EMRFD Message Archive 974

Message Date From Subject
974 2007-08-24 14:07:08 Ted Bruce KX4OM SA-612 Impedance Questions
I understand that Pins 1 and 2 have 1500 ohm input impedances. I
recently built a 40m receiver using an SA612, using the 42F123
transformer design from the SW-20x rigs. Also, I used the Coltpitts
VFO on a separate board, completely shielded, with feedthrough caps on
the power, and copper tubing and soldered shields for the coax pass
throughs for the two VFO outputs.

Originally, I used the normal 47pF/10pF voltage divider output from Q2
to feed Pin 6 on the SA612 1st mixer, and the second output off the
coil via a 10pF/270pF divider, a lower level signal, which would have
normally gone to the transmit mixer at Pin 2, to a DL4YHF 7-segment
display counter through a preamp transistor. The VFO tuning range is
3.009 to 3.094, with 4 extremely well matched (10Hz) 4.000 MHz IF
crystals in the receiver.

At first, it seemed to work OK, except that it would take the counter
about 10 seconds to start up, counting from about 6.8 MHz to 7.010
(3.99 offset in DL4YHF's standard PIC program compilation). Lately,
it won't even reach the 7 MHz range. Scratching my head on that one,
I reversed the outputs from the VFO. The counter records the proper
value immediately, and the receiver works, with reasonable complex
signals scoped at the input of the crystal filter, and pure 4.000 MHz
signals scoped at the output. Now, however, I can hear faint very
faint digital signals, repetitive, but not constant, at the receiver
output, as if something at 4 MHz is feeding through the receiver.

I'm pretty sure I have two problems. The VFO signal is not
sufficiently large enough compared to the RF signal as currently
connected. The second problem would be, if I had the cables connected
as they should be, that I don't have enough preamp gain to the
counter. I'm not sure why, when it was working correctly, that there
was a slow count-up from 6-something MHz to the correct 3.009 MHz
(7.009 with PIC offset). Now, of course, it doesn't count up past
about 6.9 MHz.

A third question is on the crystal filter. I used a Cohn-type, which
can be plugged in to the PC board for replacement with another filter,
with a 3 dB Bandwidth of 1 KHz, so my grandson can hear more of the
band noise and signals. I have the HFO set on the slope so that there
is virtually no opposite sideband audible. The filter impedance
calculates out to about 1.5 kHz. I'm wondering if I should have used a
resistive termination anyway, rather than feeding the end series caps
directly into the input to the filter and the output. It's the first
filter I've built where the filter impedances matched the connected
receiver stages.

Anyone have thoughts on this?

975 2007-08-24 18:13:12 Allison Parent Re: SA-612 Impedance Questions
984 2007-08-27 10:47:53 sp9mrn Re: SA-612 Impedance Questions
Hi Ted
Did You try a simple single fet (bf245, j310) buffer?
May be before NE612, may be before counter - I dont know ;-) what
will be better. It usually gives a good separation. I am not sure,
byt I have some troubles using that counter (IMHO excellent and very
popular in SP) with NE602 without buffer. Of course I understand,
that external VFO changes the situation, but think about it.

BTW great SMT version of dl4yhf counter by Tomek:

and buffered input for counter:

986 2007-08-27 13:47:16 Ted Bruce KX4OM Re: SA-612 Impedance Questions
Wow, great links! I used the basic PIC board from Wolfgang's site,
and I wired a simple preamp using a 2N2222 in the "development matrix"
space on the board. I couldn't use the display board that he used, as
my 7-segment display modules have the pins along the top and bottom,
which resulted in my having to use a double-sided PC board design.
With the original board and the homebrew display board, I was able to
make the combination rectangular, with the boards were aligned at 90
degrees like in DL4YHF's design. I then enclosed the ends, top and
back of the unit in scrap PC board pieces, sealing the cracks with
conductive copper adhesive tape.

Both of the designs in the links that you provided are interesting. I
didn't give any thought to buffering the preamp, as in the PY2RLM
design. All of the replies I have received point to the need for
buffering, so I will redesign the counter input. I have a lot of
J310s, both in TO-92 and SMD, so I will try those.

In your 2nd link, Tomasz's surface mount board is a fascinating
design, with the PIC board and the counter board combined as one unit.
I am going to translate his page and study it some more. It is a
two-sided board, which is of course more difficult to make using the
toner-transfer iron-on PCB process, but then my display board was
two-sided anyway. Tomasz's board is far more compact than mine, and
it will be much easier to shield and fit into an enclosure.

By the way, along with my custom Eagle CAD library "ham-qrp.lbr",
which I have recently updated with better toroid transformers,
DL4YHF's library is my most-used Eagle library of parts. I will upload
mine to the "Files" secti