[rfeenstra]

So, I have the radio up and running.  Seems to be working OK now.  It was a bit of a chore, though.  Seems the Osc/mixer is a bit sensitive to coil rewinds. Rewound the coil even though it worked prior to disassembly.  Close inspection revealed green wire due to the nitrocellulose.  So I rewound.  The original wire measured out to be 37 or 38 AWG.  I used 37 with 35 turns.  The oscillator would run with the tuning control below about half.  Died above 1100 khz.  I could watch it on the scope reduce amplitude quickly and die.  So, I rewound it again using 34 AWG thinking the cloth covered original wire gave a bit more spacing (I read somewhere here that they usually use a heavier wire for that reason).  Now the oscillator gave a better sine wave but died at 900 khz instead.  Hmmm.  Decided to remove 5 turns of wire and try again.  Voila, the oscillator worked all the way across the band but was weaker at the high end.  I removed 2 more turns and then the oscillator was more consistent.  Removed 2 more turns and now the oscillator was stronger at the high end of the band.  I added 3 turns ( a little more spaced as I had to make room for the splice) and now the oscillator is fairly consistent across the band with a nice sine wave.  I took the signal for the scope off the cathode of the 36 osc/mixer tube.  I get good sensitivity all across the band now but it seems this radio is very sensitive the to settings of the antenna and RF trimmers.  If not in a slightly compromised position, I get a lot of whistling between stations.  If anyone has insights into this, PLEASE educate me!

Here's  pic of the almost finished chassis.  You gotta love the globe tubes!

       



Here is the rewound coil for the second time:

   

The final issue I've been dealing with is the output #42 tube.  If you look at the photo, it looks as though the screen grid is glowing, and I think it is.  The screen voltage and the plate voltage are normal according to the voltage chart.  The screen is connected directly to the return from the field coil and so is higher than the plate since the plate voltage is dropped by the output transformer.  I tried a NOS 42 I had and initially, I thought it was doing the same thing.  So I tried to reduce the screen voltage by inserting a resistor.   I got it slightly below the plate voltage but still no change.  After close observations, I decided that on the NOS 42, what I was seeing was actually a reflection of the cathode on the bright wires of the screen.  I don't think that is the case with the globe 42.  I suspect that this tube, even though it measures fine on the tester, is not long for this world.  I am certainly no expert in this stuff so if any of you have input, I'm all ears!!