Quote: "I started work at Mettoy Swansea in the late 70’s in a trainee management role, where I mainly worked on electronic toys e.g. Radio Control. The idea of Mettoy producing a home computer came from the now continuing trend that traditional toys, (which were Mettoys bread and butter) were falling out of favour.
Work on the new computer project ‘SAM’ (name taken from the memory support chip the 6809 uses) was started with Tony Clarke (first M.D). Initial concept was for a 16K computer, so by rights it should have been a ‘Dragon 16’. But as the launch was drawing near the Spectrum 48K was launched and 16K was thought inadequate, so the ‘Dragon 32’ was born. The Dragon name was purely being patriotic and naming it after the Welsh emblem - in addition Mettoy already had a defunct company called Dragon Data which was originally outsourcing its mainframe computer facilities. Because the original 10,000 units build plan was in progress, the first 10,000 Dragon had to have a 16K piggy board for the extra 16K.
Other design concepts that were considered were for a similar machine to the Sharp MZ80K, CBM PET and the later Amstrad CPC 464 where the Dragon would have a built in monitor and cassette deck. These ideas got dropped, as it would mean a much higher price tag to the final machine and by the fact that most people already owned a TV, the monitor was a bit redundant.
The early days of the Dragon was quite busy which involved me hand soldering he first batch of prototypes (around 20) at home, into the small hours. Unfortunately due to the time it took to create the dies for the plastic moulding of the case, these prototypes had to have cases hand cast from Dental material, which meant they weighed a ‘ton’, but it did mean examples could be sent out for review etc. This work on the prototypes also involved many trips to the Swansea Tandy (Radio Shack) to purchase TRS-80’s, accessories and software to ‘disassemble’. This bears out by the fact that the first game ever played on a Dragon was TRS-80 Quest.
The initial idea for the SAM/Dragon project was that 10,000 unit would be made during the first year which did affect many of purchasing and business plans at the time. But soon after the Dragon 32 went on sale demand quickly outstripped supply. This meant that the Dragon Data production line was running 24hrs a day, this was also a slow process due to the 24 hour soak period that every Dragon went through, where a cartridge based tool with LED’s was used to indicate the condition of the machine. The 24hr soak period was deemed necessary due to the relatively high 3% failure rate in the first 24hrs. To help combat this high failure an investment was later made in a ‘GENRAD’ automatic test machine to try and locate the failures.
Other interesting facts from the production were that the keyboards which were originally bought in from the American company Hi-Tek at £10 a go, at this time the build cost of a whole Dragon was working out around £70. During the first year of production approx. 35,000 were manufactured and sold. All the PSU production were out sourced to Flowtronics, who made most of the Dragon’s PSU’. This meant good business for Flowtronics but unfortunately it also nearly killed them off when D.Data went out of business. Another interesting little know fact about Dragon Data was the production of software for the farming community where farmers could keep track of their cows etc. The software, ‘Farm Fax’ (I think) was packaged with a Dragon and sold through a government scheme to bring farmers bang up to date with modern technology.
The 1983 move out of Swansea was in part brought about by the financial instability of the parent company. To help out this situation ‘Prutech’ (the high risk, hi-tech arm of Prudential) was involved to part finance Dragon Data and separate it from Mettoy. Hence the move out of Swansea and away from Mettoy’s home ground so that Dragon Data could also be seen to be more independent. Dragon Data’s real financial troubles began with the pending ramp up of the production of the Dragon Alpha (aka. Professional). This ‘Money Suck’ was mostly caused by the buying in of components for the pending production, which involved buying in IC components for 10,000 Alpha’s plus 600 Sony chipsets (at £60 a time!!) for the then revolutionary 3 ½” floppy drives used in the Alpha. This was all well and good but the designs for the Alpha weren’t finalised which meant many of the components already purchased where becoming redundant and ‘dead stock’ overnight. In addition investments in employees were made - it seemed - with little thought in cost. At the same time the cyclical nature of the product in terms of design upgrades and sales were not properly understood.
During this time I along with others built the new Alpha machines for ourselves - something that was achieved quite quickly without the normal day to day business constraints. After the sad downfall of Dragon Data I stayed on and moved to Spain with Eurohard and helped them set up the initial production - this was in 1985 and the factory was located in Caceres in the Extramadura part of Spain. It was a very poor region and I always remember my first day as the bus I arrived in drove over the main power cable to the factory - after this I always stopped the bus early and allowed it to bump over the cable without me on board. It was quite amazing how they purchased Dragon Data and ever made anything as they didn’t appear to have any real money either."