4L/min. I have done just a bit of reading on variable control valves which is what I “think” I need but I’m a bit out of my depth. I’m a novice working in hardware and particle. My background is in application development but am trying to prototype an idea I have.

Is there a better approach I’m not thinking about? Do you know of an inexpensive control valve I could use?


#2

@jakedempsey, do you need 5 different outputs, each with a different fixed flow rate? Or do you need 5 different adjustable (non-fixed) outputs?

If I understand correctly, you could use a manifold to go from your pump to 5 outputs each with a solenoid and manual flow rate valve that you adjust for each output, making sure your expected total rate dosen’t exceed your pump’s max output.


#3

Can you explain the system in a little more detail?
For instance: will the air be released at a constant depth in a fluid ( a defined/known discharge pressure) ?

The air pump has a defined Pump Curve that you can reproduce if it’s not easily available. For any given system pressure, there is a defined flow rate the pump will produce.

The discharge piping and endpoint make up a System Curve.
This curve relates the headloss at various flowrates.

The actual operating pressure and flowrate for the Pump are where the two curves intersect.

If the 5 outputs have substantial differences, this would get tricky but not impossible.
Similar to what @peekay123 suggested:
You could use a flow meter on each of the 5 outputs, each with a proportional valve.
Add 1 more output and valve for a waste gate if you need fine tuning.
Proportional pneumatic valves and tiny flow meters are pretty cheap, but not extremely trustworthy or accurate.


#4

I’ll try my best to explain. Again, I’m trying not to sound dumb lol but this is new territory for me. I’ll be using a Koge pump found here: http://www.koge.com/file_data/20160803-081406.pdf

My thought was I have an air line going from this pump to a manifold. On that manifold are 5 variable control flow valves each with their own output and check valve to prevent backflow. I basically just want to “split” the air coming out of the Koge pump into 5 outputs but want control at each output how “open” it is. I’ll use a separate control to control the voltage to the Koge to adjust its total flow. The 5 outputs share that total air flow.

The air is not being released in a fluid. I am using the output air at each of the 5 locations to create a bernoulli effect for atomization of liquid.


#5

So as a really simple prototype I have a Photon connected to a potentiometer which is read using an analogRead and controls a pwm pin to a transistor which is being used to control the voltage to a much smaller 6v pump. I did this to just show I ‘could’ easily have 5 separate pumps and control the voltage to each pump to have 5 variable outputs but that of course requires 5 pumps. I’d rather have 1 pump with a way to control the volume of air for the 5 outputs.


#6

This is my current testing board… I was playing with buttons, switches, and the led so sorry thats in the way a bit. I don’t have a digital pot so using this manual one. Ideally I’d have a digital pot and control it via my volume up and volume down tactile buttons for each output.

I found the setup for controling the pump via the transistor somewhere online… took me a while to do and I have no idea if its right… it does work but I don’t know why each piece is needed as I’m a total noob.

So right now my switch will activate the system. If its off the oled reads OFF. When its on the Photon will read the pots value then create a mapped value of 0-255 for the led and write to it so the led intensifies as you dial it up. It also does a mapped value of 230-255 to do a write to the separate breadboards pin tied to the base of the transistor. The oled will display the values being used for the led and the motor. I picked 230 so that the pump starts at 3.5v which is its operating voltage min. At 255 its only getting 4.6v b/c of my 4.8v power source I’m using for testing.

So my thought was I could swap the manual pot for a digital one. Have a volume up and down button for each digital pot (5 pots since I want 5 variable outputs). I could replicate my current pump setup. Read from 5 pots… write to 5 pwm pins of 5 transistors etc.

However, I think it would be better to have one stronger pump… use one pot to adjust its total flow… and have a manifold with 5 variable control valves (still using 5 up/down btn setups with digital pots) but I don’t know if those are expensive or how to drive them or if its just dumb.


#7

We’d typically do that with a relief valve on the 6’th outlet of the manifold, to set the maximum manifold pressure, and let the pump run at this steady-state condition.
image
But that might be hard to do accurately at such a low pressure.

Your project may be simplified by using (5) individual pumps that you originally considered.
It would definitely be cheaper.

Here are some pumps:
Quick Search

I’ve had decent luck with this tiny little guy:
3V

And great luck with this one (probably too large for your project x 5)
12V


#8

I’m thinking the smaller pumps may also just be the easiest… I can’t use the manual control manifold as in your message b/c I need to control that flow via the photon as well as the overall output of the singe pump. I think just using multiple pumps is the answer … I’ll continue down that path for now. I have some small pumps now… I have the koge pump and its actually a great pump! I have the Diminus pumps as in my screenshot which are only 6v and work well for my quick prototypes but will look at the ones you suggested as well. Thanks again for all the help!