Re: Stone Soup

Lev
 

On Wed, 05 Sep 2018 09:55:42 -0700
"Jerry Gaffke via Groups.Io" <jgaffke=yahoo.com@groups.io> wrote:

Nothing logical about building a new one-off homebrew HF transceiver.
But it can be fun!

None of these are necessarily bad ideas, but here's my yeah-but's:
:-)

1. Tuned IFs
Both IF's are tuned by crystal filters already.
Yes, but if we didn't amplify unwanted signals, the XTAL filter has easier task. :-)

2. Filter the output of the LOs
Diode ring mixers are commutating switches, not exactly the same as the
true multipliers of a Gilbert cell. So work best with a square wave as
the local oscillator, not a sine wave. We could use higher levels at the
local oscillator port of the mixers, perhaps just need changes to those
resistors. 
Okay, but the harmonics can cause unwanted products.

3. Push-pull preamplifier ( https://www.sv1afn.com/j310preamp.html )

More gain in front of the first serious crystal filter on RX is not
necessarily a good thing for a wideband HF receiver, just makes it
easier to overload on some local AM broadcast transmitter.

4. RF transistors (BFS17P)
Yes, we could use something better than 2n3904's in places.
Though using buckets of $0.02 2n3904's/mmbt3904's (at lower gain where
appropriate) has a cost and stocking advantage.
I guess the BFS17P is also cheep... but whatever.

5. PA redesign (normal push pull design, two stage of IRF530 ) use
BFQ19S for drivers I plan to look very hard at the G0UPL linear, sounds
like it will be clean and flat and robust and cheap.

6. More filters on the output

If we're talking transmit LPF's, the only failing on the uBitx is
routing there, not the filters themselves.
Ok.

7. Maybe change of the mixers to JFET mixers
Perhaps, but what are we trying to solve?
To be honest, nothing! I just want to build a JFET based mixer. You can
balance JFET based mixer easily. Or we shall use DBMs.

Raj's result suggests the spurs are not caused by the mixers, but rather
by harmonics entering the mixers also stuff getting coupled into the IF
amps from the nearby power amp due to board layout.
Okay. So that is why we shall filter the signals of the LOs.


73s de HA5OGL

--
Levente Kovacs
Senior Electronic Engineer

W: http://levente.logonex.eu

Join BITX20@groups.io to automatically receive all group messages.