Let's talk about servos - Whatever is on your mind

[DanWard]

Thanks,

So then the gimbles on my radio (DX7i) has potentiometers and the transmiiter has an analog to digital converter that converts the potentiomiter analog signal to digital values based on the resolution (number of bits) in the A to D, So as you say it is determined by the transmitter.

I was obviously messed up. I though inside an analog servo the position feedback was from a pot. The receiver told the servo where to go in the form of an analog value and the servo moved until analog feedback matched analog setpoint.

So my next assumption was a digital servo eliminated the analog pot in the servo and replaced it with a position encoder that could be nothing more than pulses generator or an absolute position encoder.

I'm learning. Thanks,

[Turner]

Have a look at these:

http://www.futaba-rc.com/servos/digitalservos.pdf
http://www.rchelicopterfun.com/rc-servos.html

[AEAJR]

Turner, those are great links.

[DanWard]

Most excellent artical. I was way off base in my understanding.

Sounds like the PWM (pulse width modulation) frequency is 50hz (cycles per second) on analog servo and 300hz on digital.

This has left me feeling a bit less "in the dark".

Thanks,

[Turner]

Yeah, I learned a lot too. Good information.

[Voltron]

I'm glad I found this thread. I have a strange application for rc servos that I'm pretty sure that no one else on this forum has done yet. I've been monkeying around with RC stuff for over twenty years and this is a first for me.

I have a friend that lives in another state that does special effects makeup and costumes for low budget horror films. She emailed me today asking me if it was possible to use rc servos to control the movement of six snakes on a Medusa headpiece she is making for their next movie. Now there is a request that you don't get every day.

I'm thinking, sure, why not? The hitch is though, that she needs the servos to cycle back and forth without using a transmitter. She needs to be able to turn them on and have them cycle automatically to move the snakes without someone controlling them. That request takes me out of my comfort zone with what I know about servos.

Does anyone have any ideas how we can accomplish this?

I have to try to keep the cost at a minimum for her. Her and a group of friends self finance these movies and do some pretty cool stuff for what they have to work with. Although they do the movies for the fun of it, their last movie "Wet Kiss" was accepted by a distributor and has actually turned a small profit. It is also available from Netflix.

[Turner]

Maybe you should think in terms of continuous rotation.

There are servos available for continuous rotation or you could just use motors.

[Voltron]

Turner, thanks for the reply.

I thought about using small DC motors but I'm not sure if it would give us the effect of snakes writhing around. It might work with an eccentric or cam to get the desired motion. Motors would certainly be easier to control if I can make them run slow enough with enough torque. The speed and torque of a servo would be just right, but they would be harder to control without a transmitter.

[DanWard]

Hello,

Very informative thread. I have enjoyed and learned quite a bit. I now think I'd like to use digital servos on the plane that I am now building.

So the plane calls for 4 @ Hitec HS-65MG.

• I'm really not big on metal gears. Have yet to see the need. Maybe some day a bit tyo late I will say "should have used metal gears" as I strip a gear.
• So I will look up dimensions, weight and torque for this recommended servo and then look for a match in a digital servo.

Anyone have a recommended band for digital? I don't dive a lamborgini nor do I drive a wore out cheap junker. Just a good chevy or dodge truck sort of servo. Anyone willing to point me in a good direction?

And then on the subject of servo centering.

• Don't you want to have a 90 degree relationship between the servo arm and pushrod when at zero? The servo travels in an arc so if you start at 90 degree relationship the pushrod travels the same x and y distance when servo rotates in either direction. Right?
• My experience is no two sewrvos match one another and seldom if ever do the splines on the shaft allow 90 degree relationship to pushrod when everything is at zero (sub trim, trim, etc).
• So I do my best to get as close to 90 degress using splines.
• Then I use sub trim until center line of servo arm is close to 90 degrees to the pushrod.
• Then I adjust pushrod length until the flight surface is at zero as well.

I think this is right but really nvever had it confirmed. Would appreciate the feedback.

[rcers]

Hitec is an excellent maker for Servos and they make a digital version of the HS 65 MG
http://servocity.com/html/hs-5065mg_servo.html

You are dong well on setup - you do want the servo arms to be close to 90 degrees. It is not super critical it is exact but that helps to get the same throw each direction.

Mike

[Turner]

Sometimes you actually want differential movement of a control surface so you will intentionally stray from the 90˚ setup. This can be used to build in differential on ailerons whether using one or two servos. Flaps too might deviate from the 90˚ norm.

[DanWard]

Recommended servo is analog HS-65HB at about $20 each.


Next step up is still analog HS-65MG at about $30 each.


The digital equivalent is HS-5065MG at about $40 each. And I have seen some other brands that even cost more.


Maybe champaign taste on a beer budget but at twice the cost of the recommended...well...


So it doesn't hurt to ask. Is there a more economical solution without them being junk?

[DanWard]

My local hobby shop gave me a decent price on HS-5065MG so I guess I'll give them a try. I also purchased an ICE50 ESC and seperate CC BEC. I read that digital servos draw more current plus the seperate BEC provides some seperation should the ESC fail.

[dahawk]

I'd say you've got top of the line gear ! and using a ubec is always a good thing. Frankly, for the type of craft I fly, I'm okay with the cheap Htx servos from HK. Have not failed me yet but if they do, it's not like I'd lose a $500-$1000 ARF.

Nice buy !

-Hawk

[kyleservicetech]

Originally Posted by Voltron

I'm glad I found this thread. I have a strange application for rc servos that I'm pretty sure that no one else on this forum has done yet. I've been monkeying around with RC stuff for over twenty years and this is a first for me.

I have a friend that lives in another state that does special effects makeup and costumes for low budget horror films. She emailed me today asking me if it was possible to use rc servos to control the movement of six snakes on a Medusa headpiece she is making for their next movie. Now there is a request that you don't get every day.

I'm thinking, sure, why not? The hitch is though, that she needs the servos to cycle back and forth without using a transmitter. She needs to be able to turn them on and have them cycle automatically to move the snakes without someone controlling them. That request takes me out of my comfort zone with what I know about servos.

Does anyone have any ideas how we can accomplish this?

I have to try to keep the cost at a minimum for her. Her and a group of friends self finance these movies and do some pretty cool stuff for what they have to work with. Although they do the movies for the fun of it, their last movie "Wet Kiss" was accepted by a distributor and has actually turned a small profit. It is also available from Netflix.

Yeah, I've done this many times. Its all done with one of those little Microchip PicChips. This chip can be programmed to directly operate a servo with only a 5 volt battery, the PicChip ESC driver, and the servo. The whole thing is about the size of a dime. The project was built on a part of one of those Radio Shack perf boards. A circuit board could easily be laid out for this, but they would cost about $60, for a dozen or more boards. The boards will have to be sawed out with a little hack saw.

What would be needed is exactly how they want the servo to move. A lot of things can be done with these PicChips. These PicChips can be programmed and reprogrammed some 100,000 times for different effects. I can build up and program a few of them, but for this application, it might be a good idea to simply provide how to program them. I've got basic programming notes that show how its done, and how to make the servo do different manuvers.

What is needed is one of those MicroChip "Pickit3" programmers for about $45. The software is free from Microchip. The PicChip (P12F629) is about $1.00 each from places such as www.digikey.com. The wiring is extremely simple. This chip has 8 pins, with three parts installed. The chip, a resistor, and a capacitor. And, the programming could be figured out by a knowledgeable 14 year old, based on what I've done so far.

My last project was a PicChip/Servo system that used the servo to push the shutter of a digital camera every 60 seconds.

Take a look at the attache JPG. This PicChip driver operates the servo to left position for 2 seconds, 35% position for 2 seconds, right position for 30 seconds, and back to left position. Far more steps can easily be made with these PicChip microcontrollers.

[MrTapper]

Originally Posted by AEAJR

A note on servo arms.

When you look at servo ratings they report ounce/inches of torque.

If you put a 1" servo arm on the servo, it should be able to deliver its rated ounces to that arm at 1" out from the hub of the servo.

If you put a longer arm on the servo, say a 2" arm, a 2" long servo horn ... then you expect less force avaliable at the end of that arm. This is due to the principal of leverage.

So a servo rated for 10 inch ounces can deliver that at 1" out from the servo hub. At 2" out is should deliver 5 inch ounces. So if you need 10 and you put on the long arms, then your servo better be rated for 20. Add to that our comments about inflated ratings and some room for safety you probably want a servo rated at 25 to 30.

Engineers, did I get it right?

Originally Posted by Stevephoon

AEAJR,

I don't think this is correct. It's the motor torque or force that servo’s list. That same amount of force is available everywhere on the control arm.

....

So, if you hook your control rod to the inner most holes of both the servo and control side and get say 1 in of deflection and if you move the control rod to the outermost hole on both and still get 1 in of deflection, the force available to move your control surface is still the same. One is not any better than the other as far as force goes. The difference is that you have more control rod movement back and forth.

Does this make sense? (I think this is correct!)

Steve

Hi all,

After reading through these couple of posts a few times, I was unsure of the general conclusion reached (Sorry if this is just restating what's already been said). I figured I'd make things clear to anybody that felt the same...

AEAJR was correct in nearly everything that was stated, but I think the source of confusion was "...a servo rated for 10 inch ounces can deliver that at 1" out from the servo hub. At 2" out it should deliver 5 inch ounces."

This implied that the Torque available at various lengths away from the servo changes, when it is actually the Force.

Thus, a servo rated at 10 inch-ounces will provide 10 ounces of Force at 1" away from the centre of the servo. With a 2" servo control arm, the servo will provide 5 ounces of Force.

This, I believe, is what Steve was initially trying to say ("It's the motor torque or force that servo’s list. That same amount of force is available everywhere on the control arm."), but the terms Force and Torque were reversed.

When Steve said...
"...if you hook your control rod to the inner most holes of both the servo and control side and get say 1 in of deflection and if you move the control rod to the outermost hole on both and still get 1 in of deflection, the force available to move your control surface is still the same."
...the terms force and torque were reversed again.

Finally, when attaching the pushrod to the innermost holes of both the servo arm and control horn, the servo has to apply the most Force to deflect the control surface. When attaching the pushrod to the outermost holes of both the servo arm and control horn, the servo applies less Force, but the same amount of Torque is used.

Theoretically, in both of these cases the control surface should deflect the same amount using the same amount of Torque since the pushrod is connected at equal distance from the servo centre and control surface.

This brings us to the topic of mechanical advantage:

If one were to connect the pushrod/linkage to the innermost hole of the servo and the outermost hole of the control horn, the control surface requires the least amount of Force to be deflected and the servo can apply the greatest amount of Force. This allows for very fine (reliable) control of the control surface, but a very small range of deflection. Minimal strain is applied to the servo.

If one were to connect the pushrod/linkage to the outermost hole of the servo and the innermost hole of the control horn, the control surface now requires the most amount of Force to be deflected, but the servo is applying the least amount of Force. This allows for very large deflections, but applies the greatest strain to the servo... a setup best used with a much stronger servo and/or slow moving aircraft that will not see high pressure on the control surfaces due to airspeed.

Ok, I think that should clear everything up for now (Aerospace Engineering studies finally paying off )

Sorry for the long post

Tapper

[Turner]

Torque is always measured at some specified distance from center. From a known servo, closer to center delivers greater force, further from center delivers less force. Range of motion is a different issue entirely and unrelated to torque.

http://en.wikipedia.org/wiki/Torque

[Rockin Robbins]

And regardless of where you measure the torque, whether 1" from the center of rotation, or 5" the torque measured will be the same.

Let us suppose that at 5" from the axle we measure 5 ounces of force. But our standard measurement is inch-ounces in this case. So our calculation is dividing both the force and distance by 5 gets us one inch-ounce of torque. That is called normalizing our measurement. In the metric system you would normalize the force to newtons.

So with a servo: torque is always the same. Force is directly proportional to the length of the moment arm.

[Stevephoon]

Originally Posted by MrTapper

Hi all,

...
Ok, I think that should clear everything up for now (Aerospace Engineering studies finally paying off )
...

Tapper

Trapper, That does help me... I knew that I was mixing up force and torque somehow...

Originally Posted by Rockin Robbins

...
So with a servo: torque is always the same. Force is directly proportional to the length of the moment arm.

This is what I had thought all along, but never stated it correctly....

Steve

[DanWard]

1. I'm using digital servos for the first time. My DX7s radio has a setting for frame rate which is normally 22mS. but it can be changed to 11mS but is only recommended if digital servos are used. I am wondering, since I have digital servos, what is this frame rate thingy all about. Seems to work fine at the 22mS setting on my bench.

2. The leads on my wing servo is to short so I bought extensions but worry about them coming apart. Is there any semi-permanent way to connect them? I won't take them apart again unless I crash otr something fails. I was wondering about hot glue. I've used it to glue servos into foam planes and yet when you need to it will come off the servo.

Anyone have any suggestions?

3. Then inside the fuselage I have connectors for when I take the wings off for transport. My fear is them coming apart in flight. Any suggestions? Do I have to buy some sort of clip thingy or is there a home brew version?

[kyleservicetech]

Originally Posted by DanWard

2. The leads on my wing servo is to short so I bought extensions but worry about them coming apart. Is there any semi-permanent way to connect them? I won't take them apart again unless I crash otr something fails. I was wondering about hot glue. I've used it to glue servos into foam planes and yet when you need to it will come off the servo.

If you've got a Harbor Freight store near by, you can purchase shrink tubing that will fit over the connectors. Hit the tubing with a heat gun, and it's permanent. But a very sharp knife can be used to slice off the shrink tubing for removal. Just be careful to not cut any wires. If you're planning to leave this connector extension in place for a year or three, it's best if the pins are gold plated.

A roll of 10 mm tubing works well.

http://www.harborfreight.com/10mm-x-...ing-98067.html

[DanWard]

Now why didn't i think of that? Most excellent idea. Light weight, affordable, and i already have what I need.

Excellent and thanks for the help.

Now to figure out what frame rate is.

[kyleservicetech]

Originally Posted by DanWard

Now why didn't i think of that? Most excellent idea. Light weight, affordable, and i already have what I need.

Excellent and thanks for the help.

Now to figure out what frame rate is.

OK, Frame Rate:
Our radio systems operate with a "Pulse" output from the receiver to the servos. This pulse output is a voltage that rises from zero to several volts (depending on receiver type) for a short period of time that ranges from 0.001 seconds to 0.002 seconds (one millisecond to two milliseconds) That would correspond to flipping a headlight switch on your car for 0.001 seconds. Hard to do, but with our electronic stuff we use, it is easy to "Flip" an electronic switch a million times per second, or a lot more.

Now, 1 millisecond pulse will drive your servo to its "counterclockwise" position, and 2 ms pulse will drive your servo to its clockwise position. But, there is also a "pause" between those pulses sent to your servos. That pause is the length of time that signal voltage is at zero volts. So, you could have a pulse signal of 1 millisecond "on" and 20 milliseconds "off", for a frame rate of 20 milliseconds. Or, a pulse signal of 1 millisecond on and 10 milliseconds off for a frame rate of 10 ms.

If you've ever seen an oscilloscope, this is what is used to directly display the voltage magnitude of that 1 ms pulse, its pulse width, and its frame rate. (I've got three of them, Tektronics 2216's and 2236. That 2236 will show pulses of one pulse every few seconds to pulses at the rate of 100,000,000 times per second.)

What's the difference? Not much, unless you've got speedy servos that can notice the difference.

[CNY_Dave]

For connectors and other items I want to secure but remove, I went to lowes/home depot and bought a roll of no-adhesive rubber electrical tape.

Sticks to stuff a little and itself quite well but removes easily and is reusable 5-10 times. A narrow strip wrapped around a connector and it won't come apart, but can be unwound quite easily.

I even wrap it around a clevis instead of using heat-shrink or fuel line.

[DanWard]

OK in my RC career (which was about 5 years up through 1993 and then I picked it up again a few years back) I have always used analog servos. But I'm building a nice 3DHS 47" SHP and I'm trying to do it right.

CC Ice 50amp ESC
CC 10amp UBEC set at 6 volts
Omege 130g motor
And while the plane called for Hitec HS-65 servos I was intrigued by digital servos so I purchased Hitec HS-5065.

I tested all servos with no load before installing in the plane and they all appear to work fine. I have also checked them now that I am having troubles with them with no load (after install) and they all work fine.

I also checked all flight surfaces before connecting to servos and nothing binds. I have also checked them now that I am having troubles by disconnecting from servo and moving by hand and everyone is smooth as glass. But I am getting ahead of myself.

Finally tonight I'm about finished with my build. Time to set subtrim and servo endpoints.

Much to my dismay every flight surface has a problem. The ailerons work some of the time but rudder and elevator almost none of the time. If I jerk the stick 0 to 100% ailerons will work most of the time but even this doesn't do much to help elevator and rudder.

It is most notable when you move a stick slowly. The surface just stops anywhere from 10% stick to 60% stick and doesn't move anymore even though you move stick to 100%. While the problem is repeatable the point at which the servo stops varies every time.

When the servos stops it is dead silent, no buzz, no hum, no nothing.

If you touch the elevator when servo stops it will move, sometimes to full travel, sometimes not and requires a second touch. When the rudder servo stops you can move the rudder a little further and it will stay where you put it until eventually it reaches full travel.

I have also noted they are bad about not returning to servo. If you hold the stick at 100% and release it it will go to servo most of the time. But if you return to neutral stick slowly they stop shy of zero almost every time.

Not knowing any better I would swear the servos are stalling.

I have used a Fluke volt meter to check UBEC and it is darn near 6 Vdc on the nose and rock solid.

I have tried moving all flight surfaces 0 to 100% in a jerking fashion at the same time (which is when servos work best) and I see no drop in UBEC voltage nor does it make servos perform worse.

I have gone back and studied the HS-65 and HS-5065 and while I knew they were identical torque and speed what I did not know previously is they are programmable and I don't have a programmer. Hitec assures me I should not need a programmer. But among the list of features available in an HS-5065 is something called overload protection. It says that from the factory it is turned off.

While I do not believe all four servos are overloaded the fact that when they stop they don't buzz or hum makes me think the motor stopped running as if it thinks it is where it should be or it thinks it was overloaded.

I am exasperated. Sometimes I'd swear a little black cloud follows me around. My bet is if I had analog servos I would be ready to fly.

Anyone here have any recommendations? Hitec says they will replace them. Man am I tempted to have them exchange them for HS-65. The performance advantage of digital is not speed or torque and if a disadvantage is what I am experiencing right now...arrrg

Anyone have any ideas?

In the mean time I am going to go back and read the previous links and try to figure out why I bought digitals if speed and torque are identical

PS - this is the same DX7s radio that flies 3 other planes just fine. Also moving servos from channel to channel no matter where servos and rudder are plugged in they always perform the worse.

PSS- I am having troubles holding flucke leades on receiver plus and minue pinds to measure BEC voltage but as near as I can tell my voltage is 5.95 with no servos being moved and 5.85 while wiggling all sticks at the same time. That is close enough to 6 volts to not be causing me troubles, right?