Excuse a fairly rapid reply, but it's quite late at night, and I have
a very
full day tomorrow.
> I'm interested in the high tech solution, which (currently) is a
high speed
> analog (electronic) switch driven at 4X the desired receiving
frequency. It
> has advantages in ultimate sensitivity and simplicity, even though
ultimate
> sensitivity gains aren't useful at HF due to naturally occurring
> terrestrial noise sources.
> The high speed analog switches (that I know of) limits the maximum
> receiver frequency to about 75 Mhz.
Switching mixers are always a good idea, but at the higher VHF
frequencies and
above, the use of the Tayloe type approach (which is really an
implementation
of one kind of commutating bandpass filter) starts to fall over
because the
switches just aren't fast enough. However, we've used diode switches
for a
long while, and conventional diode ring mixers still work extremely
well into
the mmWave region. Again, because of the speed of devices (largely the
spreads of propagation delays with temperature and manufacturing
tolerances)
the use of a digital quadrature phase shifter, like a Johnson counter,
isn't
too practicable above a few tens of MHz. For narrowband applications
simple
delay-line phase shifters can be a practical means of generating the
quadrature components, as can packaged phase shifters eg.
Mini-Circuits, or
coaxial-line hybrids. I used the last approach in a Weaver type 2m
transmitter I made in the early 1970s.
There is a problem with a discrete component approach, however. To get
the
best performance from a direct to baseband receiver, it's quite
important to
balance-out any DC offsets in the the two quadrature branches of the
signal
path. This requires some sort of adaptive system. It can be done
either under
software control or in hardware. Because of the thermal gradients
which will
inevitably present, It's always going to be more difficult to do this
with a
discrete component solution than on silicon. It's not impossible to do
with
discretes, though.
The AD device which I'm playing with at 1.3GHz has built-in analogue
balancing
circuitry. There's a lower frequency (50 - 800MHz) version of the part -
don't ask me the number at this time of night(!) - which would make a
very
nice heart of a 2m receiver or transceiver. I guess that the performance
would be of the same order as typical amateur receivers.
> I would think the technology would really shine at 2 meters (and
higher),
> where many mixers can be eliminate by direct conversion.....but,
where do
> we get a switch that can operate at 600 Mhz? I think counters can go
that
> high, so clocking the electronic switch is currently possible.
> Is there a lower tech approach that does away with the 4X local
oscillator
> and switching speed requirements?
The AD chips don't use digital dividers to establish quadrature drive for
their mixers. Rather they use broadband polyphase all-pass (I think)
passive
filters integrated onto the the die.
> I'd like to see some discussion on alternative means of direct
conversion
> that offer high performance and opposite sideband rejection.
> I can't even imagine anything working well at 10 GHz. Show me the way
please.
In fact there is already very nicely designed direct-conversion
transceiver
for 10GHz designed by S53MV. The signal path could be easily adapted
to drive
an ADC. Matjaz's design is a few years old now, and there are now
components
available which could simplify it considerably.
Have a look at the Hittite web site. They do integrated SSB passive
mixers
with 35dB unwanted sideband suppression covering 10GHz!
The major problem at 10GHz is making a suitable local oscillator source.
Although you could use a 100MHz region vxo and multiplier strip possibly
locked-back to a stable reference, I think it's now possible to make a
(relatively) simple synthesiser in the 2.4GHz region with adequate
performance. That's a job for over the winter...
73
Chris
GW4DGU
Yahoo! Groups Links
<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/soft_radio/
<*> To unsubscribe from this group, send an email to:
[EMAIL PROTECTED]
<*> Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
