16' clock with moving hands, need ideas for control and programing

I'm working on a Christmas Carol build and were balls to the wall right now. This show is massive. One element is a 16' diameter clock with a minute and hour hand. I need to be able to control the movements of the hands. The biggest hand is 8' long and weighs in at right around 3 lbs. We used our cnc to cut the hands out of ACM board to keep them light and durable.

Currently I'm looking at a set of Nema 34 - 861 oz-in holding torque motors paired with a microstep drive unit powered by some generic power supplies.

While I believe that I have all the mechanics figured out the major hangup is the HMI. Our run crew is mostly undergrad students and a few overhire. I need an HMI that is stupid simple, point and click buttons that are easy to program and wont break the budget by changing what the controllers are. I've looked into visual studio, Arduino and some other proprietary automation software.

Right now were either over budget or it's to labor intensive to program.

Any helpful hints or ideas would greatly be appreciated!

Jason
 
Well, my approach to the task would be an Arduino or other easy to prototype MCU with one of the many RS-485 -> RS-232 converters ( DMX Shield ) so that you can either set it to an exact time with a scene, or slowly rotate with a fade effect using a DMX compatible lighting board or software suite. If you break down the rotation into
 
I would do with an Arduino based on the constructs you have laid out. I would also start managing expectations with the design team and director. It will take some time to program and any changes to a sequence will take some time and possibly affect the later cues. Operation will be simple but it will take some time to prep.

I am imaging you are driving both hands with one motor and using gearing to get the ratio correct.

You will need to home each hand. I would use a hall sensor like this one to locate the hands in space. If you only have to hit a few targets it might speed up programing to put sensors at every end point of the move. They are cheap enough...

Here is a good youtube video with a basic set up. I'll block diagram how I would program this for your needs. I am by no means an ardunio expert, there may be better ways to do this with but I think this will get you rolling. If you can get a Nav license we could do this via Modbus and it will be much sexier

The HMI would be a box with a toggle switch, a pair of buttons, and a rotary switch. The toggle would switch between Homeing and cue mode. The rotary switch would select the cue. The stop button works as an interrupt, pushing it sets the system into a safe state. The go button runs the selected cue in cue mode or starts the homing sequence in homing mode.

For programing I would wire up a pot for speed control and leave my computer hooked up to the arduino so I could use the serial monitor to watch what was going on.

During the testing and integration phase the first thing I would do is write the homing sequence.
IF toggle is home THEN run HOME - Spin the arms until both sensors are found. Set steps count to zero.

Next I would figure out home many steps it takes to do a complete revolution. I would write a cue that more or less ran the homing sequence but instead of setting steps to zero printed the current step count to the serial monitor when it found the homing sensors.

Now that we can home the clock and know how many counts it takes to do a 12 hour move we can do the math to figure out how many counts it takes to do any move or what delay is needed to the clock to tick in real time.

Thats a lot of stuff, does that seem like a direction you would want to move in? If so I can flush out the cues a little more or help you with any concerns.
 
I would do with an Arduino based on the constructs you have laid out. I would also start managing expectations with the design team and director. It will take some time to program and any changes to a sequence will take some time and possibly affect the later cues. Operation will be simple but it will take some time to prep.

I am imaging you are driving both hands with one motor and using gearing to get the ratio correct.

You will need to home each hand. I would use a hall sensor like this one to locate the hands in space. If you only have to hit a few targets it might speed up programing to put sensors at every end point of the move. They are cheap enough...

Here is a good youtube video with a basic set up. I'll block diagram how I would program this for your needs. I am by no means an ardunio expert, there may be better ways to do this with but I think this will get you rolling. If you can get a Nav license we could do this via Modbus and it will be much sexier

The HMI would be a box with a toggle switch, a pair of buttons, and a rotary switch. The toggle would switch between Homeing and cue mode. The rotary switch would select the cue. The stop button works as an interrupt, pushing it sets the system into a safe state. The go button runs the selected cue in cue mode or starts the homing sequence in homing mode.

For programing I would wire up a pot for speed control and leave my computer hooked up to the arduino so I could use the serial monitor to watch what was going on.

During the testing and integration phase the first thing I would do is write the homing sequence.
IF toggle is home THEN run HOME - Spin the arms until both sensors are found. Set steps count to zero.

Next I would figure out home many steps it takes to do a complete revolution. I would write a cue that more or less ran the homing sequence but instead of setting steps to zero printed the current step count to the serial monitor when it found the homing sensors.

Now that we can home the clock and know how many counts it takes to do a 12 hour move we can do the math to figure out how many counts it takes to do any move or what delay is needed to the clock to tick in real time.

Thats a lot of stuff, does that seem like a direction you would want to move in? If so I can flush out the cues a little more or help you with any concerns.

Thanks for the info! two stepper motors 1 for each hand. I was thinking I would que this monster with something like showexpress and really deal with it more like pan and tilt on a mover to get the hands where i want them. The run of the ques would just be a push of the que button on the screen and I could tweak all the speeds in the showexpress software.
 
Thanks for the info! two stepper motors 1 for each hand. I was thinking I would que this monster with something like showexpress and really deal with it more like pan and tilt on a mover to get the hands where i want them. The run of the ques would just be a push of the que button on the screen and I could tweak all the speeds in the showexpress software.

That will make programming it much easier. You will still need to have a homing sequence. Perhaps in addition to hour and minute hand position you have a separate control channel.
 
I'm having a heck of a time getting this setup to even move. Finally got some movement out of the steppers last night. I was able to change speed but not distance. Right now I'm just focusing on trying to understand the programming side of things before I add in the DMX sheild. ANy info or links to help would be greatly appreciated. BTW here is a picture of the clock!
proxy.php
 

Users who are viewing this thread

Back