View Full Version : Linear actuator injector

07-19-2012, 12:59 PM
first I want to apologize, because I am very inexperienced and I have no interest in building a robot. Having said that, I still dare to hope that someone will help me with my chemistry project :-)

So, I want to build automatic liquid injector using linear actuar, timer and syringe. I did some research and I learned that actuators that you people use are very precise and are controlled by electric impulses, which come from control board (one impulse=one step). Because of my special need, I need actuator that has at least 10 cm range of motion. The precision of some actuators I seen around is 0,01 mm per step and better, which is astonishing and quite overkill for my purposes. The strength is also overkill, I could do with as low as 50 N. I dont want to use computer to control it, just timer.

Ideally, I would like to program the control board, then take it off, plug it into electricity with a timer and transformer and when the timer goes on it moves for certain distance and stops. Then it repeats.

So I was wondering if anyone got an idea for suitable setup ? I have already found actuator that I think might be good. Its called NEMA 17 (http://www.osmtec.com/linear_stepper_motor_nema_17_e.htm). By the way, I have 50 dollars budget. Thank you for your help

07-20-2012, 01:58 AM
With a $50 budget it will be hard to find any fully functional linear actuator setup like you are looking for I imagine. Have you considered scavenging old/broken electromechanical devices for motors, gears, and other parts? Old printers are a dime a dozen and probably contain some DC motors with gears you could use.


07-20-2012, 05:32 AM
As long as it doesn't have to be very fast, you can use a stepper motor with a pinion plus a fitting rack (like a flat gear) to make a linear actuator. To actually make the stepper motor "go" you need a motor driver, and to actually drive the motor driver, you need a microcontroller. Looking at the bargain basement Chinese suppliers I know, I could probably make the Bill of Materials come in right under $49, although that includes things like soldering an AVR microcontroller to breadboard instead of buying a ready-made circuit for the microcontroller.

If you're handy with the soldering iron, and the coding, and the mechanical assembly, needed, then you could do this. If you're not that multi-disciplinary, then you're probably looking at a higher cost, and might as well buy a dedicated linear actuator or at least a dedicated linear motion assembly (such as used for small X/Y tables and CNC mills.) The total will probably run you around $200, though.

07-20-2012, 08:56 AM

Regarding the mechanical side, another option for you to consider would be getting a section of threaded rod, attaching it to the motor and using it to drive a partially constrained nut back and forth along the axis of rotation. Extend the nut off to one side and attach your syringe to it. The nut may be replaced with a plate that has been drilled and tapped to fit the rod. Given that your accuracy and force requirements don't appear to be that high this might be acceptable. If 0.01 mm and 50 N are overkill, what would an acceptable range be?

You'll still run into control issues as jwatte has mentioned though, and I'll second his advice on the uC and making one.

07-20-2012, 09:47 AM
first of all, thank you for your replies and suggestions. I understand that the budget is too low, so I am willing to invest around 120 dollars. To answer your questions, I am handy with mechanical assembly and usual electric components, not coding though. My idea for mechanical solution is similar what you suggested. By the way, I need at least 50 N, 5 mm per step, speed doesnt matter as long as its done.

Is it really that expensive and problematic to control these relatively cheap motors ? Are there linear actuators that dont require control board ? I had an idea, where you got actuator that moves at constant rather low speed (could be lowered with regulator) and moves when there is electric current. I would set the timer to go on for example for 1.5 sec, allowing the actuator to move only for 1.5 sec, thus moving the syringe for desired distance. I was unable to find such actuator, does anyone know which would do the job ? Or maybe cheap control board that would send electric impulses whenever it has electricity? Thanks

07-20-2012, 10:29 AM
With those requirements, 5 mm per step gives you a fair bit of leeway. What tolerances can you afford? 5 mm +/- 3 mm is one thing, but +/- 0.1 mm would be another.

This timer you have, what kind of output does it give? If its output is constantly on until switched off, there may be some hope for it. If it only gives a single pulse when time's up, things may get a little more complex with a latching setup.

Given your requirements, if you can solve the timing issue then a simple DC motor controlled by a MOSFET might be acceptable. But for steppers or servos, then yes it is that expensive since they both require specific switching or timing patterns to get anywhere. The issue with finding a "cheap control board that would send electrical impulses whenever it has electricity" is that such a board would be so limited that it wouldn't be worth making, like how you don't see 1 cylinder engines on cars. You'd also have issues when you want to go in the other direction and when it's first initialised.

A possible motor for you might be made by Tamiya, part number 72001 (I think that's the part number). Consists of a motor plus a planetary gearbox to reduce speed and increase torque. Neat little piece of equipment, and since time isn't an issue for you (partially tongue-in-cheek: is one movement in 3 days acceptable for you?) it might be sufficient. Control would depend on what kind of timer setup you have though. Maybe you could rig something with a 555 timer and a MOSFET, but I haven't played with such things in ages.

07-21-2012, 05:14 AM
Thanks for your reply. As for the tolerance, I want to inject milliliters, I thought I will make the mm/ml ratio 1:1, it moves for 1 mm and injects 1 ml of liquid, but of course I can pick different volumes of syringes to make this ratio more suitable for the motor, for example 5:1 (mm:ml). For the actual tolerance, I want precision +-0,5 ml, with total volume of the syringe at least 20 ml. Now for the timer, I dont actually have it, but after two minutes of googling I found this one http://www.mistking.com/Digital-Seconds-Timer.html, its ordinary timer with ,,Ability to precisely set ON times on per second basis,, so it works just like usual turn on/turn off switch except its done not by hand, but the timer. As for the speed, it should be done in, I dont know 2 hours max and I need to inject once a day. I really like the motor you suggested, is it still suitable after I specified my needs ?

Thank you for your interest and help, its really appreciated :)

07-21-2012, 06:02 AM
Motor and gearbox should still be suitable if you find a suitable regulator for it.

Apart from different volume syringes you could also work on different pitches for the screws to increase the resolution, e.g. an M6x1.0 mmthread will move the nut by 1 mm for every turn, while an M3x0.5 mm thread will move it by 0.5 mm.

Once you've found an appropriate regulator I think that should be all the electrical components you need. I can't see any glaring issues, although you might have a bit of fun figuring how long the timer should be on for each cycle and what speed the motor will need to go at.

07-21-2012, 06:19 AM
Lol we live in amazing times, liquid soap dispenser :-) I even found one where you choose how many ml you want, has 400 ml (!) tank and is triggered by movement, all that for 20 bucks.

07-21-2012, 06:41 AM
"[email protected] syringe tool" (http://fabathome.org/wiki/index.php/Model_1_1-Syringe_Tool)
"reprap paste extruder" (http://reprap.org/wiki/Category:Paste_Extruders)
"stepper actuator" (http://www.anaheimautomation.com/manuals/forms/stepper-actuator-guide.php)

I know I have seen a tutorial somewhere where a guy pulled the rotor out of a surplus hybrid bipolar stepper motor, used a vice/hammer/arbor press to remove the original shaft from the rotor, and used a tap to create internal threads in the actual rotor of the motor. He then reassembled the rotor into the motor and used the appropriately sized threaded rod with a nut at the end to keep the rod from rotating, thereby converting the cheap rotary stepper motor into a precise linear actuator. The volume extruded (assuming incompressible fluid with sufficient viscosity to keep it from freely draining from the syringe but not so high that the fluid flow lags too terribly after the force is applied to the plunger) should be the syringe's cross-sectional are multiplied by the pitch of the threaded rod and the rotation of the motor's rotor (V = CA [mm^2] * p [mm/rev] * N [rev]). Most hybrid bipolar stepper motors that I have seen use a step resolution of 1.8 [degrees/step], giving a volumetric resolution of ((CA * p)/200) [mm^3/step] (V = CA [mm^2] * p [mm/rev] * 1/200 [rev/step] * S [step]). Pololu sells some rather nice stepper motor driver boards that require only two pins to control, one for selecting direction of rotation and the other to make the motor increment a step. This could allow even the simplest/cheapest microcontroller to select a direction, start an internal timer, and increment the stepper motor one step every time interval (for a certain number of intervals) to produce a desired extruded volume.