[Flexradio] The inelegant Keying solution

[Tayloe Dan-P26412]

I think your basic problem with your solutions has been you are trying
to minimize the current load to the keyer.  This small current shorting
to ground is the true source of your problem.

The real solution is to increase to current.  Moving to a relay does
exactly that, but a mechanical  solution like a relay will have contacts
that will eventually have the same low current contact problem.  Someone
else posted the same contact problems on a receive only application (no
current drain) that could be fixed only by using mercury wetted contact
relays.

I suggest instead returning to one of your earlier buffering experiments
and greatly **increasing** the current through the contacts to perhaps
20 to 50 mA.  Experiment with what current it takes to be reliable.

I would suggest a simple 2n7000 with a gate pull up resistor to +12v
such as 470 ohms (~25 ma).  The paddle grounds the gate, drawing 25 mA
through the paddle contacts.  Since this is an inverter stage, a second
stage is needed to get a non-inverted signal. Use a drain resistor of
~10K on this first MOSFET, and drive a second 2N7000 whose gate is
connected to the drain of the first 2N7000.  Bypass both gates with a
0.01 uf cap to ground for RF bypass.

For both sides of the paddles, that is four 2N7000s, four resistors (2
10K, 2 470), and four caps, a rather minimal circuit.

Alternatively, you could possibly look at the keyer input of the
SDR5000/SDR1000 and add pull up resistors right there at the jack to
either +5v or +12 depending on what the input circuit needs to see as a
"high" voltage using a resistor mounted internally that sets the current
to a high enough value.

The higher contact current should keep the contacts from oxidizing and
make things more reliable. 

- Dan, N7VE

-----Original Message-----
From: flexradio-bounces at flex-radio.biz
[mailto:flexradio-bounces at flex-radio.biz] On Behalf Of Lee A Crocker
Sent: Thursday, July 03, 2008 8:38 PM
To: Flexradio
Subject: [Flexradio] The inelegant Keying solution

I normally run my F5K with an external keyer through the serial port,
which is a hold over from the SDR-1000 days.  It made for an easy way to
switch the keyer between 3 radios, by merely opening up PSDR SETUP and
turning on the correct serial port to the radio I wanted to key.  I
tried using the internal PSDR/F5K internal keyer using my Begali paddles
and found some intermittent trouble with the keying, dropped elements or
the keyer would hang on a string of dits or dahs and would not turn off
till I hit the paddle again.

I have been working on different solutions to this keying issue with the
internal keyer.  I have built various circuits, trying to find a simple
solution.  I have a Begali Graciella paddle and it has gold contacts.  I
have a Begali Simplex Mono which has non gold contacts, and neither are
they coin silver.  The problem appears to be that the amount of current
pulled from the KEY input at the nominal 3 volt voltage on the Key port
makes my Begalli keys behave as an incompletely forward biased diode and
not a conductor, and that signal appears to be ambigious to whatever is
sensing it inside the radio.  I tried using such things as PNP
transistor buffers on the KEY input on the F5K to no avail.  The
junction of the transistor behaved the same as the paddle itself.  I
tried fooling around with a NE556 to act as a signal
conditioner/debouncer and didn't like the result of that either.  I
wired up a couple of 12v reed relays and this seems so far to have
 cured the problem.  I measured about 5ma current drawn through the
relay coil using a 9V battery and at that current level the Begali
operates in an unambitious way to the relay coil, and the relays thus
far have keyed the KEY input flawlessly..  So this maybe a simple yet
inelegant way to solve the keying issue.  

I used a couple of relays out of the junk box, but Radio Shack does sell
some 12V reed relays that should work fine.  The circuit is basically V+
or the + of the battery goes to the top of each relay coil and then each
coil goes to either the dit or the dah of the paddle and then the ground
of the paddle goes to the - side of the battery or ground.  One side of
each relay switch goes to ground.  The other side of each relay switch
either goes to DIT on the KEY jack or DAH on the key jack.  For the
relays I used polarity is not an issue, but some relays have built in
protection so you may have to watch that.  The battery only draws
current when the contact is closed, so that part is elegant, and it will
cost under 10 bux to build, which beats the hell out of some 40 dollar
micro-pic solution.  Eventually I will hook it up to the station 12V
supply and probably rebuild it inside a pill bottle with a 1/4 inch
stereo phone plug but I wanted to make sure it
 was going to work before I went to that trouble.  If you build it the
way I describe, the 12V will be dropped through the coils and will be
self protecting, so if you get a screw driver across the paddle contacts
it wont be like shorting out your 12V 35 amp power supply to ground.
(Mine is 75 amps so it would be a hell of a bang)

This may be a nice little project for the 4th.  My final parts list will
be 1/8 stereo phone jack, and 2 x N.O. 12V reed relays and one 9V
battery and one stereo 1/4" phone plug  a little piece of perf board and
a pill bottle to stuff int all in,   

V+-----|\|\|\|\|---------/ .--------------gnd
bat+         coil     paddle dit                bat-

F5K dit--------------/ .--------------gnd
KEY input            dit  relay

Sorry for my poor attempt at a schematic  This represents the dit side
of the circuit and will be duplicated for the dah side.  

73  W9OY


      
_______________________________________________
FlexRadio Systems Mailing List
FlexRadio at flex-radio.biz
http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz
Archives: http://www.mail-archive.com/flexradio%40flex-radio.biz/
Knowledge Base: http://kb.flex-radio.com/  Homepage:
http://www.flex-radio.com/