>
>Rich said;
>
>>However, with an oscilloscope, when I connected the load,
>>the output V briefly spiked downward around 50%. When I disconnected the
>>load, the output spiked upward about 50%.
>
>I'm not sure if that happens with a tuned choke
It doesn't. This is why Collins, Henry, Hughes, Continental, on and on
use resonant choke filters.
>(I don't like calling it
>'resonant' because we've seen it had better not be!) but those spikes are why
>you need a good big capacitor on the output to keep the dynamic regulation.
>
A capacitor on the output would have to humongous to correct this
problem. Resonating capacitors are typically 0.1uF to 0.2uF at c. 2.5x
the DCV out.. The resonance point is typically c. 2.45x the electric
mains frequency.
>One reason that you can see the technique used in big (5kW and up) commercial
>tx's with relatively small capacitors (10 to 20 uF) is that most of those run
>ISB type services, frequently with multi channel type teleprinter signals
>on the
>other sideband: this significantly reduces the percentage plate current
>change
>and reduces the significance of the effect.
>
>G2DAF wrote about the problem over 40 years ago, and in the design of the
>G2DAF
>linear (?), he didn't use a smoothing choke. However, he did use 866
>rectifiers with capacitor input, and enough capacity that he was pushing the
peak
>current ratings of those tubes.
Exceeding the peak current rating is more like it.
>Now whether or not that was symptomatic of Dick's
>design approach, I leave you to judge.
>
No surprise.
>
cheers, Peter
- 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
|