Line which is as slippery and thin as Dyneema, is almost impossible to handle with bare hands and even with a gloved hand, tensions more than a few kilograms are difficult to manage. Throw in very long line lengths and high tensions and some mechanical means of controlling line in and out are essential. The winch with a capstan and reel is the machine for the job. In its simplest form the winch or windlass is just a reel with a shaft and winding knob on it'd side. This is represented by the fixed fishing reel and the garden hose reel. Long line lengths, if wound in directly to a storage reel, accumulate tremendous compression on the reel centre. This may result in implosion of the reel and breakage of the reel sides. I discovered that after I building my first big reel (see below left). I also used many large electrician's reels in my first 18 months of exploratory flights over 1,000 ft. I soon tired of hand winding over 2,000 ft. of line although it developed the biceps!

Designing and building a winch is not difficult for a person with a modicum of mechanical knowledge, the ability to draw simple designs and basic tools such as a hammer, drill, screw drivers and a set of spanners. I was fortunate in that I studied mechanical engineering and engineering drafting although I never completed my degree. The design brief was to build a robust machine that could stand up to the rigors of outdoor use, transport on a trailer and was affordable with a small budget. Fortunately I could see many good commercial fishing, industrial and mining winches on the Internet so I just need to transfer the principles of these designs and scale them down to a version that could control 15 km of line at tension of about 200 lbs. The design needed to provide enough start-up torque to counter maximum line pull and enough line speed to keep a kite aloft in calm conditions. These parameters need to be measured. The line tension was measured flying one of our record attempt kites in 15 - 30 knot gusty winds. I measured 79 lbs maximum tension. I also applied some known values to flat surfaces with Cda and surface area. Maximum line tension was calculated as 160 lbs in 40 knot winds. In fact 120 lbs has been recorded in 2012 with 40 kt winds. The minimum wind speed required was measured as 5 knots or 2.8 m/s. The belt and pulley drive system gave reduction ration of 36:1. The motor's stock speed was 2850 rpm producing
80 meters per minute, 4.8 kph or near to 2.4 knots line speed. We needed double that but that was to come with the later winch version. This rebuild was forced by the failure of the first electric motor during the October 2005 incident. The earlier version was powered by a 1/2 hp single phase AC motor with reversing switch. By the second series of attempts in October 2005, the motor proved to be inadequate.

The Teco FM50 inverter is inside a custom made control box. The FM50 uses clever robust electronics in conjunction with a processor to provide a wide range of speed and torque characteristics to an AC motor. The motor can spin from a few revolutions per second up to 6000 rpm, over double its normal speed while maintaining good torque. It is not recommended to run a motor at 6,000 rpm (double its normal speed) for more than a minute at a time. I have run the motor at 4,000 rpm for extended periods without problem. It has forward, reverse and intelligent thermal protection which virtually eliminates damage from overload with temperature sensors in the motor giving feedback. I have also incorporate a forward/reverse control switch for the line layering motor having abandoned the auto reversing switch because of dust and grit striction in the toggle switch cause malfunction. It requires some diligence to observe the travel of the line laying carriage when it reaches each side of the reel. It probably is possible to use an optical sensor and a small circuit board to control the forward and reverse function of the line laying motor.

The winch system needed to not only produce the required torque and output speed but needed to run reliably every day, all day. For that we also needed a reliable generator as there is no on field mains power supply. The original generator was a Chinese made Yamaha copy and the first one failed, needing to be replaced under warranty. The replacement ran fine for 2 record attempts but unfortunately was damaged in the trailer during a trip to the Eastern suburbs for testing. This generator and electric motor combination proved marginal in power output anyway so the replacement generator was a good price and also more than double the power at 5.5 hp. It is able to provide plenty of grunt for the winch as well as laptop, fans, floodlights and electric jugs. It is capable of running a decent welder which I hope to never need.
The control for the original electric motor was just a on/off/reversing switch although this feature needed to be wired by a motor specialist. Without speed control, the motor is limited in its ability to finely attune kite and wind speed. It also requires excessive motor on-off cycles, stressing the internal motor control and the control switch. After the first motor burned out I upgraded with a more powerful TECO motor and added a FM50 TECO inverter/controller. The motor is 1 hp or 0.74 kW. The original TECO motor was 0.5 hp and fixed speed. The inverter has a programmable interface which can vary the torque curve at various speeds and when slowing down amongst a host of other parameters. It enable the standard 2,850 rpm to be doubled to 6,000 rpm for up to a minute. Normally I limit the over speed to 4,350 rpm or 50% over the standard of 2850 rpm. This over speed capability is important as it gives me the ability to give the kite a quick boost up, analogous to giving the line a tug. The kite can also be sustained if the wind drops to zero. This system is more than capable of performing all the intended roles except for high speed freewheeling to alleviate the situation that occurred on September 30th prior to line break

To fly a train of 12.34 sq. metre DT Deltas would require a winch capable of hauling 500 lbs of tension or 4 times the current load. This would use the same winch system but a significant upgrade in the motor size, chain drive for the motor to intermediate pulley and to the capstan. The belt drive to the slipper clutch on the storage reel could be retained. The current motors are 0.72 Kw and 1.1 Kw (1 and 1.5 HP. A kite train would need a 5 Kw (7.5 HP) motor at least. The inverter/motor controller would need to be upgraded. The trailer, being a light weight box trailer, may need to be tied down to the ground. The winch speed at times was adequate but required spinning the motor to maximum RPM from time to time. A more powerful motor could tolerate higher gearing to produce more line speed and maintain strong torque. A line speed of 20 kph or 6 M/S would enable strong counter winching to pull long line lengths through "dead" layers.

An overhead view of the trailer winch system. It is powered by a 1.1 Kw AC motor driven by a variable speed/inverter/controller and supplied by 5.5 HP 240 volt generator.
The frame is heavy steel rectangular members. Shafts are 25 mm keyed. Large pulleys are 450 mm aluminium and small pulleys are usually 75 mm aluminium. They may be changed at the motor and intermediate shaft to alter the gearing from 38:1 to up to 50:1. The capstan is a steel wheel from a 1984 Ford Falcon. The flat section is exactly 1000 mm in circumference, a very handy coincidence.
There are two methods of measuring the line out, both measure the capstan's revolutions. One is a distance wheel used for measuring land distances and the other is a digital counter taking signal from a magnet glued to the capstan side. The analogue counter has proved to be 99.9% accurate and tallies almost exactly with the digital counter. The line wraps around the capstan 16 times so that the low tension is less than 1 kg. There is a simple formula which includes the friction coefficient of Dyneema on steel which is 0.2, the high tension and the number of wraps. Here are some values for maximum tension of 54 kg:
Wraps    Output tension kg
1                42.2
2                33.76
3                27.07
4