[R-390] wow... the R-391A does exist - Video Update

[Barry Hauser]

Bob wrote:

> Can anyone explain how the R-391A auto-tuning is accomplished?

Not completely, but I can give you an idea and where to go to get the whole 
story.

>Given the technology available at the time, I'm guessing it might counts 
>pulses from a cam-actuated electromechanical contact, with one or more 
>stepper relays as the memory element.

Not exactly.  There are cams of sorts and relays, but no pulses in volved. 
It is totally electromechanical and has no electrical/electronic conneciton 
to the receiver, which is really a stock R-390/URR all respects.  All the 
modules are the same.  The unique components consist of two positioning 
heads -- one for the MC and one for the KC. They are identical except that 
the MC unit has the detent spring on it.  There is a motor, a sealed relay, 
a drive shaft geared to the motor with three worm gears on it.  There is 
also a control head that contains a couple (or three) rotary switch wafers 
and a one or two small relays (as I recall) ... and there's the channel 
switch.  The mainframe/chassis is the same as a regular R-390/URR with a few 
exceptions -- the wiring harness behind the front panel, the 28 vdc supply 
connection (though I think R-390's have this too on the 4-pin power 
connector), and the front panel has three mods -- hole and markings for the 
channel switch and a peephole for the channel indicator which is a 
wheel/dial with 1 through 8 on it, mounted to the front of the control head. 
(The third thing is that plastic card for marking down which channel is what 
frequency.)  The three worm gears drive the MC and KC positioners and the 
control head.  All the autotune components are mounted on a cast frame 
that's behind the lower 1/3rd of the front panel.

The autotune doesn't even draw any power from the receiver -- an external 28 
vdc supply (probably 7 amps or more) is needed.  The KC and MC positioning 
heads are strictly mechanical, with a stack of spring-loaded pawls with 
hook-shaped ends which catch into disk-cams and there's a clutch affair. 
(Paul Anderson described some of this in an earlier post.)  There are no 
electrical connections to these positioners.  Wiring is involved in the 
motor, control head, sealed relay and the channel switch.  To use the 
autotune, you first tighten the locking keys on both the KC and MC knob (or 
else!) and then change channels.  As soon as the channel switch is changed 
from the channel shown in the window, it triggers the cycling.  When you 
loosen the locking keys and manually tune the rig, you are changing the 
setting for the channel that is showing through the little window.

That's as far as I can go.

> But I know the auto-tune transmitters use a phase comparator and 
> closed-loop analog feedback, so an all-analog design using a multi-turn 
> pot would be another possibility.  Either method could  explain why it 
> seems to drive to a "home" position and then reverses direction.

Has nothing to do with it.  That much I can tell you.  Beyond this, unless 
Paul A. is up to it ;-), I suggest you check out the R-391 manual which is 
on the Bama site.  It has an explanation of the internal workings of the 
beast.  I found some of that hard to follow.  The explanation in the 
R-105/R105A (ARR/15(a)) manual is somewhat clearer and the design is very 
similar, though it is driving a much different rig -- and all R-105's are 
autotuned. The manual for that radio is also on Bama.  The positioning heads 
look almost identical inside.  They remind me of stuff you'd find in an old 
mechanical adding machine or calculator -- either that or a combination 
lock.  Keep in mind that old technology.  Remember the more complicated 
mechanical calculators that would multiply and divide - -I think they all 
had carriages, like typewriters that went back and forth?  Remember doing 
division on them? It would take ten-15 seconds to get the answer to 4 over 
2.  Well maybe not that bad, but the consequences of all the stacked parts 
'n pawls on common shafts was similar when the fine machine oil turned to 
glue or grit got into the works -- not only wrong answers, but they wouldn't 
stop going katchokity-katchokity and the carriage would keep slamming down 
one end.  All you could do was pull the AC plug.  With the '391, when the 
same thing happens and the autotune slams against the 10-turn stops, the 
same emergency action is called for -- except you have to cut the DC.

So, after all that, in brief, Bob, it's not even up to "analogue" -- it's 
relay and mechanical "logic".   And, as was the case with those 
electromechanical calculators, very often we'd reach for the keyboard and go 
"uh-uh -- nah" and do some long division with paper and pencil.

Barry