Submitted by Ludvig Friberg on Sun, 03/21/2010 - 19:17
Posted in
Just toying around with some ideas for an adjustible pan tilt head. If anyone have a good solution for the axles and the adjustible parts please chime in, I have really no idea on how feasible this is. Reason is to be able to mount different cameras, 5dmkII and REDONE/scarlet/epic.
trying to embed Vimeo, not working... code below
<object width="400" height="300"><param name="allowfullscreen" value="true" /><param name="allowscriptaccess" value="always" /><param name="movie" value="http://vimeo.com/moogaloop.swf?clip_id=10325988&server=vimeo.com&show_title=1&show_byline=1&show_portrait=0&color=&fullscreen=1" /><embed src="http://vimeo.com/moogaloop.swf?clip_id=10325988&server=vimeo.com&show_title=1&show_byline=1&show_portrait=0&color=&fullscreen=1" type="application/x-shockwave-flash" allowfullscreen="true" allowscriptaccess="always" width="400" height="300"></embed></object><p><a href="http://vimeo.com/10325988">pan, tilt</a> from <a href="http://vimeo.com/user1140134">lude frib</a> on <a href="http://vimeo.com">Vimeo</a>.</p>
I still need to get a plugin
I still need to get a plugin to allow embedding of videos, will work on that...
That's very similar to what threeaxis has built. (I'm hoping he'll show his pics soon.) It is feasible, but will result in a large amount of weight. If you want to avoid having the linear bearings/brakes used in easy adjustment, and shave off a ton of weight - you can use the economy t-nuts and bolts to make the adjustments. You just loosen them up, set things into position, and then re-tighten them.
Note that your lowest position has to at least let the tilt arm tilt free of the pan base, otherwise you'll have almost zero tilting ability at the lowest position - making it useless there.
For the axles, I suggest using single-sided dual bearing mounts (as I pointed out in the other thread) and beefy shafts. Both can be gotten (and even custom designed online at a low price) from Misumi.
What I've done is to use one end of the shaft turned into a threaded mount, wherein one mounts the final drive gear, and then the other end left plain - going through the dual bearing mount. The plain end must be wider than the threaded end (I used 10mm plain, 6mm/1.0 threading on the other end.) Then a locknut is used to press the gear tight against the stepped-up face of the shaft.
One thing I've learned from that is to get the threading wider than the bore of your gear, and then tap your gear to match the threading. Otherwise, if the bore is even slightly wider than the threading, you'll not be able to make the gear perfectly centered. As soon as I order some new shafts (7mm threading this time =) I'll be able to show, in better quality, what this looks like. I'll try and update the thread (8020 prototype) later tonight with the current pics, so you can at least get an idea as to what I mean.
!c
i second what shutterdrone
i second what shutterdrone says ;) will post shots of mine in a new thread in a sec.
ludvig, i noticed this: http://vimeo.com/9493909
not sure where you are but you should pick up a copy of the US version of wired mag. they have a small piece on the guys behind the robots that shot this: http://www.youtube.com/watch?v=nX_V7W23kwM
If you want a robot arm how
If you want a robot arm how about this Robot Ride
Here's a picture of my head
Here's a picture of my head and dolly.
It's made from 40x40 mm ITEM (same as 80/20) profiles. It works pretty stable, has hardly any backlash or play and is quite light. It's adjustable by just moving the angle brackets. If you would go for bigger cameras that a DSLR, you might want to use larger profiles, like 40x120.
The motors are Nanotec 1,8° steppers with 1,4 A. My drivers go only halfstep, so the resolution is pretty poor, but I'll try the Pololu drivers this weekend (hopefully). With 16th step it should work alright. The first tests have been promising, but the tilt axis could use a larger sprocket..
Looking real slick! Seems
Looking real slick! Seems like a good design for my first head! Are you selling this as a kit? What is the total cost? What is the gearing ratio on the pan/tilt axes? Also how is the Axle constructed? Do you have some detailed images? This is what I find hardest to grasp how to make, good and solid axles/rotarys that can take a few kilos (5-10kg)
Btw, dont you loose almost all the torque going more than halfstep? I have been thinking about running 0.9° steppers at halfstep to get 800 steps per rev. At a 10:1 gearing that is 8000 steps per rev or 22 steps per degree. For my purposes I think that is more than enough. If I find out that I can microstep further and still have torque enough I guess it helps with eliminating oscillation in the rig. I have no experience on this though, so please correct me if I am wrong.
"Btw, dont you loose almost
"Btw, dont you loose almost all the torque going more than halfstep? I have been thinking about running 0.9° steppers at halfstep to get 800 steps per rev. At a 10:1 gearing that is 8000 steps per rev or 22 steps per degree."
In theory, yes, in practice: "it depends."
The reason why I say this is, yes, microstepping does reduce initial torque - however, the torque is incremental, and increasing with each successive microstep. It never reaches the rated torque of a full step. HOWEVER, you can end up eliciting more torque microstepping within a speed band that would normally generate resonance in your motor. Generally speaking, most stepper motors produce the full moving torque listed on their plates at speeds that generate lots of resonance. This resonance eats up torque like no mans business. Microstepping greatly reduces resonance. So, at low to moderate speeds - you'll usually end up with more torque microstepping than full-stepping, as you won't be giving up power in vibration.
The real challenge is that while resolution increases, accuracy decreases when microstepping - but you'll never get the same smooth, nearly musical motion from full-stepping that you will from microstepping, and resonance (vibration) is the enemy of a pan/tilt head!
!c
Ludvig, With a proper
Ludvig,
With a proper balanced rig, you hardly need any torque at all. Obviously, sometimes you have to trade off balance for nodal position, I was lucky so far that it always worked out well. The gear ratio is 10:1 on the pan and 6:1 on the tilt, if I remember right. I'm not (so far) selling the head as a kit, it needs to be tested a little more before I do that.
The axles are constructed fairly simple. The pan axle is a 10 mm silver steel bolt with two bearings, one on the upper and one on the lower side of the horizontal profile of the head. I used adjusting rings on top and below to keep the profile in place.
The tilt axis was more difficult. Initially I used a 5 mm axle, which was way too small. I changed it to 8 mm and now its pretty good. The axle is attached to the inner profile, there are two bearings (one on either side)of the outer profile. The axle has a M8 thread on the outside, which is used to tighten it down with a self locking nut.
It's not really easy to describe. I will post a few pictures this weekend.
Matthias
Matt, Looks like we took
Matt,
Looks like we took similar approaches for the axles.
I used 10mm shafts from Misumi, pre-machined (about US $15/ea) that on one end is drilled and tapped with an M6x1 threading, and on the other end is machined down and put through a die such that half of the shaft is basically an M6x1 bolt. (Next time I will do M7 for fitment reasons.)
The 10mm half of the shaft has two snap-ring grooves for retaining, and a wrench flat. Then, the gear is threaded onto the 'bolt' half of the shaft, and a locknut is used to prevent rotating back off the threading. The 10mm plain half is then run through a dual-bearing housing (again from Misumi, pricey though at about $50/ea) that is made out of aluminum (suprisingly light), and the snap rings keep the shaft from sliding out of the bearings. The side with the gear on it is actually floating freely in the air.
The tilt arm is then attached to the axle by drilling and counter-boring for a bolt through the extrusion, and a M6x1x32mm bolt is put through the arm, and tightened into the threaded hole in the 10mm shaft.
Again, I think I'll need to show good pictures too, to make it obvious =)
!c
Here's a video I shot
Here's a video I shot yesterday. I had a Canon 5D-2 with a 24 mm t1.4 Canon EF-L an the above mentioned dolly & P/T head. Programmed with an alpha version of my Windows GUI (that I have to update for 0.82 big time...I hope I can get to this soon.
I was surprised how smooth it ran (at least in the original video), you can tell by looking at the shadows. Unfortunately, the flowers did not close on sunset, so it's not quite what I wanted, but technically, I think it's ok. If testing wouldn't be so time consuming ;-).....
Matthias
right on!
right on!