Jump to content

Banner.jpg.b89429c566825f6ab32bcafbada449c9.jpg

Home Made Remote Controlled Filter Wheel


Gina

Recommended Posts

Hi all

I don't think I se any problem with an open loop motor system that uses step countng or position encoding on the motor to control position - after all that's what most telescope controllers do. Feedback from the disk is then a single home point at most. Drive on the edge using drive wheel, or geared, drive on the side using cone or rubber wheel, drive at the hub or through the hub, all will workl fine, although the last seems a bit high impulse though and less accurate.

There is a statement saying you want high positioning accuracy - why? - surely there is an error budget based on the size of the filter, the size of the aperture, the size of light cone and the size of the detector. As long as the positional accuracy repeats within that error budget you really shouldn't care. Repositioning for flats and darks at the end of the session (assuming LRGB) are always going to result in misplacing the dust doughnuts and specks because your positional accuracy is never going to be good enough to get the few microns accuracy you need to reliably place them back on the same pixels - aren't they ?

Seems there are some assumptions that haven't been aired yet. What is the normal processing sequence for an LRGB imaging session ?

regards

Mike

Link to comment
Share on other sites

  • Replies 214
  • Created
  • Last Reply
Hi all

I don't think I se any problem with an open loop motor system that uses step countng or position encoding on the motor to control position - after all that's what most telescope controllers do. Feedback from the disk is then a single home point at most. Drive on the edge using drive wheel, or geared, drive on the side using cone or rubber wheel, drive at the hub or through the hub, all will workl fine, although the last seems a bit high impulse though and less accurate.

There is a statement saying you want high positioning accuracy - why? - surely there is an error budget based on the size of the filter, the size of the aperture, the size of light cone and the size of the detector. As long as the positional accuracy repeats within that error budget you really shouldn't care. Repositioning for flats and darks at the end of the session (assuming LRGB) are always going to result in misplacing the dust doughnuts and specks because your positional accuracy is never going to be good enough to get the few microns accuracy you need to reliably place them back on the same pixels - aren't they ?

I tend to agree. Its been said that repeatability is needed to dust speck level for using flats but if all exposures of a given colour are taken together before moving on to the next colour there wouldn't be this need for precision repeatability.
Seems there are some assumptions that haven't been aired yet. What is the normal processing sequence for an LRGB imaging session ?
Yes, I wondered that - I've got a lot to learn yet! :)

Maybe there's a reason for taking all lights, then all darks, then all flats or whatever. Maybe someone can answer this :)

Link to comment
Share on other sites

Following from the above, I would have thought that dust specks may well move about with the rotation of the wheel anyway. I think scrupulous cleanliness is the best policy and not bother too much about trying to get ultra high repeatability precision.

Link to comment
Share on other sites

I'm thinking I'll drive the wheel either by friction on the edge or the inverted belt and a toothed wheel, and use optical or magnetic position sensing (Atik use optical I think). If the wheel is driven with sufficient reduction gearing it should be quite possible to stop the wheel accurately enough. If Atik can do it I don't see why I can't :)

Link to comment
Share on other sites

Some fantastic and diverse ways of driving a filter wheel in here. Some great links too. Think we are missing the hydraulic options using an empeller. I'm joking....hmmm?

Gina - are you motorising the wheel is this because you just want to remote control it, or because you want to drive it from an ASCOM enabled port for serious automated RGB stuff. If it's the latter then you need to consider how you feedback the status and recognise which wheel is active, when it transit etc?

Personally, I might opt for microswitch & notches, or perhaps some spring loaded electrical contracts on the rim, each wired to a different resistor to feedback what filter is actually selected etc? I also found a small amount of friction is your friend. It saves you having to apply braking and holds your selection in place.

Link to comment
Share on other sites

Some fantastic and diverse ways of driving a filter wheel in here. Some great links too. Think we are missing the hydraulic options using an empeller.

Not to mention the goat-powered gin wheel. An exceptionally underused power source for astronomy since that unfortunate accident at one of Tycho Brahe's parties when they ran out of liquor and a kind though misguided guest thought they would replenish supplies from the gin house...

James

Link to comment
Share on other sites

I had 2 thought's about how you might achieve the rotation, my first thought was a rubber wheel on the motor that just engages the edge of the filter wheel.

my 2nd thought would be to use a small stepper motor, I know you don't want to go crazy with electronics, and you don't have to, you can get a 'louvre window' stepper motor that is pre-geared to 4096 steps per-revolution, a very simple driver can be built using a uln2003a (1diode!! and a chip), then you just need a method to make it step, this could be via a microcontroller. You could of course use a driver chip that has step/dir built in, again, it would be minimal components, won't require any other hardware apart from a button, a power regulator, a switch and probably a couple of caps and a method to attach the stepper to your filter wheel spindle, could easily be direct drive or 2 small pulleys and a small belt internally. It doesn't need to be computer controlled of course but could easily be interfaced to one if you wanted to go down that route.

Link to comment
Share on other sites

Some fantastic and diverse ways of driving a filter wheel in here. Some great links too. Think we are missing the hydraulic options using an empeller. I'm joking....hmmm?
I suppose it could be driven pneumatically :D Fins round the edge and a jet of air!! :icon_salut::evil: Or it could have a linear motor wrapped round the edge or like a large flat induction motor...
Gina - are you motorising the wheel is this because you just want to remote control it, or because you want to drive it from an ASCOM enabled port for serious automated RGB stuff. If it's the latter then you need to consider how you feedback the status and recognise which wheel is active, when it transit etc?
Initially just remote control or control from PC with my software but ideally I would like ASCOM control. For the latter I need to do some research to find out how this would work. I would imagine probably a USB to serial adapter with motor motion control (run or stop) and position sensing. I think it needs more than one data line for sensing as we need a home sensor and "arrival at position" sensor. So that's one output and two inputs to the computer.
Personally, I might opt for microswitch & notches, or perhaps some spring loaded electrical contracts on the rim, each wired to a different resistor to feedback what filter is actually selected etc? I also found a small amount of friction is your friend. It saves you having to apply braking and holds your selection in place.
I would prefer a digital readout of the filter ID but I don't think an absolute ID is needed - only which is the "home" or No.1 position and then count.

As for friction, if the rim is driven by a geared motor no extra braking would be necessary.

Link to comment
Share on other sites

I had 2 thought's about how you might achieve the rotation, my first thought was a rubber wheel on the motor that just engages the edge of the filter wheel.
Yes, I think that is probably the easiest though i quite like the inside out toothed belt idea.
my 2nd thought would be to use a small stepper motor, I know you don't want to go crazy with electronics, and you don't have to, you can get a 'louvre window' stepper motor that is pre-geared to 4096 steps per-revolution, a very simple driver can be built using a uln2003a (1diode!! and a chip), then you just need a method to make it step, this could be via a microcontroller. You could of course use a driver chip that has step/dir built in, again, it would be minimal components, won't require any other hardware apart from a button, a power regulator, a switch and probably a couple of caps and a method to attach the stepper to your filter wheel spindle, could easily be direct drive or 2 small pulleys and a small belt internally. It doesn't need to be computer controlled of course but could easily be interfaced to one if you wanted to go down that route.
I've decided I should do some studying of stepper motors and their control. I know the principle on which they work but details of their control I have not investigated yet. I also need to look into how small you can get them and how cheaply.
Link to comment
Share on other sites

I've bought one of these :- 9g SD90 RC Micro High Speed Servo for Model Car/Truck | eBay

and have modified it for continuous rotation. Two things normally limit the rotation to 180 degrees :-

1. A stop on the plastic output gear, which I cut off with a Stanley knife

2. The pot - a metal plate attached to the shaft which I pulled off and some restriction by the contacts on the rotor. To make sure these wouldn't catch up on the rivets at the track ends, I cut off the spring contacts with a small pair of wire cutters.

I've also removed the circuit board and will drive the motor directly. I don't need either variable speed or bi-directional motion so servo type control is not required. If the full speed is too much, I'll just run it from a lower voltage. I'll probably be dropping the voltage from 13.8 anyway.

Link to comment
Share on other sites

Hi gina, with regard to your servo motor, you will *have* to drop the voltage from 13.8v, that servo is rated upto 6v, I predict that if you don't it will either spin up incredibly fast and burn out or it will burn out instantly!!

With regard to stepper sizes, the 'louvre window' stepper motors that I am currently using are incredibly small, 5V 4-Phase 5-Wire Stepper Motor + Driver Test Board | eBay

As you can see, it's about 28mm diameter for the stepper + its internal gearing, 42mm total width with screw fitting and are about 20mm deep. Incidentally, the driver board they sell is similar to the one I 'developed', mine is missing the leds. All you need to control that board is to pulse the p1 pins from a microcontroller. Plenty of libraries around for connecting mcu's to steppers :icon_salut:

Link to comment
Share on other sites

Hi gina, with regard to your servo motor, you will *have* to drop the voltage from 13.8v, that servo is rated upto 6v, I predict that if you don't it will either spin up incredibly fast and burn out or it will burn out instantly!!
I meant taking say 5v from the 13.8v supply rather than a separate 5v supply. OTOH if I find I want 5v a lot I might use an old PC PSU of which I have several.
With regard to stepper sizes, the 'louvre window' stepper motors that I am currently using are incredibly small, 5V 4-Phase 5-Wire Stepper Motor + Driver Test Board | eBay

As you can see, it's about 28mm diameter for the stepper + its internal gearing, 42mm total width with screw fitting and are about 20mm deep. Incidentally, the driver board they sell is similar to the one I 'developed', mine is missing the leds. All you need to control that board is to pulse the p1 pins from a microcontroller. Plenty of libraries around for connecting mcu's to steppers :icon_salut:

That one looks cheap enough :evil: Might order that and have a "play" :D Thank you :(
Link to comment
Share on other sites

Noone has mentioned making the filter disk into the motor armature as a direct drive motor. Mount a few magnets on it and some coils round the edge and you can drive it to anywhere directly like a stepper motor.

I've long wanted to build one of these to drive a telescope directly to sub-arcsec accuracy.

Link to comment
Share on other sites

Checked the no-load output speed of the servo, fed from my variable bench supply. About 30RPM @ 3v and 60RPM @ 5v. So if I were to drive the disc on the edge using a rubber tyred idler wheel from the splined shaft of the servo, I calculate the filter wheel would take 4 secs @5v and 8s @ 3v to change from one filter to the next. I measured the shaft as about 4mm, the wheel is 145mm and that's with 9 positions. This seem a reasonable rotation speed to me.

Link to comment
Share on other sites

Noone has mentioned making the filter disk into the motor armature as a direct drive motor. Mount a few magnets on it and some coils round the edge and you can drive it to anywhere directly like a stepper motor.
I thought I had in post 85 - not synchronous with magnets but by induction in the aluminium. Yes, magnets and coils would make it a big stepper motor. :evil:
I've long wanted to build one of these to drive a telescope directly to sub-arcsec accuracy.
That would be fun :icon_salut:
Link to comment
Share on other sites

Oh yes, taking 5v from the 13.8 would be of course ok, as I'm sure you already know. I'm just building up an automatic dew heater on stripboard, I used a to92 lm317 to give me 100ma @ 5v, should be enough to drive the microcontroller, all it's got to do is switch a couple of transistors (for now, will swap for more efficient mosfets when I can afford them).

Those little steppers and boards are really very good value for what they are and what they can do, there is a small amount of backlash in the gearing but nothing that you can't get around with some simple code, you just need 4 pins on a microcontroller, btw. that little stepper was capable of pulling the focuser on a william optics ZS66 + canon 400d, borg 2" filter holder, wo field flattener and an extension tube vertically.

I meant taking say 5v from the 13.8v supply rather than a separate 5v supply. OTOH if I find I want 5v a lot I might use an old PC PSU of which I have several.

That one looks cheap enough :icon_salut: Might order that and have a "play" :evil: Thank you :D

Link to comment
Share on other sites

I tend to agree. Its been said that repeatability is needed to dust speck level for using flats but if all exposures of a given colour are taken together before moving on to the next colour there wouldn't be this need for precision repeatability.

I, for one, take my RGBHa shots with alternating filters. The main reason being unpredictable conditions, like you do not know exactly how long you will be able to image. I have a piece of software that creates a Nebulosity script for the capture, and it splits up the session like R G B Ha R G B Ha R G B Ha etc. Come clouds, at least you end up with something that can be used, not just 20 reds and 19 greens :icon_salut:

As for dust specks and repeatability, I never really had a problem wth my SX wheel in that respect. I do not know the positioning accuracy, but apparently it is good enough or I would have seen some serious artifacts in my images.

All the best,

Per

Link to comment
Share on other sites

I, for one, take my RGBHa shots with alternating filters. The main reason being unpredictable conditions, like you do not know exactly how long you will be able to image. I have a piece of software that creates a Nebulosity script for the capture, and it splits up the session like R G B Ha R G B Ha R G B Ha etc. Come clouds, at least you end up with something that can be used, not just 20 reds and 19 greens :D

As for dust specks and repeatability, I never really had a problem wth my SX wheel in that respect. I do not know the positioning accuracy, but apparently it is good enough or I would have seen some serious artifacts in my images.

All the best,

Per

I agree - I think accurate positioning is probably overrated.
Link to comment
Share on other sites

Here is my latest design. The motor is a servo motor modified for continuous rotation and direct connection to the motor. Providing a simple but tiny motor/gearbox. This drives the edge of the wheel using a round belt.

"Home" position (Filter no.1) is sensed by magnet and reed switch. Motor stop position for the filters will be sensed by micro switch driven by lugs on the wheel.

Filter-Wheel-9-2.png

Link to comment
Share on other sites

I'm thinking now about how to connect the wheel to scope and camera. With the weight of the wheel and camera, I doubt a threaded hole in the outer plate would be strong enough, particularly as this is only 3mm thick aluminium. I guess a 1.25" tube would suffice to connect to scope with adapters. Or, for the 2" focuser on the ED80 I could use either a 2" tube or direct screw thread. For that sort of idea does anyone know of a suitable flange fitting of either size that might be bolted to the wheel casing?

For the camera, it depends on the camera, of course, and I have yet to decide that. I think a threaded hole might suffice there, although a larger fixing with suitable adapter would allow for a change of camera later.

I know the main use of the filter wheel will be for LRGB (and NB later) wanting a mono camera but I was wondering about possible filters for use with colour such as my 1100D DSLR. For this I could get an EOS adapter ring and fasten that to the wheel casing.

BTW - Does it make any difference which way round filters are orientated - ie. which way round the filter wheel is? Filter thread towards scope or filter thread towards camera. I wouldn't have thought it would but maybe it could be a question of reflection coatings.

Link to comment
Share on other sites

The belts and micro switch I ordered have arrived and I have been looking at how things might fit. I ordered a toothed belt to look at that idea but got the wrong size. The other belt is a round rubber like material of 150mm diameter and 2.5mm thick. This gives rather more slack than I expected with the 145mm wheel disc and is also quite elastic. More thinking required I reckon :)

This is how things would fit if I were to drive the belt directly off the splines on the motor. The wheel will have small notches in the edge to operate the micro switch when each filter is in position.

post-25795-133877726463_thumb.png

Link to comment
Share on other sites

Gina, Not read the whole thread so if I'm repeating another post sorry. Have you considered putting something luke shrink wrap on the motor spindle and using a friction drive direct to the side of your 145mm disc?

Link to comment
Share on other sites

I've ordered one of these and a 10 tooth matching pulley to drive it as I'm still interested in trying this drive method. Pity it isn't 450mm circumference as that divides by 9 giving 50 teeth per filter. But I'm expecting the 455mm belt to be tight already as the circumference of a 145mm disc (if accurately cut) is 455.53mm (pi=3.14159). Time will tell just how tight the 455mm belt is.
Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.