[Heathkit] HW-12a

[Robert Nickels]

On 2/18/2020 8:15 AM, LITTLE MAIL wrote:
>   There's no obvious sign of a failed component

That's the problem, one common failure in radios of this vintage is 
carbon comp resistors that have increased in value and yet they look 
perfect.   I had one in a HW-32 that caused the het oscillator/mixer not 
to work for example, although you won't have that stage in a "12".    
Assuming you have a DMM, you can measure the value of resistors without 
unsoldering them as a first step. Disconnect the transceiver from the 
power supply.  Still, parallel paths will sometimes cause a reading that 
is much lower that the marked value, just note these for follow-up 
later.   If the measured value is within the marked value, the resistor 
is good. If it measures more than about 25% high, it's still close 
enough to work in most circuits with no impact.  Those that have changed 
in value will often be as much as 5 to 10 times the marked value and 
definitely need to be replaced.

The first thing to do in fixing any radio is to take the time to 
understand how it works.   Fortunately Heathkit provides good detailed 
theory of operation, go over it stage by stage til you know what signals 
should be present, then verify.    I always clean tube pins as well and 
visually inspect the PCB for anything that shouldn't be there.  More 
than one radio has been repaired by peeling off a bit of solder splash 
that a previous repairman inadvertently created.

The next step is to power up and do the voltage checks against the 
voltages shown in the diagram in the manual.   This will verify wiring 
connections and clear up any of those questionable resistor values - I 
just prefer to check the resistors out of circuit first because it's so 
easy to do.      The last tip I can give you requires an oscilloscope or 
spectrum analyzer (see next para).     Remove the 6GE5 PA tubes 
temporarily and connect your scope probe to the grid of one of them; 
this will let you see even a low-level RF signal that may be present.   
Remember that a sine wave tone injected into the mic input of an SSB 
transmitter will produce a CW signal output.   With the finals removed 
you can keep the radio in the transmit mode as long as necessary without 
worrying about damaging the tubes.  You should end up with somewhere in 
the range of 30-50 volts peak to peak RF at the grids.     When you do, 
then re-install the finals, set the bias, and touch-up the transmitter 
alignment.

  It's really difficult to troubleshoot a sideband transceiver without a 
scope, but a $20 rtl-sdr dongle and free SDR software with a DBFS scale 
makes a decent "poor-man's spectrum analyzer" once you know how to use 
it and which RF signals you're looking for.    DBFS readings are 
relative, i.e. if you see a signal go from -50 to -40 dB, that is 10 dB 
of gain.  So you can use it to verify that desired frequencies are 
present, that undesired frequencies are not present, and to compare the 
relative amplitude of signals. Never directly connect the SDR input to 
the radio! Use a few turns of wire as a sampling loop antenna and reduce 
the gain to make sure any signals you are looking at are well below the 
top (0 dB) of the scale to make sure the receiver is not being 
saturated.   You can also slide a tube shield over a tube, making sure 
it is not grounded, and pick up the signal from a stage that way because 
the SDR receiver is very sensitive.    This inexpensive tool is better 
than the lab equipment at Heathkit when these radios were designed!

73, Bob W9RAN