Re: 1MHz output impedance query?
 

That looks promising now. It's good to see you have a 'parasitic stopper' resistor in series with the base of the 2N3904.

The 2N3904 is a bit more lively than a regular jellybean BJT and if used in an emitter follower it can go unstable up at 250-450MHz or so if there is any significant stray inductance in the connections to the pi filter. R6 should provide good protection against instability here.

I'd expect to see the base biased up at 5-6V DC so maybe you could play with the value of R5 when you build it. I'm not sure what current gain your 2N3904 will have so R5 may have to be a select on test component because your base bias resistor values are quite large.

Re: 1MHz output impedance query?
 

A quick update, I’ve had a month off due to medical intervention at the hospital, but I did move the project forward today by testing the “completed” hardware. The 1MHz and 10MHz square waves are good but I was disappointed to find that the 1MHz sine wave output was rather low level, until I remembered that the output emitter follower needs a 12v supply. So then I tested the power supply, nice steady 5v for the logic and GPS receiver but the 12v has a nasty 50Hz ripple of about 500mV, degrading the 1MHz sine performance somewhat. That’s the next job…..

Re: 1MHz output impedance query?
 

Aha! I'd installed a 270R resistor in the place where an 8k2 should be fitted, that'll degrade the 12v by drawing way too much current. More tests tomorrow!