As far as using a ProgRock goes, see the other posting that I just made - but I thought that I'd comment on using a Ge Mastr II (or Exec II) for a propagation beacon.

Years ago (2001, actually) I worked with WA7X in assembling a 6 meter, 2 meter and 70cm beacon using cast-off Ge Mastr Exec IIs - and here's what I found:

- You can't key the beacon with the PTT line:  It's WAY too "clicky" - but this isn't too surprising.  With a crystal, you end up with chirp (no surprise) but the fact is that keying the signal feeding a string of non-linear stages (multipliers, Class-C amplifier stages) means that on the attack and decay, you will not only get a sharp edge, but you will also get a very brief burst of noise as these multiple stages are momentarily in an indeterminate state, which means that the key clicks are as much as a few 10s of kHz wide.

Because of this, the PTT line is used only for a "beacon disable" function (via remote control) and, instead, a closed-loop servo power control was used, replacing the original RF output power control circuitry of the PA.  If you want to see the gory details, go here:

http://www.wa7x.com/beacon_tech.html#transmitters

Not only does this provide an acceptable keying waveform, it also provides the ability for three power steps (e.g. 1 watt, 10 watts, 100 watts) which are useful for both propagation analysis (the beacon message includes a 5 second key-down at each power level) but it also saves on the power bill:  For a long time, the WA7X beacons operate at 100 watts only during the initial "VVV", the 5 second test carrier and its announcement and the final ID, the rest of the time operate at 10 watts.

As mentioned, these beacons have been operating since about 2001 with very little required maintenance (the 2 meter beacon lost a pre-driver early on - easily replaced).

If we were to build these beacons again, I'd certainly consider using a ProgRock (one per beacon) with a TCXO, but I would do the keying the same way.

73,
Clint
KA7OEI