VARLEY Battery Eliminator

[Herald1360]

Intrigued by by the double choke I ran some simulations-

Double choke, 4u reservoir and smoother, about 1.5V p-p ripple at 80mA dc load. (With chokes in "wrong" phasing, ripple was a LOT higher!)

Single choke, 47u reservoir and smoother, about 100mV p-p ripple at 80mA dc load

In either case, ripple from the RC dropped outputs was negligible.

1.5V p-p ripple would be fine on a tetrode or pentode output valve, might hum a little on a triode but probably inaudible on speakers of the day.

Given the lack of large muffage caps, the double choke looks quite a good way to go.

Output volts a bit higher with single choke due to only half the dc resistance in circuit.

[David Simpson]

When I initially rebuilt it using the original two choke configuration & two 10uF electrolytics, like Herald 1360, I experienced a ripple, which was approx. 1Vp-p at 60mA loading. And approx. 15V p-p if the chokes were wrongly phased. However, after a soak test of a few hours, C1, which was 2nd hand, decided to give up the ghost.
So then thought to try the more conventional Pi filtering, with the chokes in series to produce 20H and fitted the two 47uF electrolytics. The HT then experienced no ripple, but instead two spiked sawtooths of either 1V or 5V, depending on the alignment of the windings. So just used one 10H choke - perfect, no ripple & no sawtooth. Now the choke, like the mains T/F, assembly had been well meggered at 500V, but the completed circuitry now contained earthing bonds to their metalwork. However, the original 1930ish circuitry obviously contained no earth bonding whatsoever. Perhaps some ancient phenomenon produced the spikey sawtooth. Anyway, I'm not going to disassemble the renovated B/El and commence lab experiments. I'm quite happy that its all tickety-boo now.
Was the double choke fitment just used by Varley, I wonder ? I haven't seen any evidence of the configuration in later 1930's mains sets made by the manufacturers we all know.

Regards, David