While digging for parts to build the little 2N7000 key interface, I came across a "Ham Gadgets"? mini keying adapter which was originally designed to allow a low current keyer to work with old Cathode keyed and high current MOPA solid state rigs that had keying currents over 100 milliamps.? Just perfect for checking this out.

Wired it up so I could plug my bug into the uBITX through it.? Baseline test, hooked the bug up direct first.? Keying was terrible and kept throwing garbage into the characters.? I put a .1uf capacitor across the bug as a quick & dirty contact bounce filter which helped a little but didn't cure it.? Next, hooked up the keying adapter and WOW what a difference.? There was still a little bounce problem but a .1uF cap right at the input to the adapter fixed it.? Might have been a bit of RF getting into the adapter through the keying line as I don't have the adapter in a shielded box, just open board on the desk top.?

It appears that the problem with the hand key circuit is contact resistance in some keys is sufficient to upset the sensing in the uBITX and will vary from key to key, with length cable and gauge of wire and a whole bunch of other variables.? In my case it appears the solution is to even the playing field by using a simple transistor switch to present as close to zero resistance between the A4 and ground on "Key Down" intervals.

I can now use my old bug with the uBITX and work on a simple transistor interface for the key paddle inputs as well.? This would give consistent resistance from the paddle inputs to ground which would help the ADC read the voltage drop across the dot and dash voltage dividers.? Just my opinion, but I intend to breadboard a simple circuit to test that theory as well.

Jim Sheldon - W0EB