[ARC5] 70+ year old errors found and corrected.

[Jim Haynes]

Not that old, but I have a Sunair transmitter system, with KW linear
amplifier and automatic antenna tuner.  Perhaps the original owner,
Virginia Highway Department, got rid of it because it was a lemon.
The antenna tuner gave me fits until I discovered a bad solder joint
on the lead to one of the servo motors.  While trying to find that
problem I caused some additional damage I had to repair.  Where the
schematic called for shielded wires they used micro-coax cable.  That
stuff has a very thin center conductor that is embrittled when soldered,
so very little motion will make the connection break.  There was no
requirement for coaxial cable; ordinary shielded wire would have worked
fine.  And, lacking ordinary shielded wire, I replaced a couple of runs
with ordinary unshielded wire and have had no problems.

Getting the antenna tuner fixed let me concentrate on another problem.
The linear amplifier intermittently gave a false fault indication.
The fault circuit where it was happening is an incredibly complicated
thing which uses three comparator chips and a couple of transistors
to implement an AND gate.  The designer must have been proud of it, 
because it is labelled "AND GATE" right on the schematic.  Well the
comparator chips are open collector, rated to sink at least 6ma when
switched on.  The designer had chosen a pullup resistor value that
resulted in about 9ma of collector current, and one of the chips was
not sinking that much.  So the output voltage hung in the middle when
it should have been low, and that caused a fault indication to occur
intermittently.  Increasing the resistor value to reduce the current
sink requirement got rid of the false faults.

Then in the transceiver that drives this thing there is an audio amplifier
chip to run the internal speaker.  The chip was probably designed for
13V automotive use and is rated at 24 volts maximum supply voltage.
They had 28V available, so used four forward silicon diode drops to
reduce the voltage.  Four forward diode drops is not 4 volts.  Eventually
the chip blew.  Since it's no longer available (well, you can get them
on ebay from people in England) I decided to replace the chip with one
in current production.  Since the pin connections are different it's
ugly; but it's out of sight.  And I used a Zener diode to drop the supply
voltage to well under the maximum rating for the chip.

One has to wonder how much of our current military equipment in the field
has similar design and workmanship goofs.

Jim W6JVE

jhhaynes at earthlink dot net