DIY Camera stabilizer steadicam design    WSCLATER    An evolving prototype, started August 2008
   Video. Stabilizer ideas from around the world
The Basics

This is a DIY camera stabilizer in form similar to the MERLIN. I used mostly scrap or surplus materials. A piece of 1/4 inch aluminium forms the camera platform. Below that, separated by spacers, is a slotted aluminium plate to which the gimbal housing attaches, allowing front to rear balance adjustment. An aluminium torch was reconfigured to act as a handle. A curved aluminium tube screwed to the camera plate holds the washers used as counterweights. The pictures here represent experiments over a period of time, and dimensions and balance configurations are different from time to time.

For these experiments, I have been using a Panasonic NV-GS60 camera weighing a total of 600mg (1.3lbs) with the large battery attached. The whole system weighs about 3lbs. I have positioned the gimbal some distance below the camera to maximise the inertial effect of the camera's mass. The different heights of the counterbalance weights cause unequal centrifugal forces to be experienced by the system, and can cause tilt when panning.


Above, new handle in black, April 2009

Below, this picture below indicates the dimensions of the stabilizer against a 5cm grid. The two blue dots indicate the camera centre of gravity, and the centre of the universal joint.

Static & Dynamic Balance

The orange dots indicate the principle centres of mass of the camera and the counterweights. The yellow dot indicates the centre of the mass of the stabilizer with no counterweights or camera. .The green dot indicates the centre of the universal joint. The system is shown against a 5cm grid. In this setup, the camera has a mass of 500gms. The forward counterweight;s mass is 170gms, and the lower weightt's mass is 210gms. The remainder of the stabilizer (excludind the gimbal and handle) weighs 320gms, and its overall influence on the balance in this setup is relatively small, although not insignificant.

The masses of the camera and counterweights (m) and dimensions (L) are represented in the diagram. I have excluded the mass of the stabilizer framework .During panning, the centrifugal forces of m1 and m2 will cause tilt in a contrary direction of that of m3. For dynamic balance, we want (m1 L1)+(L2 m2)=m3 L3. If these forces do not balance, the stabilizer will tend to tilt during panning.

The above is a highly simplified representation of the dynamics.

Interesting Tiffen Steadicam analysis of balance dynamics here

 

 
Gimbal Design #1 August 2008
I have designed and constructed my own gimbal system. I experimented with ball joints, and initially used one which I made up myself. It was good - smooth and fairly low in friction, but I was never really satisfied with it. Eventually I had the idea of a small universal joint combined with a ball bearing set into the top of the aluminium torch which forms the handle. The bearing is a standard skateboard type - avilable quite cheaply in packs of ten. Conveniently, the head of the torch dismantles, allowing the bearing to be fitted in a very satisfactory way. The ball bearing is held in place by a threaded ring. Above the bearing I have a small wheel which will pan the camera with minimal effect on the camera's tilt. Using a finger on this wheel, I can also effectively lock rotation on the vertical axis to the handle, and contol pans by turing the handle. If the vertical shaft is off-centre in any way, the camera may display a tendency to turn when the handle is tilted, so some accuracy in construction is required. This will be less of a problem with the ball bearing at the top, ie above the universal joint.

The inner end of the universal joint is fixed to a metal disc located in the gimbal housing. The housing itself is a plastic tube from an old vacuum cleaner. Near the bottom of the housing I have two small screw heads which act as locations for finger control of the system. These are in line with the centre of the gimbal.


The edge of the panning ring has been rounded and polished. The smoothness of the wheel and its proximity to the top of the handle make it easier to apply light forces to control pans.

To the lower end of the universal joint, I fitted an internally threaded tube which takes the screw which passes through the centre of the ball bearing, The ball bearing is held between two washers, and the threaded ring holds the assembly in place. On the right is the washer which is fixed to the vertical shaft, forming a wheel which can control panning without unduly upsetting tilt.

Balancing


A number of holes in the top plate accept the standard 1/4" camera mounting screw. The gimbal housing can be slid forward or back to adjust balance. The design would benefit from having a greater number of holes, mor closely spaced.


Washers used as weights at the front of the tube.These,and the weights at the end of the tube, can be slid one side to the other to adjust balance. This is not an ideal system - making the structure of the stabilizer asymmetric could upset the dynamic balance.

Below, plastic bottle caps provide a softer finish for the weights on the tube. The weights are centralised.


Washers on the underside of the camera platform are used to set the balance.


A small lead weight can be slid forward or back along a plastic channel, to fine trim the front to rear balance.


April 2009. Balance adjuster. A plastic channel at the back of the camera platform holds two UK penny coins with a disc of lead (painted black) sandwiched between them. The surface of the coins has been polished down, revealing the ferrous metal underneath. These plastic strips are used for holding documents together. The disc can be rolled from side to side to finely adjust balance. A new arrangement for the spirit level, using 13 Amp fuse clips.

Below, A similar design for agjusting front to back balance, this time using the outer ring of a skateboard bearing, filled with lead.

The whole balancing setup needs rationalising. What the stabiliser needs is two screw adjustments to move the camera position or the gimbal position, but this would require some fundamental redesign of the whole system.
Gimbal Design #2 March 2009


Design #2 is similar in concept to #1. The prinicpal difference is the use of a TRAXXAS joint. In some respects, the centre of the TRAXXAS joint is superior to the red joint. The metal centre is more highly polished, and has a smaller surface area of cantact. The yokes of the Traxxas joint are also stronger laterally.


The TRAXXAS 1951 universal joint, or half-shaft, and skateboard ball bearings. Two pairs of internally and externally splined shafts in a pack. 4mm and 8mm screws will cut a thread into these shafts.


Assembly of the upper half of the universal joint. The yoke is screwed to a metal washer. A collar made from part of one of the internally splined shafts is used to help secure the upper yoke to washer which is located in the gimbal housing. I have shaved the edges of the yokes to increase the angle of movement.


The upper yoke in position


Completed gimbal housing


Roughly cut, the bottom yoke is connected to the ball bearing with an 8mm screw. The washer below the yoke will be used sometimes for panning.

Gimbal Design #3 March 2009

This design is similar to the earlier designs, but has two ball bearings - one above, and one below the TRAXXAS universal joint. The top of the joint has to be accurately centred in the bearing, otherwise the balance of the system can be upset if the joint turns. There are some attractive advantages in this design, especially when the handle is off the vertical. However it is not possible to control panning with the wheel used in the earlier designs.

Gimbal modification. April 2009.
Friction coupled slip ring for panning

This coupling allows the panning ring to 'slip', allowing smoother pans. The vertical shaft of the gimbal is fixed to the centre ring of a ball bearing, and the outer ring is actuated by hand. Three short sections of soft plastic sit between tthe inner and the outer rings, and provide the necessary friction. The plastic grips the outer ring, and there is movement between the plastic and the inner ring. The ball cage has been removed. There is no lubricant in this design, other than a trace film of oil. Too much oil or grease produces inconsistent friction.