[R-390] PTO learnings and sensitivity un-mod

[jgedde at optonline.net]

I've been tracking down a sensitivity issue with my R390A.  I found a few bad 2.2K resistors which I changed, but sensitivity still wasn't up to the gain stage charts in the TM.  Slowly I came to the conclusion that my PTO was feeble (apart from having endpoint drift).  This is where it gets good.  The PTO was indeed weak and putting in a new tube just made it worse.  So, I got under the hood and found the 220k screen resistor came in at 290k.  It was the subject of a change from 56k to 220k at some point.  Changed that out, but got a far from flat amplitude vs. frequency.  So, I began to wonder what that 220k change was all about...  After some reading I decided to experiment.  Put in a resistor substitution box and found output amplitude got better and flatter vs frequency with a decrease in value.  I made the decision to revert back to the 56k pre-mod value.  Wow!  What a difference!!!!  PTO output went from 750 mV to 3.3Vp!  It is virtually flat from 2.4 to 3.5 MHz.  The receiver is WAYYYY more hot too!  I haven't gone back and checked stage gains, but this is very promising.  Carrier meter picked up 12-15 db!  Very pleased.  The 56k un-mod also made the PTO much more insensitive to tube changes...

As far as endpoint drift, this seems to be related to mechanical stress on the Oldham coupling.  As far as I can speculate, the two ends and the center plate shouldn't be tightly sandwiched together.

I learned all about how the Cosmos PTO is linearized and yes I had to take off a turn of the endpoint coil, but now I have a drift free (so far) PTO linear from end to end within <300 Hz.  40 little tiny screws and a few hours and I'm a happy camper.

It's a good R390A day!

John