"Tim Shoppa" wrote in message
oups.com...
Cool stuff. Way cool.

Just trying to sort out the bandswitching from the pictures:

Antenna input is switched to one of six bandpass+optional RF amp
boards. It looks like the antenna switching is by the black relays in
the back?

Eight bands, six band boards, a couple of the bands must be close
enough together that they utilize a common board? Or maybe the IF *is*
one of the bands...

There appears to be a preselector knob on the front, how does this go
into the bandpass/RF amp boards.

The output from the RF band boards is then brought to some sort of
combiner board that I cannot make out anything but a bunch of SMA jacks
on. Is there switching there too, or (because of power switching to
the RF boards) is this just a simple combiner?

And that Eddystone dial is to die for. Oh, man.

Tim.
Hi Tim,
An excellent job of sleuthing- I never could have done that.The RX is
essentially an 80M RX with xtal controlled converters for the other bands.
You are correct, the rear board switches the antenna to one of 6 front end
BP+LP filters- 10M 28-28.5 and 28.5-29 share a filter and 80M is bypassed
through this section directly into the preselector (which works on 3.5-4MHz
only). The board along the power supply side switches the filter outputs
into the 1st mixer. A relay on the output of the 1st mixer selects either
its output ( 3.5- 4MHz) or the 80M signal directly from the antenna. This
signal then goes to the preselector- which is a very sharp tunable BP.
The relay board on top of the chassis toward the rear selects the
appropriate L.O. and routes B+ through the coax to the L.O. 160M and 49M
share the same L.O. but different input filters.

Again, brilliant work on your part.

Dale W4OP

That's very well said!

The "optimum" design is, as you say, very much a matter of the
particular situation. In my career doing circuit design, I've worked for
a company where products were custom or nearly custom -- a production
run of 10 was a big deal. In that situation, if I could save half an
hour of design time by using a $20 part instead of a $5 part, it was a
very good trade. I've also designed a product for a mass near-consumer
market. There, a week of extra work spent in getting a few pennies out
of the production cost was worthwhile. We all have to remember that:

1. There's no one best way for everybody, every project, or every market
or situation.
2. What's best for the mass manufacturers, or for that matter another
homebrewer, isn't necessarily the best for us.
3. The best way to do a design is constantly changing, as both techology
and our own situations (e.g., the amount of time and money we have) keep
changing.

Design is a creative process, and there's a lot of art as well as skill
in it. Finding a better way to do something requires both, and knowing
the appropriate time to declare it "good enough" requires art, skill,
and discipline too.

Roy Lewallen, W7EL

Michael Black wrote:

"gb" ) writes:



I see you used a 9 MHz IF.
Any particular reason that you used separate (LO) oscillators for each band
(rather than a single oscillator with switchable crystals) ?


Surely because it works out to be simpler.

People often make the mistake that extra circuitry is bad because it
complicates things, but the reality is often that extra circuitry simplifies
things.

Back in the days of tubes, their bulk and filament current (or slow
startup if you switch the filament) meant that one did give thought
before adding another tube. It probably was simpler to switch crystals
than switch oscillators, and compared to the cost of tubes and sockets,
the switch was relatively cheap.

Once transistors came along, their size, low current and instant on
meant the game changed. If adding another transistor in an IF stage
meant that all the stages were running at a lower level, better to
do. It added little to the cost, and pretty much nothing to the current
drain.

If you switch crystals, you limit layout. You need that switch near
the crystals (and in some cases may need two poles per position), or
you need to use relays (I've seen it done in mods to tube equipment)
or diode switches. Those diodes can often cause some problems, depending
on choice and useage. Switching crystal oscillators means none of this
applies, and the cost of the transistor and passive components is
nothing compared to the cost of the crystal.

This is a theme of some of Hayward's work. Various times, he's described
transceivers, and not only is he using separate IF strips for receive
and transmit, but he's got a separate IF filter for each function. His
reasoning is that it makes the chain cleaner by not having the switching.
Given the cost of commercial filters, I'm not sure that's the best choice,
but certainly there is a tradeoff. If adding a few dollars for an extra
RF stage rather than switching tuned circuits requires a cheaper switch,
or makes it easier to do the switching, it may be worth spending those
few dollars rather than fuss too much.

Jerry Vogt (I"ve spelled that wrong) who seemed to be connected
with Hamtronics (at least, he wrote about their stuff for Ham Radio magazine)
pointed out in an article about preamps that a manufacturer needs to
cut costs. They skimp on parts because any single part saved is
multiplied by the number of units built. But a hobbyist building a single
unit doesn't have that multiplier. If they toss in an extra bypass
capacitor here, or build each stage into a metal box for shielding,
the cost is right there. They don't have to worry about 10,000 more
capacitors, or whatever.

Michael VE2BVW

Dale, that receiver is among the best homebrew rigs that I have seen
in 45 years! Incredible job!

Ted KX4OM

On Wed, 23 Mar 2005 22:08:37 GMT, "Dale Parfitt"
wrote:

I had posted this elsewhere, and it has created some interesting feedback.
Based on Wes W7ZOI's Preogressive RX from the late 1980's. A wonderfully
balanced design to which I added some bells and whistles I like in a
receiver.
These are some shots of my almost finished 1980's theme HF receiver.
All stainless hardware, teflon wiring. Rock stable VFO, homebrew xtal
filters ( 5/2/5/0.5), selectable and defeatable AGC, notch filter, tone
control, QSK input,built in front end BP filters for each band, preselector,
AM/USB/LSB/CW.
Enjoy:
http://www.parelectronics.com/pics/W7ZOI11%20copy.jpg

http://www.parelectronics.com/pics/w7zoi13jpg.jpg

http://www.parelectronics.com/pics/w7zoi14.jpg

http://www.parelectronics.com/pics/w7zoi15.jpg

Dale W4OP
for PAR Electronics, Inc.