Servo question

[Red Squirrel]

I've never used servos before but I think they might be just what I need for my hvac project for my server room.  Basically I need to create a bunch of dampers to control air flow supply and exhaust  to take/exhaust air from/to different locations based on various environmental conditions.   They do make electrical controlled dampers but they're very expensive, like $100+ a pop and I need like 8 for my project.   So from my understanding of servos  they move to a specific position based on the duty cycle of a wave form sent to it, correct?  So a 555 timer could be set to move a servo to a certain position, with two 555 timers and a switching circuit I could toggle two different positions.  Does this seem right?  Guessing frequency matters as well so I need to ensure the servo and 555 timer I use work on same frequency?

So I'm thinking a servo that moves wood flap inside a box (need to think up of the actual design for this) would make a good damper, 90 degree off, 0 degree down (or w/e depending how it's setup). Now question is, can a servo be used in a 24/7 fashion like this?  Or are they designed to be put in a certain position then have their power cut off until they need to move again?  Also, are they hard to turn manually?   Ex: will air hitting the flap be able to force it to move around?

For servo I'd probably be looking at something like this:  http://www.amazon.ca/Hitec-31311S-HS-311-Standard-Universal/dp/B0006O3WVE/ref=sr_1_1?ie=UTF8&qid=1423394663&sr=8-1&keywords=servo

[IanB]

The servo you linked to is an RC servo for model applications, not an industrial servo. An RC servo needs constant power to hold its position, and is unlikely to be designed for 24/7 operation (how many RC models run for hours at a time?).

If you want a servo to hold position with power off between commands, it would most likely need to have a worm drive or high ratio gearbox. Possibly such servos are made for RC applications--you would have to research it.

[SL4P]

Specs:  Keep this in mind when designing your mechanicals...
Torque oz./in. (4.8V/6.0V): 42 / 51, Torque kg./cm. (4.8V/6.0V): 3.0 / 3.7

[SeanB]

They are not very high power devices, and will fail if left powered, and will tend to blow around in the air.

Alternatives are small stepper motors with a built in gearbox, like those used as flap motors in split aircons, or other small reduction box motors with a split capacitor winding that you can control as to direction, and use a hard stop to block it and actuate for a short period so it reaches the limit then it runs stalled till power off. Those will work as the friction in the gear chain keeps them in position.

[paulie]

RC servos do not "fail if left powered". There are some issues with gears stripping if outside forces are great but that is unlikely unless these dampers are used in a hurricane.

Also note that most will hold position quite well when powered down. Specially that particular Hitec model which has considerable  resistance due to the gear ratio. Again in hurricane force winds this may not hold true. Of course in this application there is no point in powering down.

These days cheap chinese servos not only cost 1/10th as much as those old domestic units but often outperform and outlast.

Last but not least it's worth mentioning that a single MCU can replace many of these 555 devices for a fraction of the cost. MCU-phobia will probably prevail here and prevent the simplest and most cost effective solution.

[Red Squirrel]

I have not fully decided on how to do the electronic side of things so I'm open to ideas.  What's a MCU?

Also maybe servos is not the best way of doing this, is there a better way?  I was even thinking some kind of spring mechanism, and electro magnet.   Perhaps even modifying a manual damper.

[IanB]

An MCU is a microcontroller unit, or just a microcontroller for short. It is a programmable device that can read various inputs, like position switches, or temperatures, and then manipulate various outputs, like on/off relays or servo positions.

A servo is nothing more than a motor that can move to a given position on command. It is probably the best way to move an adjustable damper.

[Red Squirrel]

Oh ok, I thought it was something else.  Yeah this will have a micro in it.  I was thinking of having the 555's mostly so that the micro just needs to send an on or off signal.  But perhaps that's adding more complexity as I can just do the PWM with the micro controller directly as long as I can do it at whatever frequency the servo requires. Will probably use Arduino or RPI.  The micro will have temp and other sensors which will be the data it uses to determine which damper is open or closed.

So really if I do go the servo route I'm fine leaving it powered on then?  Idealy I want to find the highest torque one and experiment, as if it's solid enough without being powered then I can only power on when I need to do an adjustment.   I may also experiment with electro magnets and springs.  Not sure where I would find springs though, I guess I can probably make my own if I find some steel wire somewhere.

[TerminalJack505]

If you use an RC servo then there are a couple of gotchas that you should be aware of.

First, if you plan on switching between just two positions, without specifying any of the intermediate positions, then beware that the servo will move fairly quickly between the two different positions.  The rate that it will move will basically be out of your control so you can wind up breaking things or just simply putting more wear and tear on things then necessary.

The second thing to be aware of is that RC servos have no feedback of where they are.  This can cause a problem similar to the one above when you power the servo on and set its position.  If the servo's position is far away from where you command it to be after power-up then it will move rapidly and uncontrolled.

You can avoid both of these particular issues by using a MCU.  The first problem is avoided by having the MCU specify some of the intermediate positions so that the speed at which the servo moves is under your control.  The second problem can be avoided by having the MCU remember (by using non-volatile memory) where it last positioned the servo and use this position at power-up.

[IanB]

So really if I do go the servo route I'm fine leaving it powered on then?  Idealy I want to find the highest torque one and experiment, as if it's solid enough without being powered then I can only power on when I need to do an adjustment.   I may also experiment with electro magnets and springs.  Not sure where I would find springs though, I guess I can probably make my own if I find some steel wire somewhere.

A high torque servo will have high ratio gearing, which means it will hold its position quite well when powered off. If you only move the damper infrequently then it would be sensible to leave the server powered off between moves. Since there is basically no load on an air flow damper there is not much need to hold the position under power.

[Red Squirrel]

Hmmm yeah if it takes off too fast with a heavy load it could probably put stress on the gears.  I could probably get away with using rigid foam or something lightweight for the damper plate though.  I'm guessing if I find high torque ones then they'll be slower as well and that's probably what I want.  Found this one too:

http://www.amazon.ca/33322S-HS-322HD-Standard-Deluxe-Karbonite/dp/B0006O3XEA/ref=sr_1_22?ie=UTF8&qid=1423452331&sr=8-22&keywords=servo

I'll have to do more searching when I want to start this project, but might order one just to build a prototype.

[SL4P]

I made a board some time ago with an low-end PIC on board... ServoStick
Used at the time to pull the choke in / out for small industrial engines.

• single wire input (open / max +18v), or closed to 0V)
• output moves between two preset servo positions.

The two end positions were programmable in situ (with a link) - and saved in EEPROM.
Small, easy and effective.
I'll look for the design files and code - and send if needed.

[electr_peter]

The second thing to be aware of is that RC servos have no feedback of where they are.  This can cause a problem similar to the one above when you power the servo on and set its position.  If the servo's position is far away from where you command it to be after power-up then it will move rapidly and uncontrolled.

Power-up for cheaper (common) type of RC servo always provides a jerk movement irrespective of control signal. Solution to this problem - better servos with smart control.
Common RC servos do not have feedback output for MCU - additional wires can be wired to provide (poor) feedback to MCU ADC. Smart servo will provide feedback though.

[w2aew]

As others have stated, an RC style servo might not be the best solution, but could be made to work.  And yes, a 555 could be used to provide the control signal for the servo.  A couple of months ago, a friend asked me if I could show him how to control one of these servos with an analog voltage - so I did a video on this.

I'm not saying that this is how you should do it, but simply providing this video link (below) to help you learn bit more about how these servos work:

[SL4P]

I'll look for the design files and code - and send if needed.

Code only via PM if you're still interested.

[Red Squirrel]

Thanks for the info, I think I'm good as far as code goes, as this will be built from scratch anyway so easier to just incorporate my own code but thanks anyway.

I just typed "servo" in Amazon to see what comes up, so I'm open to ideas of RC may not be the best kind, is there other types of servos to look into?  Don't want to break the bank though. I'll probably need like around 8 or so for this project.  Feedback would be good to have, but not required.   Nice thing with feedback is I could ensure that it moved and if not it could trigger some kind of alarm to say that a damper failed to move into position.  Could use "door contact" type switches for that too though.

[SL4P]

  Could use "door contact" type switches for that too though.

Keep in mind any contamination of the sensor. 
Definitely a good idea to add a limit sensor if you can't trust the servo, but in many situations you could need an 'industrial grade' IP-rated type switch if the environment is really nasty and cycle rate relatively low, or an IR or proximity sensor for high-rate / less-fragile environments.