Oddball stepper question - auto-focuser

While this isn't a CNC question, exactly, it is related.

I am looking at building a PC controlled focuser for a telescope using a small stepper motor and pulleys. What has me scratching my head is how to tell the software where "home" is.

The scope is a Maksutov Cassegrain and its focused by turning a knob which moves the primary mirror. The knob rotates but it does not move in and out. In fact there are no moving parts to which I can attach any kind of switch to detect when the mirror is at the limits of travel.

So I'm trying to figure out how to "home" the focuser so that the software can reliably move to a position to repeat a previously found perfect focus.

The best would be hardware limit switches so the software could find its own travel limits during initial setup and do a "home" at each startup thereafter. However, I can't really think of a method of doing this..

Anyone have any ideas?

I would think using software limits would be the best option unless you want to take the scope apart and stick a limit/homing switch in there somehow.

Can you manually put the mirror in the same position at the start then switch the stepper drive on and tell it that it is homed straight away ? And possibly when you power it down leave it in the homed position so at restart you can do the same thing...

Also what software do you intend to control this with ? I would think if it is simply moving at a constant speed you could just use a 555 timer to generate step pulses to send to the stepper driver, though you would have to include some kind of brain so it knows when to stop. Seems overkill to hook up a PC and monitor to control this...but I guess a second hand PC is practically free these days.

Cheers.

Russell.

Hi Rhodan

To focus the scope the priamry to corrector plate reflector distance is being changed. I guess your problem is that the optical system is 'sealed' and this precludes the mounting of mechanical switches to home the separation of the mirrors. What happens when the focus control is turned to the extreems of its travel? Is the focus continuous but changes sign ie.e. is some form of cam operated mechanism? If so then it might be worth investigating using a colimated beam of light bounced off the primary mirror from a slit source mounted on the corrector frame using a photo transitor or diode on the oposite side arranged to be triggered whrn the distance is at its greatest. You will have to look at the geometary involved as it might even be possible to bounce the beam off the corrector plate as well as the extra reflections add to the accuracy.

Hope this gives you some other ideas as the microswitch may not be the most appropriate sensor for homing a folded optical system. Also could you use the diffusion to detect sharpness of a known star using a CCD pinched from a defunct autofocus camera?!!!! Can the change in drive torque (motor current) be used?

Regards

Pat

Is there a mechanical stop on the focus knob? One technique I have seen done is to issue more step pulses that it would take to move the complete length of travel and call the place where it stopped moving "home". A slightly more elegant approach would be to put an encoder on the step motor and step until you see the mechanical encoder stop moving and consider that position "home".

Both of these approaches require a mechanical stop and one that does not bind up the mechansm to move away from.

If you don't have these two conditions, you might be able to use an inductive proximity switch and sense the mirror position through the wall of the scope body (assuming a fiberglass or other non-conductive body). This might have repeatability problems though.

Good Luck,
Bob

How many turns of the knob are needed for full travel of the mirror? It must be more than one or you would not have any problem.

Mount a small gear on the knob driving a larger gear so that larger gear rotates less than one revolution during full mirror travel, and have stops on this gear trigger limit switches.

You did not mention what kind of software you intend using?
Off the shelf CNC software may not be exactly suitable, If you have used Pic Micro's etc, a small dedicated system could be made up, and if more than one turn was required you could look take a leaf out of the robot method and use an absolute encoder, no need to home either.
Al.

The focuser will be controlled by an Atmel microcontroller so software isn't an issue. On the PC the software that controls the focuser will be sending ASCOM focuser commands (generic positioning commands sort of like Gcode).

Absolute encoders? I'm not familiar with those. I've only used steppers in an open loop setup. Do you know where I can find such beasties? This motor will be small as the knob takes very little effort to turn.

I am not sure if this will be accurate enough for you, but if a precision multi-turn (x10/x20) potentiometer were used as a feed back device, and then analogue to digital conversion, a absolute position reading should be possible, I have not used the Atmel, only PicMicro, but I am sure they have the same capability.
Something to look into?
Steppers without gearing are fairly coarse positioning devices.
Al.