[ARC5] Building a power supply for an ARC setup

[Brian Clarke]

Hello Fred,

The BC-696 was designed for screen modulation. This means that the maximum 
Voltage on the plates is usually your dc supply. The RCA Transmitting tube 
manual says that the 1625 as an RF PA tube can stand 750 Vdc plate Voltage 
in Class AB2 ICAS service, which is pretty close to what the design is. 
While a choke input filter would give a better quality output, 900 x 0.9 = 
810 Vdc is a bit too high. If you can parallel the two secondary halves and 
bridge rectify into a capacitor input filter, 636 Vdc is very likely, a 
little lower if you run into a pi filter. You would get essentially the same 
output Voltages if you use full-wave rectification into a capacitor input or 
pi filter. The problem with capacitor input or pi filtering is the higher 
output Voltage load regulation compared with choke input. The Duncan  Amps 
suggestion of 528 Vdc seems a bit low, unless there is a very heavy bleed in 
a child-like attempt to reduce load regulation of the output. If the 
secondary dc resistance of each winding is 16 Ohm, at maximum current drain 
the Voltage droop should be 5.7 V, not the 108 V the Duncan Amps design 
suggests. Even if the choke were 100 Ohm, the additional Volt drop at full 
load would be 36 V. I am assuming you would use solid-state rectification; 
if you used tube rectifiers, you might get up to 100 V drop. This high 
Voltage drop using hollow-state rectifiers makes the output load regulation 
much higher (worse).

The main thing limiting the VA rating of the transformer is likely to be the 
core. You don't get more current from FWB rectifying two parallel windings 
compared with FWCT. The heating of the windings will be the same. In FWB, 
both windings heat at the same time; in FWCT, the windings heat on each half 
cycle alternately - so, the average heating is exactly the same, as is the 
average output current.

I would suggest a 10 H choke if you can. And if you are worried about its 
insulation after years of solitary shelf life, put it in the negative lead. 
In that position, you can use it to provide a bit of negative bias, if you 
need it. Bear in mind that the bias will vary with load current.

73 de Brian, VK2GCE.

On Monday, April 22, 2013 12:37 PM, you said:


>I have been hunting around for a HV transformer for a power
> supply specifically for a BC-696 (and a future series modulator) - and 
> last
> year happened on what looks like a half finished power supply consisting 
> of
> a Freed RPG-26 at a local hamfest.
>
> According to the specs this transformer can deliver 900v at 320 va. 
> Yikes -
> a bit TOO much voltage.
>
> However, this transformer has a nice feature in that the secondary is
> actually TWO separate secondaries - each one (I'm guessing) delivering 
> 450v
> on an 16 ohm coil.
>
> My question is:   Can I parallel the secondaries for 450V ac with more
> current?
>
> According Duncan Amps PSU Designer- that would give me around 528 volts DC
> IF I  can find a 5hy Inductor that will withstand the current load.
>
> Am I barking up the wrong tree?  I need an OT to straighten me out 
> then....
>
> Thanks for all your help in advance.
>
> Fred Cousins
>
> KD7EDW