[Amps] The power of an oversized electrolytic capacitor / how much PS filter C is enough?

[Leigh Turner]

Fully agreed Carl.

Cheers,

Leigh
VK5KLT

________________________________________
From: ham_amplifiers at yahoogroups.com [mailto:ham_amplifiers at yahoogroups.com]
On Behalf Of Carl
Sent: Friday, 11 May 2012 10:13 AM
To: garyschafer at comcast.net; 'Leigh Turner'; 'Jim Thomson';
amps at contesting.com
Cc: ham_amplifiers at yahoogroups.com
Subject: [ham_amplifiers] Re: [Amps] The power of an oversized electrolytic
capacitor /howmuch PS filter C is enough?

  
Id prefer to call ripple, a 120 Hz, for instance, AC riding on the DC.
Varying DC due to regulation is not ripple IMO and if that varies 
excessively IMD will suffer since the tube load line is changing.

The excessive C fad used by a few has diminishing returns beyond a certain 
point that nobody has detailed yet. I could really care less if the IMD 3 
is -32 or -32.5; or -45 vs -46

Carl
KM1H

----- Original Message ----- 
From: "Gary Schafer" <garyschafer at comcast.net>
To: "'Leigh Turner'" <invertech at frontierisp.net.au>; "'Jim Thomson'" 
<jim.thom at telus.net>; <amps at contesting.com>
Cc: <ham_amplifiers at yahoogroups.com>
Sent: Thursday, May 10, 2012 11:39 AM


> On Behalf Of Leigh Turner
 
> A large energy storage capacitance on the HV rail makes sense from the 
> point of view the amplifier load current requirements have a large 
> peak-to- average ratio in SSB service. A stiff B+ voltage also helps 
> with minimising IMD.

I have seen this bandied about but I have never seen any figures on how much
power supply ripple really affects IMD. It would seem to me that the better
the linearity the amp has the less any plate voltage variations will have on
IMD. I would guess that with the typical amplifier that 10% ripple would
cause immeasurable IMD contribution.

> 
> The tolerable ripple level superimposed on the HV plate supply above 
> which incidental AM hum sidebands appear on the radiated signal is an 
> interesting question; the anecdotal experience of Rich's friend with 
> only 2 uF of filter capacitance suggests there's a large tolerance to 
> ripple on the plate supply with the tube and tank circuit exhibiting a 
> good PSRR.
> 
> We can note the amplifier in question here uses a 4-1000 tetrode, and 
> that such tubes exhibit a plate current virtually independent of plate 
> voltage, i.e. tetrodes are a constant current device. Such a 
> characteristic would make it more immune to ripple and noise on the B+ 
> plate supply.
> 
> Leigh
> VK5KLT
>
>
A tetrode operated in grounded grid operates as a triode as there is no
interaction between grid and screen.

If the amplifier is operated in class C then any variation (ripple) in 
plate voltage would directly modulate the signal. But when operated as a 
linear amplifier there is much less chance of mixing to take place which 
is necessary to create side bands.

73
Gary K4FMX