[Milsurplus] Command Set Transmitter Keying

[Mike Hanz]

Mike Morrow wrote:
>> Steve, KB4DMF, was preparing a complete SCR-274N set ... he ran into an 
>> interesting chirp problem with keying the entire set as it was 
>> originally meant to be.  
>>     
> Steve's write up is an interesting discussion of what happens when (and
> why) the HV keying relay K52 stays energized longer than the antenna
> transfer relay K55, which results in the antenna being disconnected
> before the transmitter stops RF generation.
>
> The real problem, it appears to me, is the delayed dropout of K52.  Steve's
> installation of a DC blocking diode in the K52 coil circuit seems like
> a good way to eliminate dropout delay.  But it seems to me that installing
> similar diodes in the other relay paths would be unnecessary.  I would
> have tried installing a resistor of relatively low value at K52's coil
> terminals at the easily accessible top of the modulator, which could help
> with K52 field collapse without using a diode.
>   
.
I'm afraid we were up against a hard stop for shipment because of the 
Memorial Event.  The difficulty with a simple series resistor for K-52 
is that all of the paralleled relay coils add to the offending current 
sustainment mechanism...their contribution depending on the amount of 
stored energy each contains at the time of voltage cutoff.  It is 
somewhat of a empirical science, and we didn't have the time to examine 
the back current pulse from each relay and determine if there was one in 
particular that was contributing the most to the problem.  Since there 
is a common node at the junction of all the coils, it was *all* of them 
to one extent or another, so isolation was the quickest solution and had 
the added advantage of yielding the fastest possible keying speed if 
someone got on the station who was a whiz at a straight key.  I also had 
in mind that diodes were a much less intrusive modification and could be 
hidden in the cable harness more easily than to try to modify the 
BC-456, but Steve decided to do it as shown for the reasons he cited.  
Can't blame him - the cables were already made up by that time.

> I'd also agree that this would seem to be an inherent characteristic of
> the design, with only one observation.  If the relay opening spring
> force of K52 has weaked over the last 65 years, that would also increase
> the time it takes for K52 contacts to open. 
.
That's a reasonable analysis, Mike - a potential I immediately examined 
when Steve originally ran into the problem!  I spent a couple of years 
designing mechanical relay testers for nuclear weapon QA, so I had some 
memories to go by, despite their gradual slippage into the mists of time...

> It would be interesting to be able to compare the drop out time of a new K52
> versus that of all the 65-year-old examples that exist today...but that's
> obviously not possible.  
.
Well, true, but I had several NOS and well used samples of K-52 to put 
through a timing test with a 28v supply and series transistor driven by 
a pulse generator.  The clacking drove me nuts after a while...   :-)   
Out of 8 samples, they all came out the same within less than 10% 
variation in pull-in and dropout times, which is fairly typical for a 
mass produced relay of its type. 
.
Anyway, the problem was certainly unanticipated, but in retrospect, I 
don't know why we wouldn't expect it.  Probably because no one has used 
CW in a complete set like this for years.

73,
Mike  KC4TOS