Ken Johnson wrote:

> I have done it with a 4511 and 74hc237, it works ok but I can't get
> rid of the multiplexing noise (the signals are multiplexed).

  You pose an interesting conundrum.  You have of course left out the
*important* detail - where does the data come from to be displayed.  If
the data itself is multiplexed as you sort-of imply, you may have a
problem, because you will have hash at that point if not all the time.

  For a frequency display of course, it need only change when you are
changing the frequency (tuning) and since you can't receive as the
frequency is actually changing, no problem.

  Since the 4511 contains a latch, just use six 4511s (and an HC138 or
the 237, whatever...).

  If you are using a synthesiser and *need* to use a MPU, use the
reference oscillator (which you have to shield anyway) to clock the MPU.

> I would like to use a single-chip solution such as programmable logic
> of some sort so I can put it all in a small shielded box (it lives in
> a very sensitive 137MHz weather satellite receiver).

  The Allegro chip suggested sounds good.  If it doesn't multiplex at
all (whilst receiving), you don't need shielding.

> I though of an eprom but I need 14 output lines (7 segments, decimal
> point and 6 digits).

  Yes, hardly "single chip".

> I don't want to use a pic, then I would have to get rid of the pic's
> clock as well.

  Definitely [OT] then!

  The most interesting approach to multiplexing is that used by bedside
clocks.  They use a two-phase multiplex generated by a centre-tapped
power transformer winding (about 3V per side).  Half the LEDs go to each
end, the centre tap being ground.

  The clock has an input from the transformer which serves the dual
purpose of keeping time and gating the alternate LED sets.  The beauty
of this is that the sinewave applied to the LEDs generates virtually no
harmonic component beyond a few hundred Hertz.  Switching of the drivers
occurs at zero voltage.
--
  Cheers,
        Paul B.