The prior experiment and realization about the coil value was quite timely. My first coil was about 25% below the stock coil, coming in at around 33uH instead of 45uH, so I was already thinking about making a second one. My original calculations were with solid core wire instead of the stranded Litz wire. Then I ran into another problem. Namely, the threaded ferrite slugs I have are too tight once threaded all the way in. Even with a lubricant, they bind in the cardboard and you will eventually crack it. That was quite annoying to discover. I ended up destroying part of the cardboard former while getting the shattered slug out. Before I broke the slug, I checked the coil value... It looked way low, only about 90uH. More on that in a moment.
On to attempt #2. I ordered three difference coil forms is slightly different tolerances for the inner and outer diameter. I picked the one that seemed to have the smoothest fit, as opposed to the one that matched the original length, which is what I started with. I quick slice with a razor blade fixed the length issue.
The other lesson I learned was to put some sort of separator between the winding form the the ends of the coil to allow for better coating of the layers. I used waxed paper. This is the first layer.
After each layer, I coated the winding with nail polish and let it cure. This is layer #2. Because of the waxed paper separator, I could go all way to the end of the winding. I was also a bit more patient and let everything cure well prior to moving on to the next layer.
From the first coil, I knew I needed more than four layers. Because each layer has more wire in it, the progression isn't linear. Something around 4.75 layers would have given me the original 45uH. However, I knew I needed a higher max inductance since I needed something around 500uH. The stock coil had about a 7x range of inductance, from 45uH to over 250uH. I targeted my second attempt for about 98uH with no core. This worked out to six full layers.
Fully wound and waiting for the coat of nail polish to cure.
This time, because I was able to thoroughly coat each layer, there was no collapse of the end when I took it off the winding form.
Even so, I went ahead and separately coated the ends of the coil, to help prevent any tendency to unwind while being handled.
The finished coil.
The terminal tabs are a bit big, but I had ready access to them and they work just fine. I wasn't going to go order smaller ones, especially since this is one of those cases where the parts cost is a smaller fraction of the shipping cost.