Jump to content


Photo

Pico Dome (it's that small)

  • Please log in to reply
126 replies to this topic

#51 1965healy

1965healy

    The Snarkster

  • *****
  • Moderators
  • Posts: 8887
  • Joined: 23 Jun 2007
  • Loc: San Antonio, TX

Posted 20 January 2013 - 12:29 AM

Not ugly just a diamond in the rough! You've really made a ton of progress on a lot of fronts. Take your time tho, we've had clear skies the last few nights here in Central Texas and as you get closer to completion you're gonna draw clouds this way.

#52 dobsoscope

dobsoscope

    Messenger

  • *****
  • Posts: 409
  • Joined: 24 May 2006

Posted 20 January 2013 - 05:06 PM

So the gores are only held down at three points along their length? (top bottom and central bolt)

#53 averen

averen

    Vendor Main Sequence Software

  • *****
  • Vendors
  • Posts: 573
  • Joined: 20 Jan 2009
  • Loc: Austin TX

Posted 20 January 2013 - 06:44 PM

So the gores are only held down at three points along their length? (top bottom and central bolt)


That's how they were attached to the frame. The glass and epoxy is really what keeps everything together.

But yes. Only those fasteners on each gore. Although there are more along the top than just one.
Jared

#54 dobsoscope

dobsoscope

    Messenger

  • *****
  • Posts: 409
  • Joined: 24 May 2006

Posted 21 January 2013 - 05:19 AM

thanks, then you could I assume have done away with the ribs. I am skinning mine with aluminium sheeting which gets nailed/screwed down onto 1 inch thick ribs.

#55 averen

averen

    Vendor Main Sequence Software

  • *****
  • Vendors
  • Posts: 573
  • Joined: 20 Jan 2009
  • Loc: Austin TX

Posted 21 January 2013 - 09:12 PM

thanks, then you could I assume have done away with the ribs. I am skinning mine with aluminium sheeting which gets nailed/screwed down onto 1 inch thick ribs.


Yes, and I had initially considered building a frameless dome. But this turned out to be easier for me to build. Plus removing the ribs now would be considerably more difficult!

I made a ton more progress this weekend! I had today off and I spent a good portion of it working on the dome...with the exception of about 4 hours when I met to pickup my new 12" dob :-) (that will not even get close to fitting in the dome)

First I drilled and countersunk the rails for the shutter to ride on. The shutter will ride on teflon pads on these rails. The rails are 1.5" wide. The shutter will be held on by by pieces of wood or aluminum on the bottom of the rail.

Posted Image

I also made and fitted the lower shutter. It's not 100% done. There will also be some pieces that go on the outside of the uprights to fully seal the lower shutter. Here's the shutter in place:

Posted Image

And here is a shot of the shutter removed:

Posted Image

The lower shutter will not be automated in any fashion whatsoever. It will be held in place with 4 bolts running into the uprights. If I ever decide to image that low I will manually remove the shutter. It's doubtful that it will ever come out...with the exception of passing equipment in and out of the dome.


Other than the physical dome work I've been breadboarding the controller and making sure that everything will work there. I managed to burn up the voltage regulator on one of my microprocessors. Apparently it wasn't rated for 12V...so I've ordered a couple more voltage regulators and I'll get that fixed once they get here.

I can't recall if I've mentioned how the dome controller will work...so here goes.

The dome controller will consist of 2 microprocessors (AVRs, I'm using Arduinos). One unit will be physically attached to the computer. The other unit will be in the top of the dome.

They will each be responsible for the following pieces:

Computer attached unit
  • Rotation
  • Forwarding communication to dome unit
  • Finding rotation home
  • Calculating dome rotation position

Dome unit
  • Opening/Closing the shutter
  • Controlling my flat field

The dome unit has a TON of safety code in it for the shutter. Obviously I have to rely on some sort of wireless communication between the two controllers so there is a lot of communication built in that just makes sure that the communication is valid. And if the unit that controls the shutter doesn't hear from the rotation unit for a certain amount of time it will automatically close the shutter.

Anyways I'm really looking forward to getting some more work done on the dome. I hope to etch one of the PCBs sometime this week.

Thanks,
Jared

#56 Starhawk

Starhawk

    Space Ranger

  • *****
  • Posts: 5507
  • Joined: 16 Sep 2008
  • Loc: Tucson, Arizona

Posted 22 January 2013 - 11:54 AM

This is pretty cool. Making the templates from paper attached to the dome was definitely the right way to go. Do you have a plan for a roller track? The more circular it is, the better. However, with a dome this small, I am wondering if it could touch the track in only three places (which would give it a lot of resistance to out-of-round conditions).

-Rich

#57 averen

averen

    Vendor Main Sequence Software

  • *****
  • Vendors
  • Posts: 573
  • Joined: 20 Jan 2009
  • Loc: Austin TX

Posted 22 January 2013 - 12:02 PM

[quoteDo you have a plan for a roller track? The more circular it is, the better. However, with a dome this small, I am wondering if it could touch the track in only three places (which would give it a lot of resistance to out-of-round conditions).[/quote]

I have an idea in my head :-) I actually attached the track this weekend, it's in the photos above. And as you mentioned the shutter will have 3 contact points with the track to keep binding to a minimum. Although I'm pretty confident that the arches are very round. At the moment I'm contemplating using some angle aluminum for the front and back of the shutter and a flat piece in the center. Attached to these aluminum pieces will be some teflon pads to slide on the aluminum guides.

I'm still trying to work out the actual shutter mechanism for opening/closing the shutter. I'm thinking I my make a curved rack and pinion gear to move the shutter up and down.

Jared

#58 averen

averen

    Vendor Main Sequence Software

  • *****
  • Vendors
  • Posts: 573
  • Joined: 20 Jan 2009
  • Loc: Austin TX

Posted 27 January 2013 - 06:11 PM

I didn't get a lot done this weekend. I helped a friend pour his pier for his observatory. We used around 1 ton of concrete for his base and pier. Gives me some practice for when I'm ready to pour mine. He'll also be helping me pour mine so that was some additional incentive to help him :-) He rented a 2 bag concrete mixer...money WELL spent! I'll be doing the same when I go to pour my pier. I may try and dig next weekend and then do the pour the following weekend. I can at least use the pier while I'm still building :-)

I managed to mess up my back unloading some extra bags that he gave me when I got home. So for the majority of today I've been limping around.

I did manage to get a little work done. I built the dome rotation sensor and verified it's functional with the board.

Posted Image

The hardest part was stretching the rubber around the exterior of the wheel!

There are 16 holes around the wheel, a "tick" will be triggered at each change so that's 32 ticks per rotation. This will give the dome movement a resolution of around 1/2 degree.

Jared

#59 averen

averen

    Vendor Main Sequence Software

  • *****
  • Vendors
  • Posts: 573
  • Joined: 20 Jan 2009
  • Loc: Austin TX

Posted 28 January 2013 - 12:43 AM

I managed to get a little more done this evening.

The dome controller is almost finished! This controller is in charge of rotation as well as communicating with the other controller. As it is much simpler than the other controller I just used a prototyping shield for the Arduino rather than etching a circuit board.

If you click through there are notes on different items in the image.
Posted Image
Dome Rotation Controller

I also tested the entire system. That Wondermotor has some torque! It literally jumps to life on start up...It gave my wife, dog and I a startle when it started up the first time. After that I made sure I had a hold of it!

Posted Image
Test Setup

So that's one controller down...one to go. I was surprised I didn't have to fix anything on it and that I didn't blow it or my computer up! My electronics skills aren't all that great. I prefer the software side!

Jared

#60 cn register 5

cn register 5

    Viking 1

  • -----
  • Posts: 760
  • Joined: 26 Dec 2012

Posted 28 January 2013 - 04:03 AM

A couple of thoughts - based on my experiences in automating the dome on my Astro Society's observatory.

The jump when the motor starts may damage the dome or the motor supports. I am using a PWM output to allow the dome power to be applied and removed smoothly. It also slows the dome down as it approaches the target position. I started with a relay to set the direction and a MOSFET to control the power but after burning out a few mosfets I fitted a Poholu DC motor controller that's rated to handle 15A. We are using what looks like the same size of motor.

The other thing is that I an using an electronic compass module to get the dome position, a Sparkfun breakout board. This is bolted to the dome. It should be easy to fit this to your dome Arduino running it off the 3V3 supply and the I2C data lines.

Chris

#61 averen

averen

    Vendor Main Sequence Software

  • *****
  • Vendors
  • Posts: 573
  • Joined: 20 Jan 2009
  • Loc: Austin TX

Posted 28 January 2013 - 09:58 AM

I am using a PWM output to allow the dome power to be applied and removed smoothly.


I will be running PWM on the motor as well. But for testing I just had it full on/off. I haven't implemented the ramp up/down but I'll be doing that this evening. I actually considered using an RC speed control for the motor as well. They're fairly inexpensive and very easy to control with an Arduino. I may still go this direction. I made the mosfet "socketed" so pulling it out is trivial. I will have a large heat sink on it if I keep the mosfet.

The other thing is that I an using an electronic compass module to get the dome position, a Sparkfun breakout board. This is bolted to the dome. It should be easy to fit this to your dome Arduino running it off the 3V3 supply and the I2C data lines.


On my setup this would need to be on the shutter controller. So it would make it more complicated. I think for the time being I'll stick with the rotation encoder. If I have issues with that then I will check out the electronic compass. Thanks for the heads up!

Jared

#62 averen

averen

    Vendor Main Sequence Software

  • *****
  • Vendors
  • Posts: 573
  • Joined: 20 Jan 2009
  • Loc: Austin TX

Posted 28 January 2013 - 01:18 PM

Chris
Were you using any type of flyback diode across the motor when you burned up the Mosfets?

Thank
Jared

#63 cn register 5

cn register 5

    Viking 1

  • -----
  • Posts: 760
  • Joined: 26 Dec 2012

Posted 28 January 2013 - 04:11 PM

I had a diode across the power supply, the MOSFETs had an internal diode and there was a surge suppressor across the motor terminals. The supressor didn't survive either.

The MOSFET was OK for a while but once I got more ambitious about PWM motor speed control it failed within an hour or two. No idea why.

One thing you will find with PWM control is that you will have a singing dome. The B flat two octaves above middle C I think :)

#64 averen

averen

    Vendor Main Sequence Software

  • *****
  • Vendors
  • Posts: 573
  • Joined: 20 Jan 2009
  • Loc: Austin TX

Posted 28 January 2013 - 04:26 PM

The MOSFET was OK for a while but once I got more ambitious about PWM motor speed control it failed within an hour or two. No idea why.


Probably not a big enough/fast enough diode? High frequency PWM can cause a HUGE spike when the motor (now an inductor) is turned off. Since PWM pulses the motor constantly the collapsing electromagnetic grows very large to a point when the diode can no longer keep up and eventually...poof! This can happen when any coil is pulsed fast enough.

I believe it's also common to put in a resistor with the diode when using PWM so that you consume the spike from the motor. If you have no resistor then you're using the coil in the motor as a resistor to consume the spike. If the spike is large enough and your frequency is fast enough you might not have adequate time to consume your spike and it will just keep growing and growing each time your magnetic field collapses (motor is turned off...and in PWM land that happens a LOT)

I think I'll learn from your experience and calculate out what diode and resistor I will need...I was just going to hope and guess...maybe I'll just order a controller. Also last night I was doing PWM and noticed that the Mosfet was actually shocking me a little when I touched it to see if it was hot. Just the back EMF that I wasn't capturing from the motor. I shut it down before I did any damage.

One thing you will find with PWM control is that you will have a singing dome. The B flat two octaves above middle C I think :)


This depends on the frequencey of the PWM. If you're in the audible range you'll be able to hear it :-) Maybe I can switch PWM frequency fast enough to play the "Flight of the Bumblebee" or something as the dome is rotating :-)

The default PWM frequency on the Arduino is either (around) 500Hz or 1000Hz depending on the pin used. Both are in the audible range. The only one that is inaudible is 31kHz and I believe that's barely audible, definitely on the upper range of human hearing...neighborhood dogs and young kids would hate you.

Jared

#65 averen

averen

    Vendor Main Sequence Software

  • *****
  • Vendors
  • Posts: 573
  • Joined: 20 Jan 2009
  • Loc: Austin TX

Posted 29 January 2013 - 11:02 AM

After looking through what I would need for flyback protection I have decided to buy the Pololu controllers:
http://www.pololu.co...og/product/1376

The motors should pull around 3A so a 15A controller will be more than adequate.

They have reverse as well as ramp up and ramp down. This means the shield I made can be completely replaced by this controller...I just need a way to attach the XBee and I should be good.

Thanks for the heads up on these controllers!

Jared

#66 cn register 5

cn register 5

    Viking 1

  • -----
  • Posts: 760
  • Joined: 26 Dec 2012

Posted 29 January 2013 - 04:18 PM

I got the high power motor driver, this one
http://www.pololu.co...log/product/758
it may be overkill but I was fed up with killing mosfets.

I didn't need the USB input as I was connecting to the existing Arduino using the existing code and ports.

It will be interesting if the XBee can be connected to through the serial port on the board.

Chris

#67 csa/montana

csa/montana

    Den Mama

  • *****
  • Posts: 86174
  • Joined: 14 May 2005
  • Loc: montana

Posted 29 January 2013 - 04:24 PM

Keep us posted! :)

#68 averen

averen

    Vendor Main Sequence Software

  • *****
  • Vendors
  • Posts: 573
  • Joined: 20 Jan 2009
  • Loc: Austin TX

Posted 29 January 2013 - 04:28 PM

I got the high power motor driver, this one
http://www.pololu.co...log/product/758
it may be overkill but I was fed up with killing mosfets.


No such thing as overkill!

I didn't need the USB input as I was connecting to the existing Arduino using the existing code and ports.


I don't really need USB either. But it will make setting up the unit much easier...I think. The board supports automatic ramp up/down which will make my life a little easier.

It will be interesting if the XBee can be connected to through the serial port on the board.


The XBee will not be connected directly to this board...although I had contemplated that for my other controller to be honest. Both use 3.3V logic so I would think that the XBee would mate nicely to this board. I will likely try this "just because"

In my setup the rotation unit will be connected directly to a motor and use an XBee to communicate with another arduino which will open and close the shutter as well as control a light box.

So the rotation unit will be connected to the PC via the hardware serial. Then it will have a software serial to communicate with the Shutter/Lightbox via XBee and another Software Serial to communicate with the rotation motor controller. I had considered just using the RC Servo output but a serial connection makes me feel more comfortable.

I haven't been able to find an Arduino library for this controller yet...so I guess I'll be writing that too...should be easy enough.

Thanks again,
Jared

#69 averen

averen

    Vendor Main Sequence Software

  • *****
  • Vendors
  • Posts: 573
  • Joined: 20 Jan 2009
  • Loc: Austin TX

Posted 02 February 2013 - 11:22 AM

I got the high power motor driver, this one
http://www.pololu.co...log/product/758
it may be overkill but I was fed up with killing mosfets.


Thanks again for suggesting the Pololu controllers! They're nothing short of awesome! I picked up the "Controller" versions rather than the "Driver" version. The driver has more basic control...generally all that's needed for our purposes. But I'm lazy and as such wanted to automatic ramp up/down that the controller version offers.

Depending on your dome setup I'm pretty positive that you could get away with just using a couple of the Pololu Controllers to control your rotation and shutter. The controller has the ability to do ramp up/down as well as the ability to have limit switches for each direction (what I'll be using to trigger when the shutter reaches the end during open/close).

You can also daisy chain multiple of the controller versions together when using a serial connection. I wrote a quick and dirty Arduino library that supports daisy chaining.

The only thing I don't like about them is that the serial connection pinout does not conform to the very standard FTDI layout. So I can't directly use an FTDI cable to connect to the board. This really isn't a huge deal as I won't be controlling them this way, but it would have been a nice plus for anyone that wants to control them via a serial FTDI cable.

It will be interesting if the XBee can be connected to through the serial port on the board.


This does work nicely. With the BEC on the controller in 3.3V mode I powered the XBee and controller off of the same battery. Then I just ran the RX from the XBee to TX on the Pololu controller and did the same for the TX/RX lines. I then wrote an Arduino sketch that used the Pololu library that I created and assigned it a software serial port and easily controlled the motor from another arduino directly through the XBee.

I didn't attempt this directly with my PC as I didn't feel like writing something that would handle binary serial. I already had that library for the Arduino so it literally took less than 5 minutes to test this.

Hope that helps anyone considering these controllers for dome control!

Also they have another controller that can even take an encoder as input, so you could use that to get your dome position.

Jared

#70 averen

averen

    Vendor Main Sequence Software

  • *****
  • Vendors
  • Posts: 573
  • Joined: 20 Jan 2009
  • Loc: Austin TX

Posted 06 February 2013 - 12:16 AM

Software keeps chugging along! We had company this last weekend and I also planted a tree (good hole digging experience!) so I didn't get too much done on the dome. I hope to dig the hole for the pier and finish the shutter this weekend, I'm also trying to find a pier plate so I can pour concrete weekend of the 16th...lofty goals I know.

I feel that the controllers for the dome control are nearing completion. The only piece I have left is the ASCOM driver and that should be a piece of cake thanks to WCF and C#!

There's quite a bit of software and hardware involved since I have 2 controllers in the dome. Software wise I have written:
  • Controller software for Shutter/Lightbox controller
  • Controller software for rotation controller (also the relay to the Shutter/Lightbox controller)
  • Dome Server (seen below) which brings the functionality of the controllers together
  • Custom flatbox for Sequence Generator Pro.

All that's left is the ASCOM Dome controller which will interface with the Dome Server. And I'm sure a lot more things once I actually test everything together! I'm pretty confident that the Shutter/Lightbox controller is complete and functions well. I've been testing it for stability for the last 48 hours without a single glitch and have done a TON of other testing with that controller.

Here's a screenshot of the Dome Server and the control panel from SGP showing the light box control:

Posted Image
DomeDuino Server and SGPro

Couple of things to note on the above image:
- Brightness is for the lightbox. This was set in SG Pro and then the Dome Server updates via the device, so this is a real value.
- Shutter is closed...the Pololu controller has the ability to use a couple of the inputs as kill/limit switches. I'll use a reed switch for these and the motor controller will handle stopping the motor (I was glad the controller could do this...it kind of scared me!)
-The "Relay" portion is used to send serial commands directly to the controllers. The rotation controller is directly attached to the computer but the >RS; command is actually a command used on the Shutter/Lightbox controller to get the battery status. The >R designates that the rotation controller should pass the command to the shutter/lightbox and echo back the result.

Unfortunately there's not much else to show at this point unless I post some code...and I have a feeling that most don't want to look at that!

Hopefully this weekend I'll have a much more interesting update! However this is a pretty huge milestone for me...just not a lot of tangible stuff to post. Plus the actual structure is more fun!

Jared

#71 averen

averen

    Vendor Main Sequence Software

  • *****
  • Vendors
  • Posts: 573
  • Joined: 20 Jan 2009
  • Loc: Austin TX

Posted 09 February 2013 - 09:25 PM

Finally got the shutter construction done! (mostly...)

First I cut the arches with my router then attached them to some bracing. After that I cut out 6 PTFE pads and countersunk them to act as the slides. Working with the PTFE was a breeze...that stuff was extremely easy to work with!

Here is the shutter in place. The shutter overlaps the lower shutter by about 1.5".

Posted Image
2013-02-09 16.43.07

Once the frame was completed I sheeted it with 1/8" board.

Posted Image
2013-02-09 17.46.20

The shutter slides VERY nicely on the PTFE. I'm worried that the shutter might be a little heavy to close so I may need to add a counterweight or a spring to help getting the shutter moving. It's not that heavy but when open all of the weight of the shutter is towards the back. When closed the shutter doesn't go as far down as when open so I don't think opening it will be an issue.

The shutter also goes a good deal past zenith, not a ton but it should be more than adequate to track through. I could have actually moved the shutter back a little bit more if I would have planned this better.

Posted Image
2013-02-09 17.47.04

So now I need to figure how how I'm actually going to mechanize the shutter. I'm considering a bike chain attached to the top of the shutter and a small sprocket to engage the chain.

Hopefully I'll get something figured out. I think I'll be digging the hole for the pier tomorrow!

Jared

#72 dobsoscope

dobsoscope

    Messenger

  • *****
  • Posts: 409
  • Joined: 24 May 2006

Posted 10 February 2013 - 05:54 AM

Well done ! I have divided my slit into 5 portions. 1 part fold down flap, 2 parts sliding shutter, 2 parts fixed at the back.

How strong did you find working with the pocket hole jigs? I notice pocket holes in the ribs to bottom ring joints. Did you use the same idea to fix the ribs to the two main central arches forming the slit?

#73 averen

averen

    Vendor Main Sequence Software

  • *****
  • Vendors
  • Posts: 573
  • Joined: 20 Jan 2009
  • Loc: Austin TX

Posted 10 February 2013 - 10:27 PM

Got a little more work done today! I'll be pouring the pier next weekend (weather permitting). So this weekend I dug a large hole for the pier to go into. The footing will be 3'x3' and 2' deep. I still need to measure for the height of the pier...it won't be very tall.

Posted Image

I should also be picking up my mounting plate from the metal folks tomorrow.

And I fixed a couple of small issues with the controllers communicating back and forth. I keep adding in things that are nice to have but not really mandatory. Namely battery/temp status from both the controllers. Here the shutter controller is hooked up but the rotation motor controller is disconnected. Thus the "junk" values. Values coming from the shutter controller are valid though. Plus the "Azimuth" value is coming from the actual encoder as well.

Posted Image


Well done ! I have divided my slit into 5 portions. 1 part fold down flap, 2 parts sliding shutter, 2 parts fixed at the back.

Any pictures of your dome?

How strong did you find working with the pocket hole jigs? I notice pocket holes in the ribs to bottom ring joints. Did you use the same idea to fix the ribs to the two main central arches forming the slit?


They work fairly well. The real restriction on them is the material (plywood) in hardwood they hold extremely well. But with the amount of ribs there are a fair amount of screws. So they do their job!

Jared

#74 averen

averen

    Vendor Main Sequence Software

  • *****
  • Vendors
  • Posts: 573
  • Joined: 20 Jan 2009
  • Loc: Austin TX

Posted 16 February 2013 - 06:27 PM

What a busy week! It's amazing how much prep work has to be done before pouring concrete. First I looked all over town for a 10" sonotube. I couldn't find any. So I had to use a 12" tube. This isn't horrible but it will mean that I won't be able to track as far past zenith in certain areas or my camera will hit the pier! Then I made multiple runs to pickup all the concrete (800lbs each time...times 3). Finally I put the rebar together last night and rented a 2 bag concrete mixer.

A couple of weeks ago I helped a friend pour his pier, which was nice practice, and it meant that he would help me pour mine. We started after he arrived around 9:30 this morning. After moving all 35 bags of concrete out from the garage to to where the pier would be located we were ready to get started. I had previously built the support for the sonotube but decided I better measure for the total height before we started pouring...good thing too as I had placed the mark for the height above the support 2x4 (it was supposed to go below) so that meant that the pier was 3.5" taller than it needed to be! We poured the "dead man" and while that was setting up I fixed the sonotube and support.

It was somewhat dry today so we decided that 15 minutes was enough time to let the dead man setup before pouring the pier. The concrete actually seemed somewhat rigid after 15 minutes ... but that was a big mistake! We poured the entire pier and as we were tamping it the seal from the sonotube broke and concrete started pouring out the bottom! Not good! We shoveled out what we could and then removed the sonotube completely. Then we shoveled out the rest of the overflow and replaced the sonotube. This time we waited an hour before pouring the pier and the lower concrete held!

Here's a shot of the pier as it drys.

Posted Image
Concrete Pier

So next up I'll start framing out the building. I need to figure out the total height of the building before I start building walls. I should be able to get that completed tomorrow and then I'll start on the floor. The walls will likely be pretty short (around 3' tall) The concrete pier is only 21" tall. On top of that there will be the steel plate (seen in the above picture) and attached to that plate will be a Losmandy MA adapter. Bringing the total height to 27".

Jared

#75 palmer570

palmer570

    Vostok 1

  • *****
  • Posts: 146
  • Joined: 04 Sep 2012
  • Loc: Scranton, Pennsylvania

Posted 17 February 2013 - 02:23 AM

What do you estimate the cost of the dome to be?
How much does it weigh?

Was considering doing something similar but a smaller dome and came across this maker of Plastic Domes .






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics