Steve Thompson wrote:
>
>>
>>> However, it does show that the more you think about screen supplies,
>>> the more potential failure modes you see, and the more complex the
>>> design becomes to defend against them. Marv's is complex; Acom's is
>>> complex; Alpha's is complex; and G4JZQ's and mine is complex. We all
>>> did it differently, but we all agree that a handful of cheap zeners
>>> won't cut it.
>A couple of thoughts. MOVs are widely used to clamp screens in the event of
>a flashover. At 10s amps, the clamping voltage will be in the regions of
>600V for a 375V MOV. Many designs survive this ok, but I wonder how many
>times it's luck rather than design. There can be hidden side effects - for
>example, if the supply has a series element using a Nchannel FET, there's a
>big parasitic diode which will carry the 600V from source to drain.
>Similarly, a bipolar series element will suffer reverse emitter base
>breakdown and make 590ishV available at the base.
>
There doesn't seem to be any real need for an active series element to
feed an active MOSFET shunt regulator. The loop gain in the shunt part
is already very high, so the series feed can be just a plain resistor.
Part of the trick of using MOVs for this application is to use several
(they're cheap enough, and you probably had to buy 5, so why not?). Two
in parallel seem effective to protect the 1000V screen bypass in the
socket against a flashover from the anode, then one at the output of the
supply, and then another one directly across the MOSFET. Because of the
various series elements in the path of the flashover current, the MOVs
in the supply see less voltage and current than the ones out at the
socket, so they clamp at progressively lower voltages.
--
73 from Ian G3SEK Editor, 'The VHF/UHF DX Book'
'In Practice' columnist for RadCom (RSGB)
http://www.ifwtech.com/g3sek
--
FAQ on WWW: http://www.contesting.com/FAQ/amps
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
|