[Amps] HV fuse

[Ian White, G3SEK]

R. Measures wrote:

>>>
>>> What do you think about switching HV off with the vacuum relay ?
>>> Is it better than fuses/circuit breakers in the primaries of the PS
>>> transformer ?
>>
>> It doesn't need to be as fast as a vacuum relay. With conventional 
>>mains switching relays and an electronic anode current sensing 
>>circuit, a crowbar short will usually shut down the HV without blowing 
>>any fuses.
>
>••  Ian --- The issue is seemingly to prevent  damage to the 
>amplifying device from the current peak generated by the stored energy 
>in the filter caps. Would a mechanical relay be fast enough to do this?

Sorry, I now see that the original question was ambiguous, and I read it 
differently from Rich.

My answer was based on what I had written earlier, and meant: if you 
have a glitch resistor in the HV output, then you do need a relay, 
circuit breaker or fuse in the transformer primary... but it doesn't 
need to be especially fast.

But if the question was "Can you protect against a glitch by using no 
resistor, but switching off the HV with a vacuum relay?" then the answer 
is no. It isn't fast enough to protect against the initial surge out of 
the filter caps.


>> The glitch resistor protects the tube against the first current surge 
>>out of the HV capacitor, so there is no great strain on the relay.
>
>••  If the glitch-R protects the tube,  we need a HV crowbar to 
>protect ______?

It's either/or, not both.

You deleted the part that quoted Eimac: the glitch resistor is good for 
tubes up to 1500W, but above that power level a crowbar is better. 
Bulletin 17 explains why.

> //   What kind of a crowbar would be able to function in a 9000V power 
>supply?
>>
Back in 1956, RCA were talking about mercury vapour and hydrogen 
thyratrons, or ignitrons, for transmitters in the megawatt range. 
Perhaps John can tell us what the technology is now?

The nice thing about OE5JFL's circuit is that it uses a string of very 
cheap TO220 1000/1200V thyristors, and being solid-state there is no 
standby filament power. That technology is certainly good for 4kV, and 
should go higher.

The crowbar and the glitch resistor are both very fast, but they both 
need a slower follow-up breaker in the transformer primary.


-- 
73 from Ian G3SEK         'In Practice' columnist for RadCom (RSGB)
http://www.ifwtech.co.uk/g3sek