Hey Mark, I'd be up to giving it a try. Yes there is/was someone doing it- a guy Jim remade them and sold them as Lakeland Ignitions, he then gave up a couple of years ago and Ricks Motorsport took it on & put the price up. They all use a different casing rather than the 2 originals though, which some people don't like. I still quite like the HEI unit idea, I found a couple of old links to that on Facebook if you're on there.
According to these schematics and some others I found for the early rocketbike, it is a clugey CDI that pumps a capacitor up with 400v that is manufactured by a couple of power transistors oscillating a push-pull circuit, then an SCR fires the capacitor into a pulse-style coil upon the magnet triggers' command.
It is very much like the ones found miniaturized (to run at 20 volts instead of 400 volts) in the CB650, 1979 style. (!) Those later ones, in various versions, appeared in many 1980s-2000s era bikes. They are about as big as your thumb. Their largest Achilles Heel is that they need really clean electrical power to operate right, so the bikes get brushed alternators in order to make lots of quiet power - and then, you have to replace those brushes. which I find offensive to a reliable motorcycle...
In that era (1969) the transistors we had that could do this sort of pumping and oscillating at the same time were few, far between, and quite expensive: today it would be like building units costing several hundreds of $$ just in parts (which might explain why the aftermarket ones cost so much?). It's kind of reassuring to discover that what I thought about them was correct: they were just far enough ahead of their time to not be a mass-produce-able gizmo in those days. The transistors were running at max all the time, and when they got hot and the high voltage drooped a little bit from it, the SCR wouldn't shut off after firing, which would make one of the 2 power transistors effectively shorted straight to ground. This would then overheat it and short it out in a few seconds. In addition, the power transistors were likely each characterized from hundreds (or thousands) of similar parts, picking only the cream of the crop to build those units, and the rest were sold for other purposes (a very common practice in Japanese electronics of those days).
So, what to do?
First I think I'll start looking about for some better, modern transistors for the oscillator: if they could withstand shorting for a few seconds then a 'safety circuit' could be added that would just it down and signal "Overheated". But better yet, if the oscillator can be built robustly enough without a heavy transformer the high voltage (400 volts) supply could just pump all 3 triggers (like the original?) and each one that fires could be used to force off the previous one so no multi-firing (and resultant heating) could occur. This would make it a constant-current device, too.
How much power does the Blue Streak's alternator make? I can find the spec for the later versions (the points-fired ones) but not for the early Mach III. That will set the current limitations for the design.
