[Collins] KWM-2A low output solved

[Gerald]

On Thu, 2005-08-25 at 13:21 -0400, Bill Abate wrote:
> Some months back I noticed I had lower than normal output from the 2A on 
> 75.  A post to this group suggested I verify the meter circuit 
> calibration.  I was also reminded that the meter looks at cathode 
> current which includes screen current so my input was less than I 
> thought.  So when the 30L-1 started to act up recently, I decided it was 
> time for Collins maintenance.  The 30L-1 was misbehaving when I checked 
> into the CRA net.  Turned out to be the HV supply.  Yup, the 
> electrolytics had died.  Fortunately I bought a set of 120 mfd @ 450 
> from a JA at Dayton.  They are the PC board mount type but fit the pc 
> board clips nicely.  The L-1 is quite happy now.
> 
> Next on the list was the 2A.  Got it out of the cabinet and just 
> marveled at it.  The teflon wiring is just so neat and compact.  I wish 
> I could operate it upside down, without the cabinet.  Anyway, the first 
> task was to check the plate (cathode) meter calibration.  I prefer using 
> a constant current source for calibration over measuring the cathode 
> resistance and then measuring voltage across the cathode resistors and 
> figuring out cathode current.  This way I don't have the 2A on when I'm 
> working on it.  The 6146's have to be plugged in though.  I adjust the 
> current source for 200 ma. through the cathode circuit and check the meter.
> 
> Normally this is easy except that the 2A switches the meter from S-meter 
> to plate with a relay.  It's normally connected to the S-meter circuit. 
>   I didn't want to unsolder the coil and energize it externally (who 
> would change a work of art?).  There had to be an easier way.  So I 
> pulled K-4 relay.  The relay pins are flat.  So I merely took a short 
> piece of #16 buss wire and flattened the ends with a pair of vice grips 
> so that the thickness was the same as the relay pins.  I made two 
> jumpers that took the place of the normally open contacts from the 
> cathode circuit to the meter.  Just plugged them into the relay socket. 
>   Worked like a charm.  The bad news is that the meter calibration was 
> right on.
> 
> With about 225 watts input I was getting about 65 watts out through my 
> Bird watt meter.  Less than 30% efficiency.  One problem was the band 
> switch.  Seems I had not exercised it in a while.  Could not get DeOxit 
> on all of it because of close quarters and shielding.  So I just rotated 
> it about 20 times.  That got the output to about 75 watts.  Still not 
> good enough.  Then I found it.  The loading cap calibration was the 
> culprit.  The 50 ohm setting was wrong.  I experimented with slight 
> changes in loading and got the output to 115 watts.  Now I was happy.
> 
> There is a calibration procedure in the manual for the loading cap.  I 
> could follow what Collins wanted but I did not understand the why's of 
> the procedure.  Since the 75M position has its own calibration padder, 
> C-154, I merely found the best value of loading.  I then set the loading 
> lever to 50 ohms and adjusted C-154 until I got the same output into the 
> dummy load.
> 
> The 2A is humming nicely now.  I'm sure there are some more 
> knowledgeable people than I that can explain the manual procedure and 
> the reasons for it.  I welcome that but this worked for me and I thought 
> this might help someone, especially the meter calibration.  Note that 
> there are 4 air variables associated with the loading cap and there is 
> some interdependence depending on what band you are on.  Since I 
> normally use the 2A on 75, I did not check the other bands.  Its 
> interesting to note that the manual states that the loading is quite 
> broad but that is not what I found into a dummy load.
> 
> 73,
> 
> Bill, K3PGB

Best I remember the tuning instructions, one always starts an unknown
load with the loading capacitor at maximum C, then reduces the C until
the desired loaded plate current is reached. One alternately reduces
loading C and dips plate current. And that is exactly how the autotuned
Collins transmitters tuned. If one loads for higher than rated plate
current one should find reduced fundamental output, sometimes increased
harmonic output. Getting a TVI report from 2 miles away hitting channel
2 audio from 10 meters with a Viking Ranger showed that to my dad one
time.

When the padding capacitors are properly adjusted the loading control
should show at 50 ohms on each band while feeding a good 50 ohm load.

The manual's statement that the loading is quite broad means that it
will load a fair range of impedances, not that its uncritical to
adjustment.

For linear amplifiers I prefer to tune for maximum output rather than
for plate (and grid in grounded grid amps) current(s). It is quite easy
to tune for the specified plate and grid currents and get very little
gain or power output but to fry the tubes. When tuning for maximum
output, one achieves the optimum load for maximum efficiency and
linearity. At one MARS station with a Henry 2K3 (driven by a KWM-2A),
when I retuned for output from the settings the station sergeant had
done for plate and grid currents, I increased the output from just under
200 watts to about 2 KW. Enough to be immediately noticed in 'Nam the
other end of their paths.
-- 
73, Jerry, K0CQ, Technical Advisor to the CRA
All content copyright Dr. Gerald N. Johnson, electrical engineer