Wilbert Knol wrote:
>
>> This keeps on coming around. OE5JFL now has an English-language version
>> of his thyristor crowbar on his own site, with readable schematics:
>>
>> http://www.qsl.net/oe5jfl/flashover.htm
>>
>> That solution is better than a glitch resistor, but it also requires a
>> lot more effort.
>
>
>I had a look at the OE5JFL crowbar. It looks like a good idea. But I
>would still use the glitch resistor.
>
The 10 ohm resistor in the negative return has many of the functions of
the glitch resistor, along with the inductor in the B+ lead. Those two
components limit the peak surge current, and are also where the stored
energy from the smoothing caps ends up.
>The cascaded thyristor network looks a bit dubious though. The ones
>slowest to fire could fail, due to having to block excessive voltage.
>
>I would put one thyristor across each electrolytic/bleeder. That way,
>each cap gets discharged individually, and there is no risk of zapping
>the thyristors.
>
That isn't actually necessary.
All the thyristors in the cascade are paralleled by high-value
equalizing resistors, so that takes care of the "off" situation.
When triggered, each thyristor has its own gate drive through an opto-
coupler, but still they can never all fire at exactly the same moment.
However, that's no a problem because thyristors will trigger quite
happily due to a voltage step. What happens is that the dV/dt step at
the anode is coupled to the gate through the device's internal anode-
gate capacitance, and that produces gate current. The device then fires
in the normal way. (Usually this is an unwanted mode, because it can
cause premature firing due to voltage spikes in AC applications. The
detailed data sheets often specify the maximum dV/dt that the device can
stand without self-triggering.)
In this cascade circuit, if the later thyristors are not triggered by
their own gate drive, they will fire due to the dV/dt steps caused by
the earlier ones in the cascade. If the current is limited (which it is
in this circuit) that mode of firing is *not* a "failure" mode, and the
thyristor recovers in the normal way when the voltage is removed.
Many versions of OE5JFL's circuit have been built, and it does work.
He doesn't mention the ratings of the thyristors, but in the original
they were 1200V/8A (that's a continuous current rating - this circuit
uses the one-time surge rating which is always much higher).
International Rectifier now do a range of 1600V 30A plastic-package
thyristors for about $6, which should be just right for this
application.
>Thanks for mentioning the URL.
>
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
|