[Amps] Switch-mode HV power supply

[Ian White, G3SEK]

Steve Thompson wrote:
>On Monday 15 March 2004 07:45, Ian White, G3SEK wrote:
>
>> When I started this discussion, I should have said that one of my
>> criteria for choosing a PA is that it should give a significant increase
>> above 100W. It really doesn't seem worth the trouble of adding a PA to
>> your 100W transceiver - and it *certainly* isn't worth the trouble of
>> taking a PA on a DXpedition - unless it is capable of at least say 400W
>> output, and preferably much more.
>>
>> By rectifying and multiplying 230/240V, it is quite difficult to develop
>> much more than 1kV, and that is only enough for an amplifier in the
>> 100-200W class. The Russian page shows an example of a GI7B; and even
>> for that tube, 1kV is very low. If you tried to voltage-multiply by
>> higher factors, you'd run into severe problems with smoothing and
>> ripple.
>
>I recall an article by G4JST, which doubled 240V to get +/- 600V and used 3 x
>4CX250 to coast along at 400W. I'm sure I retained a copy, but I can't locate
>it right now. (At the time, in the UK, 250s were the cheapest/most readily
>available valve that fitted the bill).
>
I remember that one, but 1200V is extremely low for 4CX250Bs, and is 
likely to lead to linearity problems, or else a further reduction in 
output to avoid unacceptable non-linearity. Also it's impossible to use 
those tubes with the normal sockets if the cathode cannot be grounded to 
chassis.

Peter DF3KV and Roger VE3ZI have mentioned PL519s and other sweep tubes. 
Unfortunately those are now very hard to get, but they were about the 
only RF power tubes that really liked to operate at low voltage and high 
current. All the tubes that are currently easy and cheap to obtain seem 
to be much happier at 2kV or even more.


>> Also, half of the total HV is positive above ground; and half is
>> negative below ground, which creates some serious conflicts between
>> safety and good RF grounding practice. And as for the power supply
>> design... well, let's just say I wouldn't touch it!
>I agree it needs thought and care, but I don't see the problems as
>insurmountable.

Not insurmountable, but definitely requiring a lot of thought and care.

>Anyone going to the trouble of using the technique to make an
>amp that's small/light enough for specialist dx travel is likely to be aware
>of the pitfalls and design accordingly.
>
I don't think that necessarily follows...


-- 
73 from Ian G3SEK         'In Practice' columnist for RadCom (RSGB)

http://www.ifwtech.co.uk/g3sek