Peristaltic Pump

I have designed and built some peristaltic pumps for pumping cyanoacrylate (i.e. super glue). The stuff is about as viscous as water, and I want to dispense it in tiny drops. Peristaltic pumps are perfect for high pressure, low volumetric flow systems, plus there is the added bonus that you can easily replace the tubing if it gets clogged (i.e. the super glue dries out, which can be pretty often) instead of replacing the whole pump. You can buy a cyanoacrylate dispenser for a couple thousand dollars, but I think I can build a small pump for less than $100. Peristaltic pumps are very simple– it basically squeezes the tube.

I designed mine to work with a Nema 17 stepper motor. It uses 1/4″ and 1/8″ laser cut acrylic, with some tapped and counter-bored holes. The rollers are cheap (~$4) ball bearings on screw standoffs with a pololu mounting hub to attach to the motor shaft. I use some laser cut acrylic pieces to hold the tubing in place. There are a couple pieces where I drill a hole in a laser cut piece on the laser cut edge, which gives me a “real” 3D piece. (i.e. It has features in more than one axis.) My CAD model is below.


It builds out easily, though looks a bit funny with clear acrylic and clear tubing.

IMG_3351 IMG_3354The black part is a piece of electrical tape that temporarily fixed the problem I describe below.

The first issue is that I got the distance between the rollers and the siding (i.e. how much the tube is squeezed) a little big. This is because I’ve never measured the kerf of the laser cutter (how much material is removed by the actual laser cutting and on which side of the line) so I had to fix that experimentally. This dimension (or feature?) is called occlusion and can be defined in different ways but always tries to describe the amount the tube is squeezed. But I *still* wasn’t squeezing the tube enough! It was only at one point of the cycle, for maybe 15 degrees. What was going on?

I finally noticed that the top piece of acrylic — the one that holds the tubing in place against the roller — was moving slightly when the roller went around that point. I can feel it if I put my finger at the back of the stack while the roller goes by. What I think is going on is that the top left screw, which holds the stack of acrylic together, isn’t applying enough force to hold it in place when the roller goes by. To fix this I used washers with the screws, which spreads out the force from the screw to a larger area.

Now it works extremely reliably. I did some testing to determine how well it was working. I used the pump to apply drops of glue to a part, then measured the location and size of that drop with imageJ.


Drop size over time (i.e. angle). Colors represent different drop size.

One of the coolest things I saw was basically the point, every 90 degrees, where the rollers swap. This is one of the downfalls of peristaltic pumps: every time a roller loses contact with the tubing, you get a pulse where no liquid is pumped. Each line in the graph is a different drop size, i.e. a different number of steps with the motor. So you can see that for a larger number of steps, the pulses happen more frequently as it takes fewer drops for the rollers to travel of the tubing.

I plan to fix this by sensing where the rollers are and basically doing a double drop whenever the roller leaves the tubing.

Overall, so far, huge success.


5 thoughts on “Peristaltic Pump

  1. Hi Katy,

    Interesting blog on the peristaltic pump and your other projects.
    I have one question.
    If you were to change the orientation of the pipe by 90 degrees, what would be the effect on the pump efficiency?


    1. If you have the tubing turn at a 90degree angle anywhere in the system, before or afterward the pump, you would definitely decrease the pump efficiency because you’d be increasing the head loss of your tubing system. (This should make sense — you have to push harder to get the fluid to flow around a corner than to go straight.) Calculating the head loss is one of those dicey engineering areas where it’s incredibly hard to be precise but there are lots of approximations that work in most cases. Just look up “head loss for angled pipe” to get some starting places.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s