Universal, 2 Gyro Image Stabilizer

This image stabilizer can be used with any lens and camera. It works the in same way as the Hubble telescope keeps pointed to the same object during multi day exposures.

This stabilizer can be used succesfully with moderately long exposures and moderately long focal lengths.

Needed: 2 discarded harddisks (HDs)
Some old discarded computer, or parts
The part in the old computer which holds floppies en HDs at a 90 degree angle...
A woden box or plywood etc..
A hand grip
One or two strips of aluminum
A camera screw
3 or 4 car USB phone chargers
A 12 V power source (lead acid cell, a discarded NiCd cell, or (rechargeable) batteries)
Some rubber washers and a piece of inner tire
Contact glue

Your camera

Costs: something between E 0.00 and E 50.00 (my costs: E 15.-)

Time to build:a few days, including some shopping...

Tools: Simple hand tools, drill, soldering gear.

Update: look at my single Gyro stabilizer: www.instructables.com/id/Single-HD-Gyro-Image-stabilizer/

Most hard disks spin at 5400, 7200 or 10.000 RPM. The rotating parts have a considerable mass, and are very well centered and balanced. Old HDs with storage space below ca. 10 Gb can be obtained very cheaply, or even for free.

The spinning HDs working as gyroscopes in the horizontal and vertical plane (X and Y) can almost completely prevent motion blur.

When a long exposure, or tele picture is taken by hand, motion blur occurs in a combination of horizontal and vertical (X and Y axis) shaking; not so much in the back and forth (Z axis) direction.

The spinning mass in the HDs steadies the camera.

2 HDs are mounted at a 90 degree angle. The bolts to fasten are non-metric:

In Europe, the threads of these case screws are totally incompatible with anything, so the only way to fasten the HDs is to use the existing case screws from a discarded computer, which means they can only be mounted on a metal sheet or strip.

(Case screws. These screws are six-gauge wire with 32 threads per inch American National Coarse Thread (UNC) machine screws that are cut to accept a both a Phillips No. 2 screwdriver and a 1/4 inch hex driver and are 5/16 inch long.) Wikipedia.

Of course, the prime candidate for this is the disk holder in the discarded computer: It has all the screw holes already in the right place.
This way, 2 strips have to be cut from these parts, and have to be mounted squarely.

I was not able to find an old computer before the deadline (contest!), so I mounted both on a wide strip of aluminum, 2mm thick. This strip is mounted on the lower part of the enclosure.

The HDs: nothing inside is changed. I had some old 5400 RPM HDs laying around, although having a small memory size (2.1 and 4.3 Gb), they still work fine.

They still can be used as 'image tanks'; dual use. Put rubber spacers between HDs and the mount to get rid of any high frequency vibrations produced by the HDs.

Note: Taking these HDs into the field, the data they contain might not survive the rugged environment, or rough treatment. Shock damage might cause loss of data.

A poplar box of 30 x 30 cm was cut into the proper sizes to form the enclosure: although not water tight, it does protect the HDs against shorting and raindrops. Poplar and willow wood is ideal for prototyping: very soft, and almost no grain.

The pictures show how all parts are put together.

A 3 mm aluminum strip was bent at a square angle and put on the top part to hold the camera. It holds 3 holes, for different cameras. A slit might have weakened the strip to the point that vibrations from the HDs would have been amplified.

HDs need voltages of 12 and 5 DC. In this design, only one power source, of 12 V is required. 3- 4 cheap USB phone chargers convert 12 V into 5V. According to specs in the wrapping, their output is 400 mA, so at least 3 are needed in parallel. Molex connectors from a discarded computer power supply take the power to the HDs. It seems the 12 V lead is used for rotation, while the 5V lead is used for the arm: movement, reading, writing. Both leads are need to make a HD spin.

If the stabilizer is only used occasionally, 9V + 2 x 1.5 V batteries can provide power. For continuous use, or for video, a more powerful solution is needed, like a small lead-acid cell. This cell can be mounted on a belt, with wiring to the device.

The stabilizer needs a switch to be turned on and off.

I tried to solder the new wires to the small circuit boards. However they appeared to be very heat sensitive: the copper conduits came apart from the board while desoldering!

A new charger was bought; this time soldering was done more carefully! The wired circuit boards were glued to a piece of wood, which was fitted in the top part.

The space kept for the electronics was almost too small: it required some fitting to get it all in.

How much improvement is possible with this basic design?

All pictures made with a Canon SX110 IS, with maximum: 10x zoom (36 - 360 mm, if it was 35 mm format), mounted on the stabilizer of this instructable..

The exposure time is 1/15 sec: an impossible exposure time to take tele photo pictures by hand.

Pic 1 is made with NO image stabilization.

Pic 2 is made with its INTERNAL image stabilizer on , and external one OFF

Pic 3 is made with only the Gyroscopic image stabilizer (this instructable) ON, and the internal one off..

Pic 4 is made with both internal and external stabilizer on

All pics were made during the same conditions: early twilight, whole session less than 10 min.

It seems that my stabilizer outperforms the stabilizer inside the camera, and that when both stabilizers are on, results are even better!!!

Removing the platters and putting a steel or brass disk instead. It would require removing the arm and modifying electronics.

Of course another high speed DC motor with heavy disk would work as well...

Oct 8th: A single gyro stabilizer is finished: look at the instructable: www.instructables.com/id/Single-HD-Gyro-Image-stabilizer/

On stronger motors: CD/DVD and HDD spindle motors are being hacked by the RC model plane community. With thicker wiring and replacing the ceramic magnet ring with Neodymium magnets they seem to reach up to a whopping 400 W output. Machining of a new rotor (bell) and controller ('esc') is required +a high output battery pack (LiPo), which would make a gyro project no longer low budget nor fast to assemble. It could provide another dramatic reduction in size and weight though. Link:www.flyelectric.ukgateway.net/machin.htm