[Nathan Slingerland]

I carefully counted the number of turns in the primary. First I used a magnifying glass - too easy to lose track. Using the trick of running a pin over the surface was easier to keep track but in the end I found that blowing up a clear picture of the coil on the computer made it even easier.  

[Image: https://farm5.staticflickr.com/4909/4597...918d_z.jpg]

Anyway - I finally settled on a count of 60 turns of ~32ga wire. Looking at the bottom of the coil form and counting pins clockwise the primary starts on pin 3, is applied clockwise on the coil form and connects to pin 4.

I removed the remnants of the original primary and cleaned it up as described in the PhilcoRepairBench.com article.

I next cut out a plastic form the same size as the original celluloid form. 4"L x 1 3/16"H with notches in the same places. I used some tape to hold the form on the coil (just a temporary measure while rewinding it).

[Image: https://farm5.staticflickr.com/4914/3216...f930_z.jpg]

[Image: https://farm5.staticflickr.com/4835/4598...89fe_z.jpg]

I then soldered #32 magnet wire to pin 3 and started winding clockwise. It wasn't too bad to wind thanks to the #32 magnet wire being quite strong and easy to work with and hold some tension.

[Image: https://farm5.staticflickr.com/4902/4603...0eb7_c.jpg]

After resoldering it in place the set came to life!

[Image: https://farm5.staticflickr.com/4842/4421...f1c8_z.jpg]

The new primary isn't physically spaced as much apart as the original cotton covered wire but it doesn't seem negatively affect performance.