Soft start for power supply

A nice advantage of your approach is that you can save 2 parts (D1 and C1) in some cases if you want. Just use B+ from the device itself. AND it will work fine if the transformer includes the filament winding, without delaying the tubes heating up.

I didn't want to recommend that in the original post, as anyone could read it and attempt to do wiring on higher voltages areas (where higher in this case means above mains voltage - perhaps up to 4000 volts in some linear amplifiers). Many are qualified to do that of course, but far more are not.

Ed

In particular, higher powered equipment with mercury rectifiers needed the heaters on for some time before the HT could be applied. This was to evaporate mercury that had collected in the wrong place with the risk of a flashover.

With equipment like that it was quite normal to have a separate HT transformer, so a relay could switch the primary. The relay was controlled by a delay system which used a special delay switch that consisted of a heater and a bimetal strip inside an evacuated envelope. It plugged into a B9A base and looked exactly like a normal valve.

In equipment with a separate lower-voltage HT supply, a spare valve (such as the unused half of an ECC82) could be used with the grid strapped to the cathode and the relay coil in the anode circuit. The relay could not operate until the valve had warmed up and started passing anode current.

(Leaving aside the fact that this has already been satisfactorily resolved) I don't wish to cast nasturtiums on your design abilities, but you can't come up with a valid design for anything until you know the full circs of what you're designing for. In the case of the two PSUs in question, they're not giving an initial surge in the commonly-understood sense. I'll explain for the purposes of illustration, step-by-step, what happens.

1. Initial power-up/switch on. Indicated voltage immedilately climbs to about 1/3 of indicated Vout.
2. Hiatus of around 10 seconds.
3. Vout begins to climb at an exponential rate, surging to up to 50% above indicated Vout, then falls back to a point where the true Vout and indicated Vout match each other.

The above holds for both my HT PSUs. The low voltage PSU I'm using

Early utility DC links used ignitrons.

The big old Tek tube scopes usually had that arrangement.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

You could make an external box, with two dual banana jacks, that opens the path until you push a button. That could be a fun design.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

The typical circuit handles that, in general. The assumption in the typical circuit is that there needs to be a delay in order for the tubes to warm up before full B+ is applied. As you said, I don't know the specifics of your particular equipment, thus I posted a typical design, not something designed for your specific problem. It has nothing to do with my design abilities - but I do like nasturtiums! :-)

Anyway, it's not my design, it's in the published literature concerning soft start circuits. You might enjoy reading

Back to the circuit. As I mentioned, you can use it to switch the output on after delay time, which seems to be what you have decided after asking Liz about that (?) You can decide how to use it - or whether to use it - based on your specific equipment. It's simple and automatic.

Ed

Yes, well doing nothing is not an option because flicking the HT on and off is all very well - until I absent-mindedly leave it in the 'on' position (which is bound to happen sooner or later). In the mean time, I'm hoping a pal of mine will succeed in finding a 450V transformer in his ham stash. That way I'll avoid even having to build anything at all. :)

You don't waste the delay time, you use it to check that you really have wired up the circuit correctly and that there are no stray wires touching the HT and waiting to bite you. By the time you have done that, the valves in the PSU have warmed up, the valves in your own circuit have warmed up (no point in applying HT before they are up to temperature) and there is no further delay.