Servo motor question

First off, I have no experience in electronics whatsoever.

I am looking to build something in the next few months that will require clock-like movement with gears and such, and I was thinking of using a 1RPM servo motor to operate the entire piece. It will be relatively small.

My questions are:

1- How consistent are servos i.e. Do they keep constant speed/time?
2- How much torque do these motors have? It will be operating several different gears simultaneously and I don't want it to slow down during the times where it has to move multiple gears at once.
3- Can you power servos straight from the wall? I plan to have it wall powered, not battery powered.
4- Where can I get a quality 1RPM servo from?

Thanks to any who can help.


-The Truth is Out There-

Bump. Really need some details on this.


-The Truth is Out There-

easier/cheaper to use a clock movment

I don't think a servo is the right answer for what you are planning. There are transformers that willl allow you to regulate down to 6 volts, but servo use a pulse signal to operate them.

i worked at an electric motor factory a while back and some of the servo motors we made was used on cnc machines. to operate at a consistent speed they needed to go up to 9000 rpm

I originally considered getting a clock kit from hobby lobby or Michael's, but I seriously don't think they will be strong enough to move so many parts.


-The Truth is Out There-

You can check around on eBay for low RPM motors. I was looking around for them as well for a motion lamp project. Found some that run at 1-2 RPM, both AC and DC motors available. Not sure if they'll have the torque you need, but you can check.

msisaac said:

You can check around on eBay for low RPM motors. I was looking around for them as well for a motion lamp project. Found some that run at 1-2 RPM, both AC and DC motors available. Not sure if they'll have the torque you need, but you can check.

Click to expand...

good point. i salvaged a motor out of a halogen heater the one that turns it left and right. but im not sure on the rpm consistency.

If you need positional accuracy and feedback, I can't help you. I don't know of any actual "servo" motors that meet your criteria, but check this out:

https://www.sparkfun.com/products/12219

Well I say servo because I figured it would be small and easy to work with. I'm open to any suggestions.


-The Truth is Out There-

without having any specifications on what you are doing... what is it?, how big, how long is it going to be running, torque requirements, how precisely this thing need to run a 1 hz, etc...


assuming you mean RC servos...

1- How consistent are servos i.e. Do they keep constant speed/time? servos are generally for positioning something not driving around and around, they have three input wires, two for power and one for a position input. the position input will work with the servo's internal controls to make the servo turn to a specific place in it's 180 degree turning range and stop there. but you will need to supply the digital input electronics.

2- How much torque do these motors have? It will be operating several different gears simultaneously and I don't want it to slow down during the times where it has to move multiple gears at once. you can get them in different sizes so finding one with enough torque wouldn't be a problem, but it might take some experimentation

3- Can you power servos straight from the wall? I plan to have it wall powered, not battery powered. No, RC servos run off of 5 VCD, the 120 VAC out of the wall would need a wall wart (~$10) or something to drop the voltage down to a useable level.

4- Where can I get a quality 1RPM servo from? servos are cheap and plentiful, just search amazon or hobby town or any other hobby store website and take a look, there are ten million to choose from, servos are generally rated in degrees per second, so you are looking for a 6 degree per second servo, but servos only turn 180 degrees, they generally don't go around and around

servos like in RC car servos don't turn a full 360 degrees around, they only move back and fourth thriough the same 180 degrees. With a micro you can program one to teeter-totter back and forth fairly precisely in 30 or 60 second intervals to drive some other mechanism like a cam, but you can't go around and around. and you may not want to drive it like that indefinitely... You can hack a servo by removing the potentiometer and replaceing it with two resistors to make a simple high torque one way or the other motor, but you will have no speed control. or you can hack out everything but the motor and have a high torque little motor that you can control the speed and direction of, but you will have to supply the control electronics. also these motors are probably not rated to run continuously and may burn up after a while. of course you can always but one somewhat bigger than you need and since you are running at less of a load than it is rated for, it may last forever.

maintaining a constant speed on a small DC motor that can run continually would be difficult. you would need to run the motor at a higher speed and gear it down through a gear box to create the torque, and you would need some kind of speed feedback, digital encoder (?) to adjust the speed as the load changes.

if you decide to run the thing off of 120VAC, you might steal a motor out of a microvave turntable (fairly slow and torquey) and gear it down as necessary... but the absolute speed may depend on the outlet voltage which may rise or dip depending on the neighborhood circuit loading, it may speed up during the day when everyone is at work, and slow down at night when everyone is at home cooking and doing laundry and the grid dips... BUT it may not be rated for continuious use and burn up after a few days. right now I am looking at the oscillating fan on my desk and I think that that oscillator motor might be a possiblilty.

but if you can't kluge in somekind of a fan-oscillator motor or hack a servo, if you have a fairly unlimited budget, what you probably need is a STEPPER motor, but you will need a microprocessor (or a stepper motor control chip) to control it. Driving a stepper motor with a micro lets you control exactly how fast the thing turns, and often they come with a gear box attached so they put out plenty of torque. thiese are the motors they make CNC machines out of so you can get very precise control , but $$$.


My final suggestion would be look into doing a two resistor servo hack on an oversized servo to get rotation one direction at a fairly constant speed and playing with some gearing to get the speed right. you would need someone who can solder to do the hack (its actually pretty easy if you have the equipment) then if you know sombody with electronics experience, they should be able to whip up an input signal with a proto-board and a 555 timer chip and a few components. you may be able to use a cheap servo tester to drive it in the right direction with essentially no electronics experience necessary. if the tester plugs in to the wall, you won't need a wall wart... winwin. good luck

servo hacking - YouTube

http://www.instructables.com/id/Drive-Servos-with-a-555-timer-IC/

Amazon.com: servo tester

also servos can be pretty noisy... you may need some sound dampening, or go with the fan oscillator.

this is the fan I'm talking about, haha they have them at walmart.
http://www.amazon.com/Lasko-4916-De...d=1400520459&sr=8-6&keywords=lasko+table++fan

if you are really adament on not a clock. you could try to build a digitally controlled motor with a position sensor or a digital setting circle and program it to 6 deg/sec. other option is a rpm controlled motor with a worm gear for torque and then just compute your transmission ratio to find the right main gear

Ok, sooooo... It looks like servos are out of the question. The prop I'm working on is a clock, but I was wanting it to run off of power so you wouldnt have to keep winding it up like on a grandfather clock. The design I'm working with is having to be custom fabricated, so finding a premade clock system that fit the bill seemed impossible. Maybe I just need to get a clock kit and attach it to this other design somehow.

EDIT:
What about using an AC powered Synchron motor? They are relatively cheap on eBay. I have all of the gears and such designed out, and they all run off of the mechanism that moves the second 'hand,' so I only need one powered gear to run them all.

-The Truth is Out There-

Steppers.

Steppers have the positional accuracy, and they are generally hardier; steppers are used in CNC machines, for instance. And there are some pretty good off-the-shelf stepper controllers out there. If I had to throw one together quick I'd get a stepper shield, a beefy stepper motor from a surplus electronics store (or pulled out of a printer), and a real-time clock. Once you know the degrees of the stepper, you can figure out how many steps per second, and correct it back to the clock by counting steps at intervals.

Just to help out a little.

Servos. Based on what little I know of servos here are the problems I see with this option. Someone please correct me if I am wrong.
1. Servo position is controlled by sending it short pulses. The length of the pulse determines what angle the servo should be at. So you will need some sort of circuit to send it the correct pulses.
2. You will need to modify the electronics inside the servo to trick it so it will not know it reached the correct angle and keep spinning.
3. Some (most? all? none?) servos will have something on one of the gears to keep the servo from turning too far. You will have to remove that.

Steppers. Steppers also use pulses to move, but the length of the pulse does not determine how far it moves. One pulse will turn it whatever angle is it is designed for. However, you have multiple inputs that have to get pulses in the correct order to make the servo turn in the direction you want. So again, you will need to design a circuit to send the correct pulses to the stepper.

It seems to me your best bet is either a clock kit or a fixed rpm motor with some additional gears to get the rpm you need.

an AC powered Synchronous motor like out of a microwave turntable or an oscillating fan looks like your best bet. just make sure (if possible) that it is rated for continuous use and not intermittant use. i don't have tons of experience with these except for replacing the one in my microwave so it might not be that big of a deal. the cases seem to be fairly regular, you just need to pay attention to the output shaft shape and length when ordering, and don't hard gear everything together, you may want some kind of a friction clutch somewhere so you can break the hands free for setting / adjusting.

I am just throwing it out there but why not just get a standard battery powered electric clock movement and hard wire in an ac adapter. That way you know that it is made for exactly what you are trying to do. (a couple random links)

Atomic Clock Movements | Klockit
Amazon.com: Seven Star 1.5V, 3V, 4.5V, 6V, 7.5V, 9V, 12V DC, 500mA Regulated Universal AC to DC Converter with Multiple Connector Ends: Electronics

phez said:

I am just throwing it out there but why not just get a standard battery powered electric clock movement and hard wire in an ac adapter. That way you know that it is made for exactly what you are trying to do. (a couple random links)

Atomic Clock Movements | Klockit
Amazon.com: Seven Star 1.5V, 3V, 4.5V, 6V, 7.5V, 9V, 12V DC, 500mA Regulated Universal AC to DC Converter with Multiple Connector Ends: Electronics

Click to expand...

It's hard to explain without giving too much away, but the motor would have to move 11 gears/parts at once. Hence the torque requirement. I might have to scale this project waaaay down in order to to use a regular clock kit and modify it. My fear was having it large like I wanted it but the motor not being able to move the gears/parts effectively and within correct time.


-The Truth is Out There-

NormanF said:

Just to help out a little.

Servos. Based on what little I know of servos here are the problems I see with this option. Someone please correct me if I am wrong.
1. Servo position is controlled by sending it short pulses. The length of the pulse determines what angle the servo should be at. So you will need some sort of circuit to send it the correct pulses.
2. You will need to modify the electronics inside the servo to trick it so it will not know it reached the correct angle and keep spinning.
3. Some (most? all? none?) servos will have something on one of the gears to keep the servo from turning too far. You will have to remove that.

Steppers. Steppers also use pulses to move, but the length of the pulse does not determine how far it moves. One pulse will turn it whatever angle is it is designed for. However, you have multiple inputs that have to get pulses in the correct order to make the servo turn in the direction you want. So again, you will need to design a circuit to send the correct pulses to the stepper.

It seems to me your best bet is either a clock kit or a fixed rpm motor with some additional gears to get the rpm you need.

Click to expand...

Agree completely with the above comments.

Here's the deal; in older electric clocks, a synchro motor is used that is phase-locked to the cycles of the wall voltage driving it (60 Hz in the States). Which is making the presumption that the grid is staying pretty close to in tune. The next step up is an internal crystal -- which has a specific vibrational frequency -- which is then carved down quite a bit by electronic dividers.

That's basically what I was suggesting; a digital real-time clock is such a crystal set (with associated electronics all in one handy chip). Other alternative is a less accurate oscillator, and getting regular time checks over the air.

Servos are a poor choice as direct drive because they are designed for positional accuracy over a limited angle. The only way to get a controlled slew velocity with a servo is by pulsing it; going forward a few degrees, pausing, going forward another few degrees, etc. That's how I've been doing timed movements with servos on my own projects. When you strip out the position sensor (the part that makes it a "servo" -motor), it becomes nothing but a gear motor. Or to put it the other way, a servo is just a gear motor glued to a potentiometer, with a little bit of circuitry to compare where it thinks it is (the potentiometer output) with where the control signal says it should be (a pulse-width modulated waveform).

So, yeah, the advantage is that a stepper (which are also usually gear motors) is a lot more like a short stack of solenoids. Each time you activate one set of windings on the stepper, it snaps -- just like a solenoid -- to that winding set. They typically have 2, 4, or 6 sets, and by sending power to one after the other you make the thing rotate in a complete circle, around and around. But it always knows EXACTLY how many degrees it has moved every time the controller sends a new pulse, meaning the only mistakes in positional accuracy possible are gear slippage (or a load with enough torque to keep the rotor from turning as requested).



Unless you can find a heavy-duty syncho or clock mechanism (like the guts of an old office clock?) that can handle your mechanical load, the most accurate option is stepper, with stepper driver, and a source of accurate time signal, and the electronics between the parts to coordinate one to the other. Which sounds complicated, but in these post-Arduino days, really isn't. SparkFun, as one for-instance, has real-time-clock breakouts and stepper driver breakouts and I think both are even available in Arduino shields, along with libraries of the necessary software.

And a last thing here; once you are in software domain, you gain flexibility. As just one example, once you are in software control, it is easy to set to what time you want, or even jump temporarily to some new time setting to see if you like the way the hands look, or to take a picture, or to put it in an easier position to complete painting the prop...