DIY Syringe Pump Version 1
Posted by aonomus on February 1, 2010
So my last post (ranting on McMaster Carr’s international shipping policy) had a sneak peek of a little mini project that I’ve been up to over the last few days on paper, and finally last night (thanks to the folks over at Hacklab.to for their help too) I got some acrylic plastic laser cut to build the device.
So if you haven’t guessed from the title of this post, this is a DIY syringe pump – a stepper motor driven plunger to gradually pump liquid using a syringe in a controlled fashion. The principle of a syringe pump is to use a fine-pitch leadscrew to depress the plunger of a syringe, or multiple syringes – this is typically accomplished by using a stepper motor and a microcontroller to keep track of the infused volume. Now because of the high-resolution required by these devices, they are typically engineered well, and because of alot of medical applications, they are overengineered for safety reasons; this drives up the cost to at least several hundred dollars for a basic unit, to even more expensive versions when you add on more features.
I’ve noticed that many different things could use such a device, but initially this was born out of an interest in microfluidics, which may require a device to handle microliter volumes (thats 0.000001L for those unfamiliar with the metric system). Other uses that started to appear would be for aquarium enthusiasts who need to infuse medication over a long period of time, DIY CNC machines that extrude some sort of paste (read: frostruder), slow addition of reagent to a reaction (given that it would be unreactive with the syringe, but that can be solved by using a glass one and teflon tubing), or perhaps even automated titrations.
Now for my use and budget, I decided that I could hack it out of something pre-existing, but instead I wanted to make it out of really accessible parts (limited to Digikey and Home Depot essentially) so that anyone could build one if they so desired, but also so that they could repurpose it to suit their needs more specifically. Another reason why I decided to build it myself is so that I could continue to update it with more features such as interfacing with a computer, or perhaps something like a CAN network to simplify the control method down the line.
Design and Construction
As I began to think about the problem, I looked at how amateurs had previously built syringe pumps as extruders for various purposes, and examined an actual syringe pump. Most of the devices that I examined (and all the successful ones in use that have been replicated, or are commercial products) used leadscrews to actuate a plunger, or a block that depressed the plunger itself while holding the syringe body stationary. Every device also used some sort of metal rods as both guide-rails and supports for the entire mechanism, and most lacked any sort of absolute or relative position sensing.
Having seen most device designs, and lacking any sort of access to a mill, I decided to construct my syringe pump out of acrylic plastic in layers by laser-cutting each piece and assembling everything together around a pair of threaded rods as the core structure of the device.
As for the actual assembly, 1/4″-20 threaded rods were used for the structural supports and plunger guide, while a 10-24 threaded rod was used as the leadscrew (backlash isn’t an issue unless you are changing from pumping to withdrawing liquid in your program during an experiment). A standard sized NEMA 17 stepper motor was used to actuate the plunger, however NEMA 17 motors of different thicknesses could be employed if more torque or lower power consumption was required. A modification made partway through the project was to place smooth metal tubing over the threaded rod to smooth the motion and prevent binding. At the time only 3/8″ tubing was available, while ideally I would have used 5/16″ tubing for a snug fit around the 1/4″ threaded rod.
It would be somewhat trivial to scale this down and use smaller components to dedicate a system to smaller syringes, or re-design the system to use gears and increase the resolution without requiring more accurate components, and these are all potential improvements to be made once I have actually conducted some testing.
Accuracy, Precision, Resolution?
So, despite the entire device being made with parts from Home Depot and otherwise catalog parts, how precise could it be? I set out to do a few calculations and this is what I came up with
Given a step size of 3.9 degrees/step, 10-24 threaded rod (size 10, 24 threads/inch), and a 1mL syringe which I measured to dispense 0.4mL/inch, the syringe pump should dispense 1.805uL/step. With a 10-32 threaded rod, that would drop down even further to 1.354uL/step.
Given a more accurate stepper motor of 1.8deg/step, those values would decrease down to 0.9uL and 0.67uL per step.
Given that the current version built from commodity parts is already this accurate, and readily accessible parts could increase the resolution almost 3 times, I’m pretty satisfied overall. I still have concerns for the arrangement of the motor and acrylic plastic, since the syringe carriage could flex given enough strain if a syringe began to bind and change the actual vs calculated dispensed volume.
So what is next after this? At the moment the syringe pump is too large for practical use – I plan to rearrange the unit to an under-over configuration, with the stepper motor under the syringe holder itself. Also, I plan to make a few more modifications to eliminate backlash, so that pumping in both directions can occur. Finally, I need to actually place the second revision pump onto a enclosure, add a microcontroller, LCD, other control features, and some sort of computer control (perhaps serial, CAN, etc). Once its actually to this point, I hope the cost will still be under $50-100 – which easily blows away the cost of a cheap (real) syringe pump.
Perhaps other uses can be found – a modified version as an extruder for 3d printers (Reprap and the makerbot), a infusion pump for medicine dispensing for aquariums, a small pump for automated titrations, etc.
First test – lots of binding on the threaded rod
Second test – after addition of a smooth copper sleeve over the threaded rod
Second test – closeup video of a 1mL syringe with a 23 gauge needle showing small volume dispensing of fluids.