These are $2 Nano clones, and both of those crystals are almost microscopic.
USB spec says the 12mhz crystal must be correct within 0.25%, or 30khz.
The 16mhz crystal doesn't need to be any more accurate.??
But maybe they are accurate enough, and pulling either or both lower with parallel caps
will cure this issue.? Ideally far enough to miss both high and low side BFO placement.
Some 0201 caps across the IC pins may be the easiest way to hack them in.? ?;-)
But I think shielding/bypassing should be looked at first.
If the 12mhz crystal were the problem, we could just shut it down entirely when not using the USB port.
But post 40432 says tones vary when finger is on the 16mhz crystal, suggests it is harmonics of the 16mhz
causing trouble.? Note that the CH340G and 12mhz crystal are on the back of the Nano, away from the radio.?
Jerry
toggle quoted message
Show quoted text
On Sat, Feb 3, 2018 at 07:39 am, Arv Evans wrote:
Jerry
Interesting observation that the 12 MHz USB clock oscillator is on nearly the same frequency
as the uBITX IF and BFO.? If the BFO is offset to the high side of the IF passband it could well
be within audio range of the USB clock frequency.? If this is the problem then it might be possible
to pull the USB clock crystal down with a small parallel capacitor.? That should not make a
significant change in USB clock timing but could take it out of range of the receiver IF and BFO
frequencies.? Some experimenting is probably necessary to tell how far to pull the crystal.? It
would not be good to accidentally put it within the IF filter passband.
?
Arv? K7HKL
