R/C servo stepper motor conversion

[DIYASTRO]

In previous threads I have shown that I favour Servo units with PWM control to the small DC motors.

In some applications such as focusing when the need is critical the very slight over runby the motor is annoying.

I know that some DC control units such as those made by Shoestring and Hitecastro solve this problem to a certain extent but require hook up to PC. I wanted an independent control so worked out how to convert servo to stepper control without loosing too much of the amazing torque that they produce.

Posting some pics of process just now and will type up comments for them a little latter although I hope its self explanatory!

Hope others find it useful.

Boyd

[DIYASTRO]

Second part! To show the servo being used for Crayford focus operation.

Boyd

[kalasinman]

Nice work, but have to ask, if you're irritated by lack of linearity, what about the lash of all those gears? Why not use a DC RA drive from an EQ mount which would have built in op-encoder and planetary gear set?

[DIYASTRO]

Thank you for your reply Kalasinman.

I have, and still do use the motors and gear arrangement that you mention if the weight and mass are not an issue.

The backlash from the metal geared servos is almost nill. They mesh very tight and provide quite a lot of torque especially for the size and mass of the unit. Although, I have used the converted servo for a Crayford focus arrangement here, I also make my own filter wheels and camera rotators etc which will benefit from this type of servo conversion. ( see other threads by me in the DIY section, or search my posts to see pics etc)

Boyd

[skybadger]

Doesn't this end in the position of having a nice high torque motor drive but without any positional knowledge?. What is now required is an encoder...

Mike

[Ajohn]

Why is a position encoder needed ?

John

-

[DIYASTRO]

Skybadger,

Thanks for your reply.

As yet I have not had a need for encoder, I simply use the visual video output or camera image on pc screen for focus. However, you will see from the fabrication in the last photo that the aluminium plate that supports the servo and shaft gear arrangement to the Crayford has sufficient space for encoder if I feel that I need one. The use of the drive shaft or the position of encoder on the 72 tooth gear would suffice.

In this particular application it was to enable a very fine and smooth focus ability and with the driver pcb shown in the pics I have a range of 10 speeds. This in tests gives a movement of the Crayford focus a travel of 22.2mm / min or .37mm / sec at speed 9 (High speed) At speed 0 (slowest speed) it equates to .08mm / min or .001333mm / sec. Of course it has a range of speeds in between these which is more than adequate for the purpose intended.

I will be working on some encoder applications when I fit some of the modded servos to filter wheels and camera rotators that I have been building.

Thanks for your interest.

Boyd

[kalasinman]

I only brought up the alternative as I have a mount that uses a motor-planetary as I described which BTW also has an encoder. It ises a pulley and cogged belt system to drive the worm gears. It had occured to me that as my focusser (a GSO) has knobs retained by set screws to a shaft, that one knob mught be replaced with one of these pulleys and the belt drive used as well, with only bending up of a bracket and a control board remaining to have a functioning unit. I believe these motor units retail for $75. Expensive?

[Ajohn]

That's neat piece of work Boyd   I had my tongue in my cheek when I asked the question. Positional feedback isn't needed for focusing.

There are various way of using them. The normal type are driven by a 50hz square wave form and the  arms will usually move through about 120 degrees but I suspect are really meant to move over 90 degrees. The mark space ratio controls the position the arm takes up. The digital types have a dead band of 2 uSec so for say a 1 mSec total variation in the mark space ratio they have " 500 steps ". Might not sound useful but they can apply a lot of torque to the end of the arm that is usually fitted to them.

The other type that can be bought as standard uses the same drive waveform but the mark space ratio controls rotational speed, There are details about on the web concerning modifying the ordinary ones to work this way.

I believe programmable ones are available as well.

The above operate in servo mode - they will apply power to keep in position or maintain speed.

Some people just take them apart and disconnect the motor leads and drive it directly. The positional information is obtained with a variable resistor with end stops so that needs taking care of if rotational motion is needed.

There are various type of construction. Brushless motors, currently rather expensive and then combination of metal gears, plastic gears and ball races.  The most powerful ones I have seen without looking hard are just short of 20kg/cm but will  go over that with a 6v drive. They are usually spec'd at 4.8v and 6v. The digital types are usually rated to work in sub zero conditions.

John

-

[DIYASTRO]

Ajohn,

Very many thanks for your post.

Yes....interesting little things arn't they servo's! Its the torque that they develop that is so useful for their size.

I am currently looking at rehousing the gears into an even smaller unit and also sourcing even smaller stepper motors to modify the micro servo's.

As you say, the design and manufacture of these metal geared digital types have come a long way in the last few years. Hope to be able to incorporate modded versions into some of my other home made astro stuff.

My main purpose with such mods is to avoid pc control as much as possible, most of my scopes and multi cameras are analogue operated so as to avoid the need of pc.

Currently building a quad imaging rig which will have this servo type for the focus operations without the need for encoders.

Boyd

[Ajohn]

I suspect they have more to offer than steppers but need a different approach. What started me wondering was this

http://www.indilib.org/support/tutorials/143-diy-remote-focuser-with-indi.html

Connected directly to an arduino pin as far as control goes.

 Then I started think why not use them in there normal form via the arm - via a push rod, or put a pulley on the end etc. A focus for instance doesn't need a lot of travel and something of the order of 1/500 accuracy on a turn of a main focus knob is likely to good enough for any scope -   i would have thought so don't quote me. 

I like analogue too but things have moved on PC wise. It looks like the INDI set up can be driven with a netbook and a 'pi at the telescope end with an arduino to help focusing. Some one else has gone further. The whole lot is presented as a web page so can be driven from a tablet and still use INDI items via a tablet utility. That included plate solving. It's not fast but........ There are some links in a painless remote control post I started on here  Since that one has also appeared on the INDI pages too. I wasn't knocking INDI just mentioning an alternative. Maybe just maybe some one will look at presenting INDI as web page opening the whole thing up.

If some one wants to play with model servo's the testers are cheap and just need a 4 AA cell holder for power. I am not sure if  these are stable enough to form the basis of an electric focus. That's the only question really. The same question has to be asked about the servo's too.

Some encoders are cheap, ebay is full of them but is suspect when they say things like 24 pulse they are really 2 channels with 12 pulses per rev each in quadrature so that direction can be sensed. I believe people have printed their own encoder discs as well.

John

-

[Ajohn]

The site that seems to have started extreme servo hacks linked to here may have changed it's name and gone completely commercial but even so the bare bones of a linear actuator are shown based round a model servo

http://www.instructables.com/community/Hack-your-Servo-V100-Make-a-powerful-linear-actu/

The extreme hack mods seems to involve using open servo code in a replacement board and even a magnetic rotary encoder. They claim 10 bit positional precision with the encoder fitted  but the chips used can have max errors that aren't as good as that would indicate, worst case is probably +/- 1 degree.

John

-

[DIYASTRO]

Ajohn,

Had a good look at some of the references you give, interesting stuff.

I think I have finally tracked down some very small micro stepper motors that I am sure will fit the micro servo's so have ordered a few, from China so will take a while to arrive.

Had a couple of hours today and took a break from packing, ( moving house) but spent some time fitting some of the modded servo's described in this thread to my 'new multi lens camera' will try and post some pics in a new thread this evening. Although no where near finished it gives a decent impression of what I am trying to make.

Boyd