Look up the results for the Sierra Cup in 1983 and you'll see it was won by Walt Ghio in the 7th (10 minute) flyoff round after we'd flown 10 normal 3 minute rounds. I was second. Walt was using an extender disk on his KSB timer and I was using a Koster electronic. Virtually everybody else had dropped out at the 7 and 8 minute rounds because they were using standard 6 minute KSB timers (6 mins + time on d/t was less than 8 minutes on that day). A week earlier Lee Hines and I had whopped the rest of the field in the F1A Champagne Flyoff at the Livotto (Taft field) because we were using electronic timers and the rest weren't. Lee won because his Bauer ran for a bit over 10 mins while my Koster had a maximum settable time of 9 min 54.
The first electronic glider timers I saw were made by Ken Bauer, who used a 555 timer chip plus a 2224 programmable oscillator/divider to drive a magnetic actuator that was energised during the whole flight. I got two of these in 1982. They had a few problems: they were not waterproof and the analogue timing circuit drifted noticeably with temperature. The lack of waterproofing was a real problem because if the timer got wet it would only d/t after 10 minutes, though it did always d/t. The permanently on actuator ate batteries and it was powered by a 9v PP3 battery, which was bulky as well as being insufficiently dense to act as useful nose weight. That said, they were a fail safe design and would d/t if the battery failed as well as after a line break.
I saw the Koster prototypes in France in summer 1982 and got my hands on some of Thomas' first production batch in early 1983, in time to use one in my first D-box model. I trimmed it during that summer and then used it later that year in the Australian World Champs, the Livotto competition at Taft and the Sierra Cup. I had both Bauer timers and Kosters in the models I flew that year.
As far as I'm concerned the first really practical electronic was the Koster, which came out in early 1983. I'm still using them in zoom launch models though they have been out of production since the late 80s. I think that speaks volumes for their quality and reliability.
The Koster timer.
The Koster is a single function F1A circle tow glider timer, settable in 6 sec increments up to 9 min 54 secs and designed to be set/reset from a micro switch that detects when the tow hook is pulled forward. It has a moulded plastic face plate housing the d/t release mechanism and the electronics are potted in a waterproof compound. The timer was designed to fit into the same space as a Seelig F1A timer, which weighed about 28g, though without allowing space for the battery and wiring. The Koster weighs 20g with wiring and also requires a microswitch (2.2g) and a four or five cell NiCd battery (16 or 20g respectively) for an installed weight of 36.2 to 42.2g depending on the battery. The timer draws 80 or 300 uA depending on whether it is being driven by a 4 or 5 cell battery and requires 500 mA for 15 mS to d/t. This is a minute power requirement: it has been calculated that you could make a flight every 5 minutes for 18 hours and still use only 30% of the battery capacity.
Koster timers are designed around dumb electronics. Essentially they contain a 32 kilohertz watch crystal, a divider chain to slow this down to give 6 second 'tick' and two 74LS193 cascaded counters, each connected to a decimal encoded rotary switch. These select the d/t time in 6 second increments (0-54) on one dial and minutes (0-9) on the second dial. They are adjusted with a small screwdriver. The timer pulses a small solenoid to release the d/t lever. Its main weakness is that it is not a failsafe design: if there is not enough power in the battery to trip the solenoid the model will not d/t.
I lost one of my Koster-equipped models at the 1991 Poitou event on a lunch-time trimming flight when it failed to d/t. Of course I'd put it in a good thermal: we could still see it climbing and drifting away as we went out to fly the first round after lunch. So, although the model was fitted with a 27 MHz Vin Morgan tracker, there was little point in chasing it since it had specked out still climbing and the 27MHz Morgan bugs didn't have a huge range compared with current units. Much to my surprise the model was found, gift-wrapped and presented to me at the 2009 Poitou prizegiving so I was able to find out why it had flown away. Judging by the airframe damage the model had landed in a tree and either fallen or been knocked out of it, since the tailplane and fin were both missing and the tips were off. However, apart from that it was in remarkably good condition apart from rust coated piano wire and steel screws, so it must have been picked up fairly soon and kept in somebody's house for the next 18 years. The NiCd battery was long dead and the battery connections had corroded, but after I cleaned the contacts and plugged in a battery borrowed from a current model, the timer worked perfectly and so did the tracker. The model was less than a year old when it was lost, so it was probably a faulty battery rather than poor battery connections: at the time I knew very little about recognising when a battery was about to fail and had not yet started to keep a log of battery test results. The timer was one of the original 1983 batch of Kosters: they were the only ones to have coloured faceplates. That the timer still worked speaks volumes for the quality of its design and construction. It went straight back in my spares box to await fitting to another model.
In 2003 I bought a pair of MTK timers. One is in an M&K 'short' model and the other is for future use.
The MTK timer.
The timer weighs 9.3g and requires a 5 cell 50 mAh NiCd battery weighing 20g, a sub miniature (9.5g) RC servo to drive the tailplane and two reed switches (1g) to sense the hook state. This gives a flying weight of 40g. The timer is much more capable than the Koster. In addition to line break protection it provides an arbitrary set of tailplane movements during the bunt launch and glide, audible timer state indication and the ability to skip the bunt in the event of a muffed launch. A Palm Pilot is used to load settings into the timer before flight. The major disadvantage of this timer is that it will not do more than four or five launches on a single battery charge. As a result, my routine involves recharging the flight battery after every flight.