This battery eliminator was part of a mixed lot I acquired at one of the last Radiophile auctions I attended.
From the size, and socket for the battery plug, I guessed it to be an eliminator for a B114, although there were no markings whatsoever on the case. The internal construction appeared to be point to point wiring, with each component wrapped in oilskin (or something similar) for insulation. There was at least one broken connection (a resistor with only one leg connected...) so I didn't feel like powering it up.
I carefully unwrapped each component, and tried to draw out the wiring. It was fairly straightforward, with two capacitors each for LT and HT reservoir/smoothing. The LT supply was centre-tapped, with the centre tap being connected to the case (as well as LT negative).
The metal rectifiers may have worked - I didn't test them - but, in the hope of long term reliability, I decided to replace them with silicon. Likewise, the old electrolytics may have been OK, but modern types should be a better long-term bet.
The original component parts...
The replacement parts were 1N4001 for the two LT rectifiers (centre-tapped winding), and a single 1N4004 for the HT (half-wave). The capacitors were 22uF at 400V twice for the HT
(what I had on hand), and 1500uF at 10V for the LT. (Plus LM317, 220R and 47R for the voltage divider, and two 0.1uF decouplers around the LM317).
The new components were mounted on a small piece of 0.1" veroboard, with veropins for the (numerous) flying leads.
I tested the transformer with new rectifiers & capacitors, and found it to be OK. But the on-load voltages (at 10mA HT and 125mA LT) were way too high (about 5V LT, and 140V HT). For the LT, there was sufficient headroom to instal an LM317; for the HT I increased the smoothing resistor (between the two capacitors) to 2k2. I also fitted a small neon in series with a 36V zener and a 4k7 resistor to bleed a little of the off-load HT voltage.
A soak test indicated that the LM317 needed a small heatsink, which I fitted, free standing (it could possibly dissipate 1W if 250mA LT is drawn...).
The board before it was installed in the box...
I used some of the retained oilskin to insulate parts of the interior of the box; and added some plain veroboard as an extra insulation measure. The whole thing is now glued together with my favourite hot-gun glue. The mains lead (double insulated, two-core) has a new grommet around it, a cable clamp just inside the box, and is glued to the floor of the box for good measure. It terminates in a choc-block connector, glued to the side of the transformer.
The finished eliminator...
