I disassembled and cleaned the transformer, then added spliced wires with shrink tubing insulation. After connecting things with clips, everything worked well. So I reassembled the whole thing. For a few brief minutes we had sweet tone from Strat to Dickerson. Unfortunately, the transformer gave up the ghost "shortly" thereafter, proving that electronics run on smoke (let the smoke out, the electronics no longer work).

Drastic measures were in order to save the little jewel.

A bit of research showed that the transformer was a 1941 Hadley. Surprisingly, the Robert M. Hadley Company (www.rmhco.com) is still in business today!

Hadley builds transformers for a wide variety of customers, from US Government contractors to providers of ultra-high end amplifiers. Seriously, one of the applications is an $85,000 Stereo 304TL Amplifier (you can buy JUST the transformers for $20,300).

I called Jim Hadley, President of Robert M. Hadley, and he said they could rebuild my transformer. He searched their archives and found the original drawings, dated April 26, 1941 (Click here for a scan of the original specs.).

A custom rewind for a lot size of one? I call that the ultimate in customer service. It wasn't free, but you can bet it cost a LOT more to rewind this transformer than I had to pay.

While waiting for the transformer rebuild, I went to work sonically restoring the amp. The first step was to identify the flying-saucer-shaped component on the preamp tube. I posted some pics to a few google groups and was rewarded with a speedy identification (thanks to the smart folks on rec.antiques.radio+phono for their help).

The component was a rare battery bias cell used to apply a negative bias to the 6SF5 preamp tube.

The bias cell initially read zero volts. Over time the electrolyte dries out, which kills the cell. I read about several methods used to rejuvenate these cells, most of which involved drilling tiny holes in the membrane and injecting water  (www.philcorepairbench.com/biascell.htm).

Instead of drilling, I dropped mine in a cup of boiling water. Instantly we saw tiny air bubbles streaming out of the cell as the heat expanded the air inside. As the water cooled, the thirsty cell sucked some water in (nature abhors a vacuum). I dried the cell and resealed the leaky seam with crazy glue. It has read 1.24 volts ever since.

In the example diagram on the left, the bias cell is shown between the Grid and ground. On my amp, the cell is placed in series with the input jack, a common alternative in the 30s and early 40s. Someone had modified the amp after the cell died by bypassing the bias cell with a red wire (see photo above right). After rejuvenating the bias cell, I removed the red wire to restore the bias to its original state.