I remembered why I like solid state so much, I'm terrified of this bloody thing. It's a bit of deja vu, the last time I felt like this was learning to hover a large model heli. You start out thinking 'a nice harmless little hobby' and you end up thinking 'why am I standing 10 feet from a barely-controlled flying chainsaw?' Tube amplifier design - the original extreme sport.
Thanks, Kim, elliottstudio.
Armaegis, are you clairvoyant? How did you know that I look like Uncle Fester?
Everybody, please bear with me, I have to wait a month between blood tests and they won't adjust my thyroxine dose until they see the levels, right at the moment I feel too exhausted to do very much at one go, just a couple of weeks ago I was hyper. Still, at least it's not life threatening, unlike this amplifier.
-------------------------------------------------------------------------
I noticed that the first time I swapped the tubes, the lit one seemed to move. I know though, that tubes don't always show much light, although these are of identical manufacture. I checked all the tubes with an independent LV lab. supply with 1A out. They all showed a visible glow. 'Course this isn't a trick that's open to everybody, but when you're debugging something you have to use any little clue that you can get. It only occurred to me that I could do this while walking the dogs, it's best in these circumstances though not to rush at things and give yourself a chance for ideas to come to you.
Originally I drew this:-
...but I actually built this...
...which seemed a reasonable way to save 2 diodes and a slight simplification. Now I'm not so sure. I don't see any theoretical reason why it should be a problem, but I'm fairly sure that only one random LT supply at a time is working properly, so I'm going to revert to the first arrangement. Since it's all built point-to-point, it's a bit of a fiddle though.
Here's what I had to do to get enough room with this little chassis.
I'd have liked to keep the choke on the inside, but it simply wasn't possible given the other parts that needed to be fitted, and apart from slightly spoiling the simple appearance, it's actually better in electrical terms, it's further away from the OPTs and particularly the PT.
At least I didn't have to drill any more holes, I was able to use the existing ones.
The grounding system is dual star. The noisy star is located at the junction of the 220u caps. This was prewired. Wherever possible soldered joints are shrink-wrapped. Power grounds such as the CCSs are returned to the noisy star. I'm not too keen on earth-follows-signal.
Components are cable-tied to self-adhesive tie-mounts, or glued with a hot glue gun. The HT bridge rectifier is bolted to the chassis, as are the LT regulators. The 10k anode load resistors are 10W, they only dissipate 1W, but a close watch will be kept on them during the first few hours of running. The classic style of turret-board + right-angle dressing simply wasn't achievable in this space.
The quiet star is located at the output jack. Signal grounds are returned to the quiet star. A single wire joins the stars.
w
