DIY gyroscopic stabilizer using 2 quiet PC cooling fans?

Peter J. DeCrescenzo

Peter J. DeCrescenzo

Veteran
Thinking about the relative light weight of the new video-capable DSLR cams ...

... got me wondering if anyone has experimented with using 2 ultra-quiet PC cooling fans -- mounted perpendicular to each other -- mounted on a short shaft connected to a simple hand-held gymbal head to create a low-cost cam stabilizer?

I believe these fans run on low-voltage DC; if so they could be battery powered. And some of them are _very_ quiet, such as:
http://www.acousticpc.com/noctua_nf-s12b_flx_120mm_silent_pc_fan.html

There may even be quieter fans available. Just curious if anyone's tried something like this?
 
"Quite" isn´t quite quite. ;-)

When they say its quite they mean the blades.
But you dont want spinning blades on your cam.
Giros need wights. The problem is that the motors are not quite.
Maybe no problem if you dont use on board mics.

Usualy gyros are in a housing. preferably in a vacuum housing.

You could try aluminum discs instead of the blades, but they need to be made on a lathe.

Frank
 
Your comments are appreciated, Frank.

Concerning acoustic sound levels, 2 relatively quiet PC cooling-type fans might be no problem when shooting in many outdoor locations. Sound from the fans might be masked by the ambient background noise. This would also be true in many indoor locations if air conditioners or other machinery can't be turned off.

And as you say, noise from the fans may be a moot point if the microphone is some distance from the cam & stabilizer rig -- or if shooting MOS.

One aspect of this idea which intrigues me is the possible mechanical damping affect from the fans' "thrust": In addition to whatever gyro stabilization effect results from the fans spinning blades & motor, the fans might also slightly resist moving in the direction of the air they're blowing. For this reason _4_ fans -- mounted in 2 perpendicular groups of opposite-facing pairs -- might work better than 2 fans alone (of course this would also increase the fan noise level & power consumption).

Unfortunately I don't have a gimbal cam mount to test this approach; I'm thinking something like this might work as a platform for experimenting:
http://glidecam.com/product-hd-series.php
 
I've experimented with this quite a bit, and can tell you that the RPM of those motors when loaded with proper weight is nowhere near the 20k+ rpm you need to make it work right. Even if it was fast enough, it would be so loud as to be unusable.

Also- the gyros need to be mechanically linked in a way the cancels their natural movement (in respect to the earth's movement)

Also, you need bearings at each side of the shaft for stability. and SUPER awesome bearings at that.

It's a very very tough project and IMO, best left to the pros, like the kenyon labs gyros. Sorry to rain on your parade, but I wasted $400+ to make one of these and totally failed.
 
Hi Tim: Thanks for your informative post!

Question: When you say "those motors" are you referring to PC cooling fans (with their built-in motors) or small electric motors that don't have fan blades?
 
Yeah, I was talking about cooling fan motors. Usually they spin from 1k to 3k rpm. With a load, I'm sure it would be much less.

I had some aluminum gyros custom machined for me. I thought that if you take two gyros spinning in the opposite direction and lined up on the same axis, they would cancel out each others' Precession and be a stabilizing force on whatever it was attached to... boy was I wrong. They canceled each others force out completely, so it did nothing! I believe that's exactly what your '4 fans in a cube' would do as well, and even worse, since it would depend on thrust from the air-flow, it would move differently for every gust of wind. Nice idea though.

Now- Who would like to buy some lovely machined gyros? :)
 
I've done a little bit of research down this path, seen prototypes that work, and believe that low-end gyro stabilization is possible...however:

Two spinning masses a gyro does not make ;)

The problem is that even if you got the amount of spin necessary to slightly stabilize, you would slowly gain some drift in the system just from the fact that you prolly couldn't get it lined up perfectly and control it properly as well as just basic jarring motions setting it off the position you want.

So what is the solution?

You need dedicated gyro and servo motors from the RC world.

The system I've proposed in the past has been a combination of a physical and gyro controlled system....much like a melding of the first versions of the MK-V and the current system(except not just in one direction). In this way large amounts of drift can be handled by gravity while fine tuning can be handled by the RC gyro/servo's.

It's possible, but I myself have so many projects I want to undertake, and not enough money to purchase parts and time for all of them....maybe someone out there has the time/money and can experiment?
 
I've been experimenting with homebuilt gyros over the years, and am now looking for a surplus (probably Russian) military gyro as all my experiments were bitter failures.

Or I may suddenly come into $4K to buy a Kenlabs gyro (I can dream...)

If anyone's interested, the closest thing I got to working were a couple 10K rpm hard drives -- but just barely.
 
Interesting thread to me because it reminds me of early experimenting I did years ago when I started working on gyros for handheld camera stabilization. I started with brushless dc cooling fan motors by removing the fan blades (because all they do is waste energy) and replacing them with flywheels.

Here's some information you might find helpful.

#1. First, most seem to be missing a basic piece of information about gyroscopic effect. A spinning flywheel has the characteristic that if you try to apply a torque to its axis of rotation, it does not resist or act against the torque like you might think. In fact it makes a very nice right angle torque converter. Only if you mount the flywheel in a gimbal so that it can precess or rotate on an axis perpendicular to the axis you want to stabilize will you get a reaction against an applied torque in the stabilization axis. So if you build one single gimbal mounted motor with a flywheel you can get some gyro-dynamic resistance to torque in one axis (in theory you could build a double axis gimbal single flywheel gyro and get two axes of stabilization, but there are practical limitations and something called gyro lock that makes this not suitable). If you want to stabilize two axes, say roll and tilt, then you need two gimbal mounted gyros. If this all seems too complicated, I’ll just summarize by saying a rigidly mounted motor/flywheel or any number of them arranged in any fashion won’t do anything to make a stabilized platform period.

#2 The amount of gyroscopic effect you get with a flywheel is proportional to the flywheel’s geometry and mass density which determine its MOMENT OF INERTIA which when combined with its speed of rotation then determines its ANGULAR MOMENTUM. The gyroscopic effect of a flywheel is directly proportional to its angular momentum. In other words you can get the same gyro effect from a very light flywheel with small moment of inertia and spin it very fast, or make a very heavy flywheel and spin it slower. Kenyon gyros use this principle in their model steps. The KS2 uses a brass flywheel while the KS4 with the same overall dimensions and running at the same speed gets about twice as much gyro effect by substituting a tungsten flywheel which has about twice the density of brass.

#3 It actually takes a lot more gyro angular momentum than you might think to make a suitable camera stabilizer. This fooled me from the start. In part it is because of the need for the gyro to precess as mentioned above and as the flywheel precesses it goes off axis from the acting torque which reduces its effect, along with the fact that precessing needs to be minimized or at least limited for other reasons. I won’t go into detail but just say that cooling fan motors with flywheels and speeds that won’t kill the bearings are perhaps one or two orders of magnitude less than you need! Think flywheels weighing half a pound (and up) spinning at perhaps 18,000 rpm as a minimum. That’s similar to the smallest Kenyon KS2.

#4 Getting efficient small motors to spin flywheels with lots of angular momentum is a challenge for several reasons. Not the least of which is that the flywheel with sufficient angular momentum must be very precisely balanced or the vibration will destroy the bearings not to mention have a lot of shake.

#5 Finally, you have to be very careful about where the axes of rotation of one or more gimbaled gyros are in space relative to the camera or stabilizer or you get unwanted gyro effects. I’ll just mention for example, an incorrectly positioned gyro intended to correct roll could also cause panning torque.

Good luck.

Rich Greb
www.camotionllc.com
 
Hi Rich: Thank you for this information. Very interesting! I've bookmarked your website for future reference. Also a big thanks to everyone else who commented here, too!
 
You must log in or register to reply here.
Back
Top