I need help with a custom ball bearing joint

Hi,

I've been reading through the forums for awhile, alot of really valuable information here.

I couldn't locate anything that helps me, i was hoping one of you might have more information or point me in the right direction.

I don't have any background in mechanical engineering. I need to build a mount for a camera. There is going to be two axis of motion. I need my camera to sit on a platform. The camera weighs around 4 pounds. the platform the camera sits on needs to rotate about the y axis (it needs to spin "left and right"). The joint i am thinking will be a ball bearing type. That joint needs to be connected to another joint at a right angle so that the platform will also tilt "forward and backward". The motion of tilting and spinning needs to be very smooth, as you get with fluid heads on tripods. Why not just pick up a fluid tripod head, well it's a special application, so i need to make it to fit a special stand.

I'm looking for information on building my own fluid head, i think that might be one way, if anyone knows of any online resources, please let me know, info on that seems scarce.

Another way i think would be to use ball bearings, but i'm not sure if the friction on a ball bearing would be so small that it would prevent fluid movement- and i have no idea how to mount a ball bearing, resources on that would be really useful too.

Thanks!

Rob Trujillo
Vancouver, WA

As i said, i don't know engineering, so i'm limited to simple english.

I can't really undestand... does this have anything to do with RepRap?

Are you trying to make this with a reprap?


sound like you could make this really easily with a 90 degree metal tab, nuts,
bolts washers and basic tools.

if you want the camera to stay in place on it's own using bearings isn't a good idea, since they are made to be frictionless and holding the camera would require friction.

you could use a butter fly not to adjust the tension easily.

mount the camera's Center of Gravity as close to the tilt (up and down) axis as possible.

if you use a U profile you can get it exactly right pretty much by eye-balling it, but the camera won't be able to do a full turn.

in fact..... here is a better design which has everything of both ideas:

Red Profile is attahched securely to camera (can not swivel at all)
Red and Green profiles connected by two bolts on the camera's CG line (put the butterfly nuts on the outside for adjustment).
now you can tilt up and down.

Green is attached to Purple to camera to pan left and right.
Green attached to wood or anything else.

(leave enough room above green and below purple to allow for the tightening of the bolts.)


-Leav

Edited 1 time(s). Last edit at 09/27/2008 07:20AM by Leav.

Thanks for the quick response, i was out for work for a couple of days.. No, the camera doesn't need to sit on its own, that's why i thought ball bearings might be a good idea. The idea you suggested i have thought of, but instead of screws, it needs to be something that will allow smooth movement- decent tripods use fluid heads, for example. The problem with the ball bearings i've seen is that they have far too little friction. Part of the point is to filter out minor random movements, since the main support will still be a human hand. Thanks for the diagram. I've already built something similar and it works great for some applications, but this specific application requires something different. I have a model that i built that works in principle. I've hooked it to the camera and everything is great, but any movement isn't smooth- very uneven and that's picked up by the ccd.. What i used for the model is two PVC 90 degree elbows. If it weren't for the uneven movement, it would be great. That's why i thought of making a ball bearing or fluid type joint. So, a 90 degree elbow with fluid joints or ball bearing joints that can connect to another of itself is what i need.

Sorry if this is the wrong place. I found this forum while searching for info on building it myself. The top of the page says "Discuss any sort of mechanical stuff here: positioning systems, gears, extruding methods, etc.", i didn't realize until your reply that this was a reprap specific site.. That's interesting in itself, i'll be back to do alot of reading..

Thanks for the info.

Rob T.

Sure thing, I'm guessing that we'll be seeing you around here soon enough...

regarding the PVC: sounds good.... try sanding the inner tube down a bit (very small amount) and adding lubricant.

Good luck!
-Leav

Low tech idea from another source, not me...

use a rubber band to pull the head around. Loop a rubber band around the handle that you use to pan the head and pull on the rubber band. The band will even out any jerkiness caused by your hand and body and give a smooth glide to your pan. I saw this in a clip online somewhere. Worked really well then you don't have to worry about paning being jerky. Unless I completely missunderstood your post. If so, ignore me!

Demented

|I don't see how this would solve the problem since I think the problem is from the difference between static and kinetic friction.

because max static friction is higher then the kinetic friction then what happens is this:

1)you exert a force and nothing happens because it is too small to overcome the static friction.
2)you slowly increase the force.
3)you suddenly overcome static friction, the object starts moving and the force of friction is now lower.
4)you stop exerting force (or just reduce it)
5)the object stops moving.
6)go back to 1)

hence: jerky motion.

how would a rubber band help?

-Leav

I think what you need is a steadycam
[www.amazon.com]

'sid

Leav, found the link I was talking about...this should help.

[www.wonderhowto.com]

Hope that helps.

Demented

DC:
Yeah that makes sense if your hand is the source of the jerkiness... I would have thought that friction would be the culprit...

I'm sure that even the $50 tripods have bearings good enough for the friction to be a non-issue, but I was thinking that maybe a homemade one would not be so lucky.

-Leav

Ah, i see. Well, if the break away friction is the problem, then no, this won't work.

Demented

Demented,

That's a great idea, and i've used variations of it, never for live events, hasn't been dependable enough so far but i'm spending time with it..

Leav,

when i spin one of the 90 degree elbow PVC pieces (it's partially screwed onto the thread of an identical piece, using the thread as a pivot or hinge), there's the friction that i get when i start movement, the movement is abrupt at first but smooths out. Same with stopping.

Unscrewing the threads a little helps alot with that, but there is a lot of play in the heads, the less screwed in, the more play, which translates to another kind of uneven movement for the camera. That's why i was thinking bearings might work- would have to be a bearing that's packed with a very thick lube so that there's at least some friction to smooth out random movements from your hand (from what i've read, most bearings are designed to have as little friction as possible, that works against me here..). I've been reading alot and talking to mfg..


And Leave,

Yeah, the hand is the source- i have what amounts to a tripod (custom made for.. i won't go into it..). On the tripod is this special piece. Attached to the piece is the camera. This piece that we're talking about doesn't have enough stability on its own to support the camera weight by itself- that's part of the idea- if it did, the movements would be too jerky to use the footage. So, i steady the tripod with one hand and have my camera in the other. I can rotate the camera about the center axies of the PVC pieces this way. In principle, it's perfect for what i need, in practice, i need to find a better joint (bearings, maybe fluid head) as a joint between the 90 degree elbows, instead of using the threads. The 90 degree angle is needed, the threads are a means to an end.

Does anyone know of any reading material that goes into detail on how to mount common bearing types? The material i've found so far seems paced for engineers.

Also, if anyone knows resources for custom building fluid heads, that would be a good help too (same thing, most material i've found is geared for engineers.)

Thanks,

Rob

Hi Rob,

I was assuming you were using thread-less PVC pipe (the kind glued together).

You should try it, it's a tight fit, which is perfect because you could sand it down and add lubrication as necessary.

regarding bearings: I think you should try just plain old bolts through wood. worst case scenario: through in a bit of PVC pipe around the bolt to get a smooth surface.

-Leav

Or cushion the two pieces of wood with PTFE from McMaster. Teflon gives a good glide against itself, isn't that expensive, and yet isn't completely slick. You could set it up so that you have wood/PTFE/PTFE/wood with a bolt as Leav suggested through the whole bunch. The PTFE would be attached to the wood but not to each other. That, I think, should work wonderfully.

Demented

HDPE is about 98% as effective in that role as PTFE and costs virtually nothing by way of comparison.

Really?! That's pretty sweet. I'll keep that in mind. Thanks, Forest.

Demented

"Many thermoplastic materials such as nylon, HDPE and PTFE are commonly used for low friction bearings. They are especially useful because the coefficient of friction falls with increasing imposed load."

[en.wikipedia.org]

While you see numbers for PTFE for coefficient of friction of 0.04-5, the practical values tend more towards 0.08-0.1. HDPE consistently gets you about 0.2 and costs you about 10% as much. I've measured the dynamic friction coefficient of HDPE down to 0.08-0.1 using informal methods.

If you are thinking about really heavy loading of the bearing surface, however, you need to go for ultra high molecular weight polyethylene (UHMWPE) which has the requisite engineering properties.