Since Jim's reply involves issues that are likely to be of widespread
interest, I am posting my reply to the group.
>
>-----Original Message-----
>From: 2@vc.net [mailto:2@vc.net]
>Sent: Thursday, October 18, 2001 19:08
>To: Jim Thomson
>Subject: RE: 8k ultra/ "tuned input"
>>
>>Everything else in that amp would have 8500v on it as
>>well, like their glitch R in neg lead, plate block caps, by pass
>>caps, all wiring etc. Besides 300w of bleeder resistor heat is
>>a bit much. Better to go to a simple hypersil c core setup,
>>with a huge Cap. IE: cap input filter. Better to put the weight
>>and expense of an unneeded choke into a better plate xfmr in the
>>1st place. Reg is the same, and ripple is zip with >50uf.
>>
>For RTTY and AM, the resonant choke filter is the E-ticket.
>
>#### On rtty/am/fm, who cares?
// Your Emailer's attribution marks are not correct.
> Reg is perfect, cuz u are in
>a steady state plate current senario.
// Regulation is not the reason for using a resonant-choke filter. With
a capacitor filter, the peak current is c. 10x higher on the electric
mains, so one must compensate by using heavier gauge wiring to the
circuit-breaker box. The "Plywood Box" amplifier needed #4ga. Cu to
maintain acceptable HV regulation. When tuning up the PB with a tuning
pulser, the wiring could be heard making a racket in the attic -- and
the neighbor's porchlight would blink at the tune-up pulse rate.
>###### resonant choke filters are a dead loss. Commercial
>stations use em for one reason only, and that's to eliminate
>high surge currents through the diode strings, when supply is a
>single phase type.
// The Collins Radio 208-U10 amplifier uses 3-phase 50/60 Hz power. It
has a resonant-choke filter.
>- Resonant choke supplies don't need huge output caps, as the
>resonant choke setup kills the ripple.
>
>-the ripple with 3 phase supplies is only c 5% percent, and
>that's the un-filtered dc output (vs 48-50%, ripple on a single
>phase supply).
// As I see it, with an unfiltered, single-phase DC supply, the
instantaneous output potential varies from c. 1.4x the RMS V to 0V. Is
this not 100% ripple?
>To get from 5% ripple down to <1% is a piece o
>cake with a simple cap input filter + a 3 phase diode/plate
>xfmr setup.
>
>- like John Lyles sez... you need extra LC sections thrown in
>just to get rid of the harmonics of the 120hz ripple, like
>240, 360 hz etc. The resonant choke only kills the 120hz
>ripple and not the harmonics.
// I have heard a number of signals from 1500w amplifiers that used a
resonant-choke/shunt C filter without an additional harmonic filter. I
have not yet heard any ripple harmonics.
>More L/C sections or a big
>output cap, in which case, you may as well just deep- six the
>resonant choke, and stick in a huge cap.
// The aforementioned amplifiers used c. 10uF of shunt filtering after
the resonant-choke.
> Then both static and
>dynamic reg is excellent, and no ripple or harmonics of ripple
>freq.
>
>-IE: hpersil pole pig + >50uf cap.....end of story. 6A
>diodes are cheap
>
// #4-gauge Cu wire is hardly cheap.
cheers, Jim VE7RF
- R. L. Measures, 805.386.3734,AG6K, www.vcnet.com/measures.
end
--
FAQ on WWW: http://www.contesting.com/FAQ/amps
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
|