Kite Aerial Photography

So, after a season flying with Rig02, where to now? Rig01 and Rig03 were tiny webcams, for the second digital camera rig the primary aim was to stay with a 500-600g rig but increase the resolution to 5Mpixel.

Over the winter of 2003/2004 the camera was chosen and the rig constructed.

The camera had to be 5Mpixels, 200grams, Aperture priority, and have some way of setting the shutter without a servo. Another must was that camera could be woken from sleep by pressing the shutter, this saves batteries! Operation from 5V to share batteries with the servo receiver was also a want, although in the end this wasn't possible. The camera also had to get a good write-up from the reviewers!

A story in itself the Pentax 550 was chosen with it's interesting infra-red shutter release saving servos and weight. 

Pentax provide an infra-red instant shutter release suited to KAP (as well as a 3 second delay mode suited to self-portrait). Releasing the shutter by IR reduces rig complexity and makes mounting the camera in the rig much simpler and quicker with fewer connections required. Note that the IR link is only used for the last 2 inches, from decoder board to camera, the signal still travels via radio from the ground to the rig. In fact the whole functionality of the remote control was built into the micro-controller chip (a Microchip 16F84A) which controls the rig. 

I was also attracted to Olivier Humez altimeter design for KAP, and with Olivier's help this was designed in, along with various other features including:

• Full PAN and TILT control of the camera,
• Control for the IR shutter release, and zoom,
• Video downlink of the camera viewfinder,
• Automatic switching to switch off the video transmitter and save batteries when the camera is sleeping and the downlink is not required,
• Telemetry height, temperature, camera and battery condition information:
  • Available as a video downlink (shown right), and
  • Logged to a memory chip for analysis later.
• To save on development time, uses existing Ground Equipment.
  

Construction had progressed through the winter of 2003/2004, by January 2004 the two micro-controllers that are used (One for the Altimeter, and one for all other functions) had been programmed and a breadboard was fully working. 

The picture shows the circuitry as designed for testing. The circuit is on the copper board on the right. Beside it on the stool is Rig02, which is providing the Servo RX, Servo's, and Downlink video and telemetry transmitter. The whole thing is self contained, although the camera is sitting a little way off in case the servos lose control and knock it onto the ground!

All the circuitry involved (i.e. the copper board shown above) was laid-out on a printed circuit board (shown right) to make it compact, rugged, and ready for flight.

The altimeter A/D and PIC are top left, with the video I/O and switching top right. Along the bottom left to right is the Control PIC, +5V regulator and Video Transmitter Power Switch, finally the 6V to 9V DC-DC converter to power the Video Transmitter.

By the end of February all this had been field tested without the mechanical parts, as shown on the left. All the components are laid out on the plastic tray so they can be easily carried across the test field.

Picture Peter McCulloch

This ensures that the mechanical construction and balancing of the rig can be performed knowing that there will not be any further electrical changes that may upset the balance.

On the 28^th February the rig was balanced, and mechanical construction was complete. RIG04 had it's maiden flight test at my usual test site - Cairnpapple Hill. Weight 625g (1lb 6oz) all up (Camera 210g, Camera Battery 40g, Rig 335g, Rig Battery 40g) - more than estimated mainly due to using two batteries. The image right is the maiden flight, click for a closer view of the finished rig.

Picture Hamish Gentles

The shot shows the rig connected to the picavet, with the battery about to be connected and the camera about to be switched on. 

Construction is a standard double-U. The board is located under the camera, at the bottom of the rig, and the IR LED to control the shutter is mounted on the wrong side of the board, so it points up to the camera's IR receiver.

Also in this shot you can see the plastic M3 bolt used to keep the camera straight, as well as the guard plate under the lens in case of hard landings (also visible in the following shot)!

Looking down from above (right) you can see the IR receiver (beside the text Optio 550) which is immediately above the gentled IR transmitter LED. 

Although the gentled IR LED is not as powerful as a normal IR transmitter it can easily transmit the 5cm required from controller to camera at the acute angle from below the camera.

The other side of the camera lens shows the TILT servo (right) and the cables from the Printed Circuit Board (lower right) that lead up to the SERVO Receiver which is located immediately above the TILT servo.

Also on this side of the camera lens there is a connector for the tiny 2g Logger or Memory Module. This can log over 1000 events, either triggered by time or the camera shutter.

This allows a history of the flight with data recorded every 1 to 30 seconds, plus every time the shutter is pressed.

Once recorded on the rig the Logger is downloaded into a PC for analysis.

This module is plugged into a PC Parallel port (via an adaptor) to allow the data to be analysed after the flight.

The adaptor is self powered and is built into a standard parallel port hood for ease of use. Standard software is used to download the data into the PC, and an Excel spreadsheet used to translate the binary data into usable information.

Detail of the top of the rig and the PAN servo. Most servos only move through 90^o. There are various ways round this. The method employed here is that the inside of the servo has been modified to remove the end-stops and the feedback potentiometer.

The potentiometer is replaced with one located externally (left of picture), but is connected via a 4:1 gearing arrangement. This fools the servo into thinking it has turned 90^o when it has turned 360^o.

So what were the results of the first test...


Performance was excellent on the first flight, well worth the extra ground testing that had been done over the Winter, only a couple of minor adjustments were required to the micro-controllers programs and the mechanics after the flight.

With some great light for the first flight, even though the subject was familiar, we got some acceptable results including this shot 18m above ground, using a Flowform 16, at only 1^oC!  Also for technophobes:  2592x1944pixels, F2.8, 1/491, EV-0.3, focus infinity, 38mm (SLR effective).

The logger allows height and temperature to be recorded for each picture and at regular intervals during the flight. This allows a profile of the flight to be recorded as shown right with height. The above picture is marked near the end of the flight with a red dot. The drop-outs in height show a rather marginal day for wind. Good enough for a test flight over open ground where a slow descent wasn't disastrous!

This illustrates another KAP principle - the camera for this shoot averaged only 25m above ground, you don't need 100's of metres of line to take good pictures!

The Good Bits:

• Extensive field testing got a "works first time" result. It took several months of adjustments to get RIG02 fully debugged to the same stage that RIG04 first flew at.
• The IR camera link means there is only 2 connections for the camera - tripod mount and video out. This makes the connection of the camera much easier.

The Bad Bits:

• Using 2 batteries has added 40g which is disappointing, on the positive side the battery life is excellent! The plan had been to use the SERVO batteries to power the camera, but unfortunately when you power the camera from an external power source it wont go into sleep mode. This uses more battery power and defeats the purpose of sharing batteries in the first place.
• The construction of the Pentax 550 means you can't get the memory card or battery out of the camera when it is in the rig - this makes this rig slightly more awkward than the previous Nikon 775 rig with its side entry memory card.
• The aerials are still ungainly. Work was done to try to improve the aerial arrangement, but as always this is a compromise. The 470MHz Radio Amateur transmitter transmits 100mW for only 60mA input current giving good battery usage and excellent ground reception. This solution means that 10x the power is being transmitted from the rig than with 2.4GHz downlinks, to reduce interference the aerials are physically separated, other solutions were tried and failed.

Want to build one yourself?

I published a full set of mechanical drawings for RIG02. Whilst I didn't expect people to clone the rig I hoped that this material would give others some ideas for their own equipment. 

RIG04 is different, and more complex, with the addition of micro-controllers and PIC programming. The information below will give you some further information about the rig, however please e-mail me if you want source code or other information:

• Choosing the Camera,
• Electrical Block Diagram, Schematic and Mechanical Drawings ,
• PC Download Logger I²C Memory Circuit and Picture,

James Gentles March 2004 HOME PAGE KAP HOME PAGE