Submitted by ardpub on Thu, 02/10/2011 - 21:44
Posted in
[Edit: After getting some feedback and subsequently doing more research, I have abridged this post down to only what is left unanswered]
First off, let me thank everyone who has contributed to these forums as it is truly a wealth of information for anyone trying to build a motion control system or get into time lapse videography. I have tried to educate myself as much as possible before posting but I would like the community's feedback on my rig's design before I buy the parts.
I am directing a documentary in which I would like to do some time-lapse tracking shots, both horizontal and vertical.
This system will be built around the PBC Redi-Rails Low-Profile Linear Bearing. I chose this because it seems to "glide" better for using the slider for conventional tracking shots (non-timelapse). The motor/gearing will be similar to this prototype (http://openmoco.org/node/126) but I plan on turning the belt perpendicular to the guide rail (see image of mock-up).
Project Goals:
-Slider must work "manually" as a traditional slider operated by the camera operator's hands on the pan/tilt head--no motors or cranks
-With belt engaged, slider will shoot "stop, move, stop" style time-lapses and will be capable of vertical tracking (final shots will range from 3 to 10 seconds, time lapse periods will range from 10 minutes to 2 hours and carriage will travel up to 6 feet/72 inches)
-If possible, operator will have ability to manually control dolly move w/stepper driver/stepper motor for very slow tracking shots, macro tracking shots and remote mounting situations (I can add electronics to do this at a later time)
-Target budget: ~$500, a little more if necessary
Gear List (full gear description available in PDF packet linked below):
Part/Part or Model#/Qty/Unit Price/Total Line Price
Redi-Rail T Slot Low Profile Carraige G-RRL34C-B2-019F-C0 1 83.4 83.4
Redi-Rail Low Profile Guide G-RRL34R-072.000-R0 1 193.75 193.75
Anneheim Automation - 17YPG Stepper Motor w/Planetary Gearbox (19:1, 62oz/in torque) 17YPG302S-LW4-R3.7 1 83.6 83.6
8mm Bore, 16mm Outside Dia., 440C Stainless / ISO 6 Ball bearing, lube with Oil A 7Y 5MFSS1608 4 11.41 45.64
Polycarbonate Belt clamp for single and double sided belts A 6M53M060 1 5.16 5.16
3 mm GT2 Pitch, 6mm Wide, Open ended Neoprene Belt A 6R53MC060 4 9.81 39.24
8 / 8mm (Side1 & Side2), Zero Backlash Miniature Slit Type Flexible Coupling. S50MSXM19P08P08 1 22.31 8.66
GT (3mm) Pitch, 45 Teeth, Aluminum alloy Timing Pulley A6A53M045NF0608 2 0
8 (-.005/-.014) Dia, 50mm Long, 303 ST. Steel Shaft A 7X 1M080050 1 3.59 3.59
8 (-.005/-.014) Dia, 65mm Long, 416 Stainless Steel Shaft S40PX0MHG8M-065 1 5.92 5.92
+Arduino board, battery and other necessary components
Here are images of two possible designs using mostly the same parts. I will try to fit the belt underneath the carriage but if there isn't room I will opt to run the belt alongside the guide rail:
Questions/Concerns:
1) The maximimum torque of all variations of the 17YPG stepper from Anaheim Automation is 1250oz/in but the bipolar torque ranges from 34 to 100oz/in. Being that the stepper will be stopped when it needs to move, does it matter what the bipolar torque is? Can I get the lowest torque or should I opt for higher?
2) Is the 6mm wide, 3mm GT2, 333 teeth per meter belt and corresponding pulley adequate for doing vertical-tracking time lapses? (will it have enough bite to hold the carriage with Sachtler head, canon 5d or 7d and wide-angle lens or will I need bigger teeth and/or a wider pulley?)
3) Does anyone know of a tutorial on adding the necessary components to the Arduino to connect the camera and stepper motor(s)? I haven't been able to find any instructions on doing this specific to Openmoco. I feel like I must be overlooking it.
These forums have opened up a lot of possibilities for me and I am very grateful for all the info and anyone's help with my specific questions. THANKS!!!
PDF Packet with designs, and all details on parts: PDF Packet
Hello, Nice drawings. Altough
Hello,
Nice drawings. Altough I find your rig a bit large. The disadvantage of putting your motor on top of the rail is that it will show up on your photos when pointing it at the end of the slider for zoom-like shots. Why not attach the motor directly to the pulley and fix everything under the rail?
I am currently working on my slider, using different motors, the overall idea is the same. Have a look at: www.rtjacobs.nl go to Ron - video.
About your questions: The torque of a stepper is higher then a DC gear motor. so I think you are all right.
The belt is 6mm wide, as long as it is reenforced with glass fiber it is strong enough.
On this moment I don't know anything about the Arduino.
See you,
Ron
I like your rig--very clean
I like your rig--very clean and simple.
Since posting I have started to lean toward a cheaper dolly system made from two parallel shafts and ball bearing pillow blocks but the concept is pretty much the same. I opted for the motor on top because for my documentary, I want to mount the slider on two spikes in the marsh, just above the surface of the water.
Your torque must be higher
Your torque must be higher than the combination of starting torque and full speed torque required to move your payload. Starting torque requirement will be the combination of the total payload weight, the starting friction (stiction), and the loss of efficiency in the drive train. Which, with that design, you will have a good deal of. I'd suggest eliminating the stacked plate design, and keep only the bottom plate. Mount the motor directly to the bottom plate, and the pulley directly to the output shaft. Trying to align those three holes perfectly, and dealing with the slop in the bearings will be a nightmare (been there, done that =)
So, if your payload weight is 5 lbs (5*16 = 80 oz's), and you measure your starting torque (the amount of force applied to break over static friction with no payload) to be 5 oz's, you must have at a minimum 85 oz's of torque to move your payload on level ground. Increase by a factor of 1.5-4x to add pure vertical moves. So, for 80 oz's of payload, 10 oz's of starting torque, and a 2x factor for vertical moves, you would need a minimum of 180 oz/in of torque for a 2" diameter pulley. Reducing your pulley size by half doubles the torque requirement. Note that the lever for the "oz/in" torque measurement with a pulley is the radius length of the pulley.
So, actual torque elicited by motor on payload is: TorqueBadge * PulleyRadius. I.e.: 2" diameter pulley, 60 oz/in motor = 60*1 = 60 oz/in of actual torque can be elicited on the payload, a 3" pulley results in 60*1.5 or 90 oz/in of torque.
For a tutorial on connecting the camera, see: http://openmoco.org/node/88
For how to (basically) wire up some easy driver's to the arduino: http://openmoco.org/node/133
!c
Shutterdrone: Thanks for your
Shutterdrone:
Thanks for your post! Due to some budgetary delays, I haven't started ordering parts yet and I am very happy to have gotten this info prior to committing to the design. I will start working on an updated design tomorrow. Thanks again.
Ardpub- have you actually
Ardpub- have you actually seen or used this rail? The cam roller system looks very well designed, but its a very narrow rail. I can tell you from my experience using a wider Igus rail, that a heavy lens moves the center of gravity outside the track and changes the drag parameters dramatically. Thus the "sticky" criticisms of the Igus design. Well balanced over the rail, it works great. This Linear Bearing may be different due to its use of rollers, but food for thought.