[R-390] HP-8640B vs. HP-8654A/8655A/11710B

[2002tii]

Dick wrote:

>Tucker had a truckload of [11710Bs], between them and e-Bay,
>there were a lot of them. I got one, NIB, for about $50. I was not
>too impressed with the design.

It's just a 50 MHz oscillator, a mixer, a one-transistor post-mixer 
amp, and some filtering and switching (there are also 5 MHz and 1 MHz 
outputs divided down from the 50 MHz oscillator, but they are not 
pertinent to the primary function of the unit).   Their big claim to 
fame was that the conversion gain/loss was within 1 dB over the full 
frequency range, so you can rely on the output level set by the 
8640B.  Not always an easy trick with a mixer output covering 10 kHz 
to 11 MHz and input levels from 0 dBm to -110 dBm.

I reduced the amplifier noise a little by replacing the 
single-transistor RF amp with a two-transistor feedback amp in one of 
mine, but that by itself made very little difference -- the noise 
added by the 11710B is mostly from the 50 MHz oscillator, as I 
mentioned in a previous message.  I replaced the oscillator in the 
one I modified with 10 MHz from the shop 10 MHz distribution amp 
(which can be fed by a GPSDO or a Rb source) and a 5x multiplier, 
with appropriate filtering.  That one is nice and quiet, as well as 
holding the mixer LO frequency to the same < ppb tolerance as the 
generator, which is also referenced to the shop source.

Best regards,

Don