Society of Robots - Robot Forum

Mechanics and Construction => Mechanics and Construction => Topic started by: meidou on November 21, 2011, 02:25:12 AM

Title: Very slow Linear Actuator
Post by: meidou on November 21, 2011, 02:25:12 AM
Hey Guys
This is my first post,and I am in need of some help.I am attempting to construct a syringe pump using a linear actuator similar to the following DIY:http://panic.berkeley.edu/~ghe/DIY_SyringePump/.The main reasons I am attempting to make the  linear actuator rater than buying one is cost and the ones I have seen are too fast for my needs.I need the arm to move at a rate of ~500um/sec.In constructing the arm I will be following the following schematics: http://codinglab.blogspot.com/2011/10/diy-linear-actuator-out-of-scanner.html (http://codinglab.blogspot.com/2011/10/diy-linear-actuator-out-of-scanner.html) and in regulating the speed I am thinking about using a voltage regulator.
Now here are my questions:
1.Do you guys know where I can find/purchase a motor that has range of 1-10RPMs.
2.I am attempting to use a steeper motor from an old scanner, will this fulfil the requirements?
3.To control the speed of the motor would a voltage regulator be effective?

I hope you guys can offer some insight on my questions.Any help is highly appreciated.

Thanks again
Meidou
Title: Re: Very slow Linear Actuator
Post by: SeagullOne on November 21, 2011, 02:59:39 AM
Hi, miedou.

Servocity has some gear motors that are 6 RPM and 10 RPM. Some are even as slow as 0.5 RPM. A lot of torque, too. Looks like they are under 30 dollars each.
I've never used stepper motors before, but I think they require expensive circuitry to control.

I don't think a Voltage regulator would be ideal to limit the speed of the motor. A very tiny motor maybe, but not most motors. Most consumer voltage regulators are made for low electrical current and lo wattage applications (like to power a tiny microchip). I'd imagine they can fry easily after that.

How will you be controlling the arm? With a microcontroller? Manual controls? If this is an autonomous application, you could try sending a analogue or PWM signal through a transistor or relay that feeds more or less power to the motor. Something similar can be done for manual control using a potentiometer dial.

Let me know if there's anything else I can do or explain.  :) I'll do my best.
Title: Re: Very slow Linear Actuator
Post by: newInRobotics on November 21, 2011, 03:01:15 AM
Hey Guys
Hi there   :)

This is my first post,and I am in need of some help.I am attempting to construct a syringe pump using a linear actuator similar to the following DIY:http://panic.berkeley.edu/~ghe/DIY_SyringePump/.
What sort of sick virus mutation related experiments are going to use it for?!  :(

1.Do you guys know where I can find/purchase a motor that has range of 1-10RPMs.
You can use any motor that has enough torque to depress the syringe. Speed control is done by PWM (http://www.societyofrobots.com/member_tutorials/node/228).

2.I am attempting to use a steeper motor from an old scanner, will this fulfil the requirements?
It will, if you have suitable stepper motor driver. You can buy one or make one from microcontroller and two H-bridges. IMO, it is an overkill for what you need. I'd use plain motor, threaded rod, some nuts, couple of bearings, something like in the image below, just smaller scale.

(http://www.lirtex.com/images/robotics/CNC/Threaded-Rod-Bearing.JPG) (http://www.lirtex.com/robotics/diy-cnc-machine/)
Click on the picture to understand better.

3.To control the speed of the motor would a voltage regulator be effective?
No, because voltage regulators waste energy by converting it to heat. You'd be much better off using PWM.
Title: Re: Very slow Linear Actuator
Post by: bens on November 21, 2011, 04:03:05 AM
Hello.

This 1000:1 micro metal gearmotor (http://www.pololu.com/catalog/product/1596) can achieve speeds from 1 to 14 RPM and might work for your application.  In general, I think a DC motor is more practical than a stepper motor for an application like this: you only need one H-bridge to control a DC motor vs two for a stepper motor, and the DC motor will generally draw far less current.  I suggest using a motor controller or motor driver to control the speed and implementing limit switches at each end.  If you set them up right (this involves using a couple of diodes), they will only disable the motor in one direction (the one that would break something) while still allowing it move in the opposite direction.

- Ben
Title: Re: Very slow Linear Actuator
Post by: newInRobotics on November 21, 2011, 04:41:54 AM
I suggest using a motor controller or motor driver to control the speed and implementing limit switches at each end.  If you set them up right (this involves using a couple of diodes), they will only disable the motor in one direction (the one that would break something) while still allowing it move in the opposite direction.
Or Current Sense Resistor circuit to keep it neat  ;D
Title: Re: Very slow Linear Actuator
Post by: meidou on November 21, 2011, 01:21:45 PM
Hey Guys
Thanks for the prompt replies.From them I have surmised the following:
1.Usage of a stepper motor requires alot of circuitry but at the same time the stepper motor offers a great degree of precision which I require for my experiments.
2.I have been searching and i have found 2 DIYs in which the creators were able to modulate their stepper motor:
http://www.instructables.com/id/Easy-To-Build-Stepper-Controller-from-a-Recycled-M/?ALLSTEPS (http://www.instructables.com/id/Easy-To-Build-Stepper-Controller-from-a-Recycled-M/?ALLSTEPS)
http://channel9.msdn.com/coding4fun/articles/Computer-Controlled-Stepper-Motor. (http://channel9.msdn.com/coding4fun/articles/Computer-Controlled-Stepper-Motor.)

What I would like to know if IYO if the above methods can be applicable in modulating my linear actuator.

I have to apologize if the above questions are a bit stupid but I am a Biologist and my knowledge of all things electronics has been based on 1st year Physics.

Thanks again
Meidou
Title: Re: Very slow Linear Actuator
Post by: newInRobotics on November 21, 2011, 03:22:12 PM
1.Usage of a stepper motor requires alot of circuitry but at the same time the stepper motor offers a great degree of precision which I require for my experiments.
Well, stepper requires twice as much circuitry as normal motor, that is 2 H-bridges. Now, are You sure You need that much precision? If You have a threaded rod (which, I gather, You want to go for) with threads being 1mm apart from each other, then with one rotation of the motor a syringe would be depressed 1mm. If You need to be precise as 1/64mm (or more), then go for stepper.

What I would like to know if IYO if the above methods can be applicable in modulating my linear actuator.
What do You mean by modulating my linear actuator?
Title: Re: Very slow Linear Actuator
Post by: Soeren on November 21, 2011, 06:29:00 PM
Hi,

I am attempting to construct a syringe pump using a linear actuator similar to the following
[...]
The main reasons I am attempting to make the  linear actuator rater than buying one is cost and the ones I have seen are too fast for my needs.I need the arm to move at a rate of ~500um/sec.
As already mentioned, a threaded rod (in a stainless steel alloy that agrees with auto-claving if that's applicable) and driven from a geared down DC motor would be much better than a stepper.
A rod with 1mm pitch should then turn at 30RPM, to get the wanted 500µm/s.
With a worm gear in between the motor and the rod, a worm wheel with 60 to 100 teeth will get it into a range of 1,800RPM to 3,000RPM, which agrees well with a lot of small (and dirt cheap) DC motors.

A 10 segment shaft encoder on the motor will, with a gearing of 100:1, give you 1000 pulses each shaft revolution, which, in an ideal world, would mean a pulse each µm, but a cheap threaded rod will likely not be that perfect. Still, compared to how cheap it can be made, it's extremely good and in any case, can be made with more precision than you likely need (you didn't add the acceptable tolerances though).

For the best precision, you should run it one way only (until done with whatever dosing job at hand), as going back and forth, the backlash will introduce some (small) errors.

A linear actuator is build on the same principles, just with less gear reduction.


Now here are my questions:
1.Do you guys know where I can find/purchase a motor that has range of 1-10RPMs.
2.I am attempting to use a steeper motor from an old scanner, will this fulfil the requirements?
3.To control the speed of the motor would a voltage regulator be effective?
1. I'd prefer a worm gear (as this is self locking) and a cheap DC motor. Any irregularities in motor poling will more or less vanish, as each complete turn nullifies it, so only the last uncomplete revolution may cause (extremely small) offsets and this divides with the gear factor. A stepper set up for direct rotation of the shaft will be more irregular.
Not sure if you just want a constant feed speed in a single run.

2. Don't. Other people already told you of the added complexity and depending on how you wanna use it, be aware that steppers needs ramping up and down of the speed, to keep it from loosing steps.Further, you'd need to limit current, as a motor used on a 12V system may in reality be eg. a 5V motor using a current limit method to make it step faster.

While I don't know your abilities in programming and tuning of control loops, I know you'll get better result from a DC motor, with the least amount of needed experience in mechanics, electronics and programming.

3. To control the speed of a DC motor, Pulse Width Modulation (PWM) is far superior.
To control the speed of a stepper, you just clock the changes at the needed rate (with a careful eye on the dynamics of the motor).
Since you only need to change the direction of the motor once in a single syringe emptying, you don't need an H-bridge for a DC motor, as the PWM can be made with a single transistor and a manual switch (or a relay) can change the direction.
Title: Re: Very slow Linear Actuator
Post by: meidou on November 25, 2011, 09:41:46 PM
Thanks alot guys.I have decided to use the threaded rod approach.