[SignalOne] Driver

[Mark, K5AM]

2003 January 08, 0321 GMT

John, K0JD, wrote:

 >>>>>>>I was curious from the start why they
didn't key the bias on the whole driver stage, like
many modern solid-state rigs I know.  The constant
heating, conducted to the chassis, is surely a major
cause of drift  ...<<<<<<<<

I agree completely, John! Thanks for your
interesting posting to the list.

In my homebrew radio, the driver is very similar to
the CX7 driver, but
1. The bias on all the bipolars is switched off in
receive. (In the CX7, an early factory mod switched
off Q3.**)
2. Each bipolar has its own bias circuit. (In the CX7
driver, Q2 is biased from the Q4 bias circuit.**)
3. Each bipolar bias is adjustable with a trimpot.
4. The last two stud-types are sinked to a small
external heat-sink.
5. The driver is positioned at the opposite end of the
radio from the front-end local oscillator. (In the CX7,
the driver is directly above the front-end crystals.)

** I hope I got the numbers right. I am out-of-town
with no CX7 circuit, with solar-powered eMail in a
vacation cabin at 7900 feet. Also tired, from working
out in the snow all day (good exersize for an old
man, says XYL Lisa).

All these variations are fully documented and
discussed in:

HF Circuits for a Homebrew Transceiver, QEX,
Nov/Dec, 2001, 20-42.

The above article includes specs for bias levels and
bipolar currents. The circuits could easily be used to
modify a CX7 if needed, but I have not done so in
my six CX7s. However, I have sometimes shunted
the big bias resistors for Q4 with something to bring
the idle current up to specs. This is good for
linearity. The original bias resistors may increase in
value as they age, which drops the current. I always
measure all the voltages, and calculate the currents,
and measure all the rf voltages, and never change a
part unless it is proven bad. Working on the driver
inside the radio must be impossible. Back in 1970 I
found I had to build a test jig for testing the driver
on the bench; it's been used ever since.

All the articles (8 up to now, 3 more planned) on the
homebrew transceiver are listed at:

http://www.zianet.com/k5am/publ.html

The main article on the homebrew transceiver,
giving the general plan (based on the CX7 frequency
mixing scheme), can be downloaded at:

http://www.zianet.com/k5am/qex/hb1.pdf

---

Always fun to chat about the CX7!

Great eMail list! Best regards to all.



----

73,

Mark, K5AM

-----
Horse Mountain DX Ranch  DM54wa  NM
7900 feet.
k5am@zianet.com
------
Home: DM62ni
Mark Mandelkern
Las Cruces, New Mexico, USA
k5am@zianet.com
First callsign: W9ECV, Milwaukee, 1948.
10 band DXCC confirmed. 134 countries on 6 meters.
Homebrew station:
<http://www.zianet.com/k5am/ncj/ncj.html>


----

030108.0321