[Amps] RF Chokes

[w1itt at att.net]

Tom & Phil
My network analyzer is out on the west coast right now, awaiting the repair of a 500 foot crane that was going to be used to lift a couple TV broadcast antennas.  Just for grins, when I get it back to the real world, I will look at a couple electrolytics swept from 160 through 10 meters.  You've got me wondering now what sort of "load" they present.....
Always fun to hear from the lads here on "amps".  We cover all the stuff they left out at school!!
Thanks and 73
Norm W1itt
 -------------- Original message ----------------------
From: "Tom W8JI" <w8ji at w8ji.com>
> > coverage. 11-12 mhz must be avoided. This is fine for most 
> > of us, but the L
> > is not close to optimum for 160 meters, and it is 
> > physically much larger
> > than the Harris choke. I get by with the RF Parts unit on 
> > 160m by using .01
> > ufd @ 10 kv bypass caps at the base. This keeps most of 
> > the RF out of the
> > power supply; not so important with oil-filled filter 
> > caps, but very
> > important to protect electrolytic capacitor banks found in 
> > most amateur
> > amps.
> 
> Let's look at that factually for an AL12 series amplifier:
> 
> The choke reactance is 3120 ohms.
> 
> The load impedance is 2500 ohms at 1.75kW
> 
> The RMS voltage is 2100V
> 
> The RMS choke current is .673 amperes
> 
> The net bypass capacitance on the HV, ignoring 
> electrolytics, totals .005 uF or more, which is 17.7 ohms
> 
> The RF voltage across the bypass capacitance is about 12 
> volts
> 
> The electrolytic series impedance is over 70 ohms at 1.8MHz 
> in sample units I tested. The RF current in the 
> electrolytics is under .18 amperes if we assume worse case, 
> but it typically measured under 75mA.
> 
> That 75mA is on top of the 2-3 amps of effective ripple 
> current (peak current is MUCH higher).
> 
> How much life increase should I expect if the bypass 
> reactance was changed from .005uF to .015uF, and ripple 
> current caused by RF dropped from 75mA to 25 Ma or less??
> 
> The answer is, none. The 2-3 amperes of effective ripple 
> current at 120 Hz dominates the problem.
> 
> This is not as bad as many Internet suggested mods are, 
> since the additional C at the bottom of the choke won't 
> hurt. Unfortunately some mods on Websites actually hurt 
> reliability, but this suggestion is benign. The bad ones 
> include resistor grid fuses, increasing power supply 
> capacitance, and removing caps across diodes, and increasing 
> resistance across electrolytics.
> 
> Food for thought...
> 
> 73 Tom 
> 
>