[Johnson] Rotary switch issue

[Robert Nickels]

On 2/12/2023 3:06 PM, John wrote:
>
> Can you see flex along the shaft that isn't allowing the farthest 
> wafer to go "home" to its correct orientation?  I suppose this could 
> come both from the shaft itself AND/OR from the stack of switches 
> itself (which must be in synch with the detents on the front) which 
> might flex too.

All that and more, John, see below for more detail.

Thanks for all the ideas and suggestions.  I thought that loosening and 
re-tightening the nuts on the long screws  made sense but the Pacemaker 
bandswitch is a little different in that the metal standoffs between 
switch wafers are soldered to the partitions between stages.   This can 
be seen in some pics I've posted here:

https://share.icloud.com/photos/0b1iGFSZNAmePaIWBhLNvbeEw

Since each section of standoff is rigidly attached to the chassis this 
kind of ended my thinking that the wafers had rotated.   I started 
studying just how this bandswitch worked...but a brief diversion...

Since this is a Johnson list,  the following are my thoughts about the 
Pacemaker as an example of early SSB tranasmitter development against 
Rodger's comment that it isn't EFJ's best effort...feel free to read or 
disregard as you wish!

The first of anything is always the hardest, and since Johnson abandoned 
phasing for filter type SSB generation maybe they felt that phasing was 
the source of the Pacemaker's problems.   But its SSB generator is right 
out of the same books others used successfully and once aligned (which 
modern test equipment makes much easier), it does the job as well as 
most of its kind, which is to say "good enough for casual operation" but 
not the level of suppression you can get for the best phasing rigs.  Of 
course it was much more challenging given the knowledge and test 
equipment of the late '50s.

The real problem is the mixing scheme, especially on 80 meters...and I 
haven't got a clue what the Johnson engineers were thinking.    
Developing SSB "on frequency" like the B&W 51SB-B or Heathkit SB-10 did 
would have been a straightforward way to make an all-band SSB 
transmitter - all they needed to do is to add a SSB generator to a 
Navigator, or a Ranger sans modulator.   But EFJ came up with the notion 
of using a low-frequency VFO that tunes 3.0 to 3.5 MHz and developing 
SSB on the required frequency, to mix with the VFO and produce output on 
the various ham bands. In so doing they created a white elephant of a 
transmitter that has neither the straightforwardness of the on-frequency 
approach or the advantages of the Norgaard/CE heterodyne method.

With only 500 kHz separating the desired mixer output signal from the 
VFO on 80 meters,  the tuned circuits can easily be mistuned to the VFO 
frequency rather than the desired mixing output, and the image is also a 
problem.   Johnson's approach works better on the higher bands where 
tuned circuits can reject the unwanted mixing products,  but VFO 
harmonics present a different concern. For 80 meters the book even 
cautions about mistuning and that's how  I spotted this bandswitch 
problem while using a spectrum analyzer to verify the tracking of each 
of the three gang-tuned exciter stages because normal receiver type 
tracking alignment just wasn't making sense.     Understandable since 
the exciter tuning capacitor wasn't even being connected!

It's easy to understand the attraction of tracked exciter tuning, as EFJ 
had done in the high-end 500, but it makes one-knob bandswitching 
especially complicated in a sideband transmitter. The main drive pulleys 
behind the panel knob have to operate three separate switch mechanisms, 
not unlike the lineshaft in an old factory.   Straight ahead is the 
bandwitch in the PA compartment, which has it's own detent and hard 
stop.  There's also one of EFJs clever cams that operated on 10 meters, 
and two dial cord pulley drives, one that runs through the main chassis 
to yet another bandswitch on the SSB oscillator box, and the one in 
question which drives the long four-deck bandswitch for tuned circuits 
in the balanced modulator and three exciter stages.    I made a little 
video to show the delay in operation as the knob is turned, but close 
observation showed that while the wafer closest to the drive point 
operates properly and the mixer (#1) is mostly engaged, but the 
amplifier #2 contact is only about half-engaged and the rear-most driver 
contact is barely touching (if that) at extreme CCW rotation of the 
shaft.   For the higher bands the ball detent mechanism properly aligns 
the shaft, but on 80 meters there is a hard stop that prevents the shaft 
from moving past the detent.

I decided to bend this stop out of the way to allow the shaft to rotate 
enough to engage the rear contacts and if so, to grind off part of it.  
However rotation was still limited by a hard stop on the switch in the 
PA compartment.   Once it was bent it away I could "over-rotate" the 
shaft enough to properly engage all contacts in the CCW 80 meter 
position.   With the stops temporarily out of the way there was enough 
slack that I could loosen and pre-load the dial drive drums that operate 
the exciter bandswitch and gain enought rotation so the rear wafers were 
properly engaged with the 80 meter contacts.   The ball detent still 
works like it does on all other bands.      Bottom line: after fiddling 
with the drive pulleys I got the bandswitch to work.

All said and done, you have to be a bit of a masochist to want to run a 
Pacemaker ;-)   In my case, because I had the Johnson 10 watt audio 
amplifier I wanted to use the Pacemaker to drive my desk Kilowatt rather 
than the usual Ranger, so I could play vintage SSB with it once a while. 
   Plus, I just thought it would be cool to replicate the EFJ setup from 
the color photo in  the EFJ 1957 catalog.   All the Pacemaker has to do 
is switch the audio and provide the required amount of RF drive, which 
should be easily done by virtue of the carrier level control.   Which it 
did nicely for some time, and since I'm not going to be changing bands 
all that often I'm hoping it will continue to do so.

73, Bob W9RAN