>Rich says:
> >>Now, when two capacitors are placed in series ACROSS DC, let
>alone AC, the
>> >>voltage divides inversely proportional to the capacitance.
>>
>>True enough. However, you are trying to fly on an assumption.
>
>Which assumption?
That the capacitances are substantially different.
>Your assumption is that the voltage will divide unequally.
Without equalization, the 250v reverse surge happily divides unequally.
Approx. 101v is dropped acrosss the 101v diode and approx. 149v is
dropped across the 190v diode. Since no withdtanding voltage capability
is exceeded, there is no diode failure. Add the two 10k-ohm "equalizer"
resistors and this is undoubtedly no longer the case.
>However, if the diodes have equal capacities, then the 250 volts will divide
>to give 125 volts across each diode. If the capacitances aren't equal, you
>don't know how the volts will divide.
At 50Hz, with 10k-ohm "equalizer" resistors, the 101v diode is subjected
to more than 101v no matter what its capacitance happens to be. Remove
the equalizer resistors, and the piv capability is not exceeded. It is
my opinion that the Handbook's Chap 11 author, (who wrote about "damaging
current" from the equalizer resistors) knew what he was talking about.
Presumably, diode stack manufacturers know this as well. They abandoned
so-called equalizers at least two decades ago.
>
>Ian G3SEK says:
>
>> but so long as the diodes are reasonably well matched, eg full-spec
>> brand-name diodes from the same
>>bandolier, we should be pretty safe.
>
>I said
>
>>If you have branded diodes from the same date code, and/or guaranteed
>> controlled avalanche characteristics,
>
>
>which is I guess, pretty much the same thing.
>
>Having had 21 years in the semiconductor industry, I'm probably rather more
>cynical about it than most people, so I don't place too much trust in non
>specified or non guaranteed parameters - and not much trust in guaranteed
>ones, either.
Amen. The only piv that one should treasure is what one actually
measures.
>
>Gerald says:
>
>>If one designs a circuit so poorly that the equalized diodes
>>do not have a safety factor of at least 4 for just such an occasion
>>and did nothing on the mains input to protect from over voltage
>>they get to go back to the drawing board and forget about a raise
>>that year.
>
>I don't know about a safety factor of 4, but the concept I agree with. But
>what's this 'raise' business? Don't see too many of them around!
>
With so called "equalizers" a safety factor of 4x makes sense to me.
>
cheers, Peter
--
FAQ on WWW: http://www.contesting.com/ampsfaq.html
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
Search: http://www.contesting.com/km9p/search.htm
|