[Johnson] Pacemaker RF Phasing Unit

[Robert Nickels]

On 12/31/2014 5:37 AM, David C. Hallam wrote:
> I have a Pacemaker sitting on the back shelf in my workshop with the 
> same problem.  It's been sitting there ignored for a couple of years 
> after I gave up on getting 90 degree phase shift on 20M

Having found and fixed the problem, I thought I'd post a few notes here 
for David and others who might encounter this issue in the future.

Once I overcame my trepidation and popped off the dial cord drive, the 
module can be easily accessed and tested without unsoldering anything.   
As expected, both the coil and 68pf capacitor were good, and of the 
correct value to create  the required leading and lagging 45 degree 
phase shifts at the 11 Mhz mixing frequency used for 20 meter SSB 
generation.  Since adjusting the slug had no effect, I did expect I 
might find a bent or broken switch contact but that proved not to be the 
case either.    The problem, as often is the case, was 
self-induced...and having to re-string the dial cord was suitable 
punishment.

Johnson specifies that the crystal oscillator slugs be adjusted to 
produce a 1.5V level at the output of the phasing network.   I thought I 
did so when I went through it previously but I also remember using my HP 
410C VTVM for some projects before realizing it had drifted badly out of 
calibration, so that might be a factor. In any case, the problem was 
caused by excessively high output from the 11 Mhz oscillator, and 
reducing it to the level specified allowed the unwanted sideband to be 
suppressed.   For some reason, the 11 Mhz crystal produced MUCH more 
output than the others to begin with, so that is why it may be more 
problematic.

I did a little playing with the levels and found that when Johnson says 
1.5 volts, they mean it!     Since most of us have decent scopes now, 
adjusting the output levels to 4.2V P-P on the scope is probably easier 
than messing with an RF Probe and VTVM.   I found that "less is more" 
when it comes to sideband suppression; increasing the output level to as 
little as 5 V P-P was enough to reduce opposite sideband suppression by 
10dB or more. (Incidentally, if you're not using an SDR as a spectrum 
analyzer to align phasing rigs, it's the best $50 investment you can 
make - drop me a private email for more information).

After carefully setting the carrier oscillator levels to 1.5VRMS, I was 
able to obtain the specified 30dB sideband suppression on all bands.   
So,  removing the phasing assembly wasn't reallly necessary after all, 
but if you do so, proactively replace the paper plate bypass cap on the 
cathode follower stage while you're at it.   And take good notes/pics of 
how the dial cord is routed as none of the dial cord patterns are 
documented in the manual.

The Pacemaker is far from Johnson's best effort.   Maybe they were 
trying to avoid patents, but you don't even need a slipstick to see that 
the 7th harmonic of the 3 Mhz VFO is going to fall right at 21 Mhz and 
will go right through the transmitter along with the desired  signal.   
30dB of sideband suppression into a KW amplifier isn't going to make you 
very popular on a crowded band but no one is going to run the Pacemaker 
as a contest station anyhow ;-)   The addition of a simple low-pass 
audio filter would have helped to restrict the audio bandwidth but 
Johnson left it out.    Based on a quick sweep, the output only drops 
about 10dB from 1 to 5 kHz (don't tell the hi-fi SSB guys!).   Using a 
Shure 444 or other type restricted range mic should help produce a clean 
SSB signal, along with proper loading and drive adjustment.

The Pacemaker is kinda like owing an AMC Pacer - you don't do it because 
it's the greatest design or most popular, but because it's unique and 
still does the job despite its quirks.    A big thanks to Lee Craner for 
his diligent efforts in compiling useful documentation and for creating 
a vastly improved schematic of the audio/RF phasing generator.

73 and HNY,

Bob W9RAN