Skip to Content

8020/Acrylic Worm-Gear Pan and Tilt Prototype

Posted in

Here are some photos of my latest prototype, using 80/20 extruded aluminum for the framing, and CNC-cut acrylic panels for the gear housings. I'll follow this up later with parts list, etc.

While this design largely works, it has two critical flaws:

1) Getting the shaft ends to line up perfectly is nearly impossible
2) The small set-screws on the gears and the hubs used to attach to the camera arm and the tripod are too small to hold well even on a d-profile. They require an extra groove added to the shaft to capture the screws, and still can come loose when too much torque is applied while setting up, etc.

For both of these issues, I've got some modifications coming up, removing the need for bearings on two sides, and set-screws.

But, here are some photos for now - I shot a quick video earlier after finishing up some new housings for the motor drivers, so I'll get that uploaded when it's done processing.

Note that I used a lot of corner connectors. The final size (in this case, it's 8" long x 6" wide x 7" tall) hasn't been determined yet. Once I get a good final size, I'll get rid of the corner braces, and tap/drill the ends for a lighter, cleaner connection.

Here's a quick test, after tightening everything up and putting the drivers in their new housings. It was getting dark, so I just set it to run at 100 and 30 steps, with a very slight ramp to it. (Unfortunately, left auto-wb on!) http://www.youtube.com/watch?v=MM1ta04LvjU

!c

Awesome Chris! Looks

slau's picture

Awesome Chris!

Looks absolutely massive and very majestic!

Thanks Slau - this is the

Thanks Slau - this is the kind of design I was talking about. It's not the cheapest in the world, but overall value is amazing. (Weight to performance, cost to flexibility, power consumption -- the worm gearing allows power kill between motion, so batteries are only consumed by motors when actual movement is needed. This also allows really pushing the steppers, so those tiny steppers can run up to 10lbs of kit!) It's not as big as it looks in the picture, either. It only weighs about 6lbs. The final design will remove almost 1 lb of weight from it. Removing the camera arm allows it to easily fit in a LowePro CompuTrekker pack. If one doesn't care for nodal centering, another lb or 2 can be shaved off. The other neat thing (not shown)? You can quickly invert the tilt 'tower' on the pan base by adjusting two bolts -- allowing one to put the camera outboard of the pan, allowing for full 360' tilt rotation.

The primary changes waiting to be made are in the shafts and bearings. If the new changes work great, the only thing that one couldn't order from a vendor from existing stock tomorrow would be the acrylic panels. I'm figuring for someone w/o the tools or wherewithall to make them themselves would run about $50 for laser-cut service.

Anyhow, the design changes are that the shaft part supporting the weight will be increased from .25" to 10mm (0.394") and all of the hubs will be eliminated by using an M8 bolt to attach directly to the center of the shaft (Edmund optics sells an adapter from M8 to 1/4-20 for about $5 for attaching to tripods). The hubs are major source of cost and frustration.

Additionally, to eliminate the issue of the set-screws on the worm gears, the part of the shaft that the worm gear is mounted on will actually be M6 threading. Since the shaft will step down from 10mm to M6 threading, there'll be a nice face to press the gear against, and then a nylock nut is used to compress the gear against this face - virtually eliminating any slip under any torque level, and not requiring constant re-tightening when one elicits too much torque on the arm.

This also, implies that _anyone_ could put it together with a few basic tools. No custom work needed!

The shaft would then be supported on only one end, by a dual-bearing support (10mm bore size), eliminating mis-alignment issues and increasing the strength of the bearings against downward/sideward force.

!c

This design really looks

This design really looks great! Are you switching of the power to the motors by controlling the EasyDriver? Does the Duemilanove have enough pins for this?

I myself am very interesting in the Camblock design. This design looks like I can put the motor and gear in a similar box. Or do you see any issues when this gear setup is used in Camblock like boxes? My prototype could be build based on folded aluminium to build the boxes.

But again, great design. I only have a Duemilanove, some EasyDrivers, motors and some parts on a Breadboard on my desk -).

Yeah, you just set the sleep

Yeah, you just set the sleep pin on the ED's (or whatever driver). The nice thing about a sleep function, is you only really need one pin. It would be a waste of pins to have separate statuses for things that rarely change. So, just use a switch on each driver - use this to override "sleep" for any axis that shouldn't sleep, and then you just use one pin for all drivers.

The Camblock design works fine from what I've heard. The problem putting everything in one box (as camblock does - not just motor and gears, but also some electronics) is that everything has to be what it's designed for to get that compact design. Making one change means you have to re-design the enclosure. Which is fine for most people's projects, but by trying to provide a starting point where anyone could start from, with parts anyone can buy, makes it a lot harder to make things more "generic". In this case, I separated the drivers from the axes (not shown, but in little black boxes that bolt onto the extrusion), and put the motors on the outside of the gear box. This way any NEMA15 motor can be used with the design, not just the little 1/4 stack-length ones I've chosen. And, of course, if someone wants to upgrade their motors, having the drivers separate means they can upgrade those too without re-designing the whole thing.

Putting them all in one box tends to make things more compact, definitely. I would approach that route if that's your need. However, if you feel you're going to experiment with different hardware - I'd suggest going more modular in the overall design. It will save you lots of headaches and boxes of useless panels =) Here's a real example I experienced:

I originally put my drivers inside of the enclosure, with the motors inside too. I designed a circuit board for the ED's, that had inputs/outputs/ and chains through to another controller (one cable to run two motors). I machined in the mounting holes for the driver's carrier board, and a hole to adjust the current limit pot on the ED. Everything was great, until I got to assembling, and then realized that the side I had placed the hole in required me to have proto board that was copper clad on both sides - which I didn't, and the headers/wiring I used prevented me from being able to mount the board upside down w/o hitting a gear, and then no entry could be had from the top side because a gear blocked the pot! Oops! I had to open up the enclosure every time I wanted to adjust the stepper.

So, while I'm leaning towards more general designs to share - you can make it as specific as you need for your application, I'm just trying to reduce the points for mistakes as I share my designs - even if that means they're sometimes a little too simple for most people's needs. =)

!c

Here is the secret. I want to

Here is the secret. I want to use the system underwater. Have a look at: http://www.elysia.nl. I am an underwater photographer and I truly believe this should be possible. I only need the shaft to be watertight. That's no problem that will just need some o-rings. A 100m-depth rating (330ft) should be possible in my opinion! The rest, like watertight plugs etc. can be bought of the shelf. The Camblock systems seems to be the best way for me if I can find someone who can CNC the blocks out of one piece of aluminum. A lid with one big o-ring can be used to close it all up. I will probably need bigger motors because of the environment where I want to use the system. The block needs to take the stepper, ED, Gears and plugs for connecting it. The first (main) block will take the Arduino too. There should be enough room for that don’t you think? The size of the motor and how it connects to the gear (straight of angle) is important, but your setup shows how that can be done.

This will take some time and money. The final boxes are the big issue. I will start with a normal setup like the one you built. The final version will be a block if I will ever get there. I am very enthusiastic about the project and I really want to give this a try.

Silly question, but I've got

Silly question, but I've got ask... *grin* Why not just put a normal setup inside of a big, watertight, acrylic box? Seems like it'd be way cheaper that way, unless you have other requirements. (Not to discount the coolness factor, in any way =)

Can get anything machined out of aluminum around here, I take it you're in NL? Not sure what the machinists scene looks like there, but there are about 50 machining shops within a few miles of my house, caveat emptor... *grin*

If you really wanna spend some bucks, but wanna go easy - there's emachineshop.com, can use their CAD software for free, and then, if you like, have them build it. I use their CAD, 'cos it's easy as all get-out, but do the machining myself.

!c

Because Archymedes taught us

slau's picture

Because Archymedes taught us about displacement.

Ever tried holding an empty plastic bottle (filled with air) at 10m underwater?

... and that's why I'm not a

... and that's why I'm not a diver! Being of the mind I normally inhabit, I would just throw a bunch of lead bricks in there and let whoever had to get it back to the top to sort things out, hehe.

!c

Hi there, This design looks

Hi there,

This design looks amazing. Are there plans to add this pan/tilt design to a dolly soon?
If so, do you (or anyone you know) take commissions to build these rigs?

I am keen on getting a timelapse rig with pan/tilt and dolly functions that can be programmed to incorporate ramping with starts and stops for all axis'

thanks,

Marden

Stay tuned Marden. We're

milapse's picture

Stay tuned Marden. We're prototyping and testing like crazy. What you've described is exactly what we've got going on here. The plan is to get both dolly and pan/tilt in kit form so you can bolt it together and dream up shots.