[GreenKeys] Stevens-Arnold Polar Relay

[Don Sentz via GreenKeys]

Hi Jim,
Wow this is a great story- thanks!

 Yes I realized that the winding resistance seemed a lot higher than what I had seen in my books for other polar relays. I am pasting below an excerpt from my present design documentation for my version of the W2PAT TU described in the 1963 ARRL handbook. The work is on-going at this time.
-------------------------------------
I think I am ready to prototype this circuit, and test it. The changes relative to the W2PAT design are as follows;

•	Change the keyer cathode resistor 
       o	from 1000 ohms 1 watt to 1700 ohms ½ watt
•	Change the two keyer plate circuit resistors
       o	from 270 ohms ½ watt to 600 ohms ½ watt
•	Add 4240 ohms, 1 watt in B+ line after the 2nd power supply filter cap (3Kohm + 1.2Kohm standard values)
•	Use my Stevens-Arnold 1400-ohm relay, instead of whatever 400-ish ohm relay the W2PAT design used
-----------------------------------------

Thanks again for the great story- it is amazing to so quickly meet someone who knew Mr. Stevens and Mr. Arnold
-73
Don
--------------------------------------------
On Sun, 8/2/15, Jim Haynes <jhhaynes at earthlink.net> wrote:

 Subject: Re: [GreenKeys] Stevens-Arnold Polar Relay
 To: "Don Sentz" <dr.sentz at yahoo.com>
 Cc: greenkeys at mailman.qth.net
 Date: Sunday, August 2, 2015, 4:14 PM
 
 I remember the name
 "Millisec" associated with Sigma relays at one
 time.
 
 Those winding
 resistances are awfully high for use in ordinary TTY
 circuits.
 
 At
 one time - mid 1960s - Stevens-Arnold was making packaged
 tuning fork
 oscillators.  I was working for
 Teletype, where we were using unijunction
 transistors as bit-rate timers.  They had to
 be adjusted.  It seemed to
 me that a tuning
 fork oscillator might be a lot nicer - just plug in one
 with the frequency needed.  That's fine
 for transmitting, where the bit
 timer just
 runs all the time; but for receiving we need one that
 stops
 between characters and produces the
 first pulse a half period after being
 told
 to start.  I asked Stevens-Arnold if it would be possible
 to make
 a tuning fork oscillator with a
 hold-off of some sort so the oscillator
 could start and stop.  Next thing I knew was
 that Mr. Stevens or maybe
 it was Mr. Arnold
 told us he was coming to visit with a sample.  Must
 have thought we were thinking of something that
 every Teletype machine
 would use, not just
 the small quantity we would use for the high-speed
 tape product line.
 
 So we put it on the bench and he showed us that
 it could indeed be
 started and stopped and
 give us the first output pulse about half a
 period after being started.  And the asking
 price seemed quite
 reasonable.  But there
 just wasn't any enthusiasm in the company
 for changing the design to make use of it. 
 Perhaps because the cost
 of adjusting the
 existing bit timers fell mainly on the telephone
 companies that used the product, rather than on
 the factory.  Or maybe
 they were so stable
 that they almost never needed adjustment.