[GreenKeys] GreenKeys Digest, Vol 74, Issue 41

[John Nagle]

> Date: Fri, 26 Mar 2010 15:07:01 -0700
> From: gil at baudot.net
> Subject: Re: [GreenKeys] TTY-Connect PCB Wanted
> To: "Steve Ripper" <steveripper at comcast.net>
> Cc: greenkeys at mailman.qth.net
> Message-ID:
> 	<20100326150701.617ae442d0f8f631a9b3312d6b10cde7.222e282329.wbe at email.secureserver.net>
> 	
> Content-Type: text/plain; charset="utf-8"
> 
> Hey Steve:

> As for a new tty-connect board, I was planning to try a current-source
> design for the high-voltage loops, to maybe cut down the standby power
> and heat dissipation.  Anyone have a current-source design they have
> used for tty?  I would only change the 60-mil HV loops, as the LV loop
> is fine as-is.  I was also considering adding more 1/4" jacks to the HV
> loops.
> 
> Another thing I want to add are drivers and connections for motor
> relays.  I have also been considering adding an ethernet port, but I
> don't write pc application code, so I'm not sure how useful it would be
> to have a tty-connect with a tcp-ip socket port, if there is no app to
> connect to it.  Anyone write that sort of code?
> 
> The primary purpose of tty-connect is to provide an ascii-only computer
> interface, to get away from the special hardware and apps to drive
> baudot out of a pc serial port (this gets harder each year, although it
> is done quite well in heavymetal and rtty-art).  Of course, you can
> leave the PIC off of tty-connect and jumper the opto-isolated 232
> section to a tty loop section -- then it becomes an interface board for
> an app like heavymetal.
> 
> But if the baudot-to-ascii-converting PIC is installed, a logical
> extension of tty-connect's ascii serial interface is an ascii socket
> connection on an ethernet port -- however, it is all a matter of network
> applications, which don't exist, and I can't write.

     I have a board design available at

	http://www.aetherltd.com/connecting.html

Schematics, board artwork, parts list, etc. are all freely available.
This is for 60mA current loop Teletypes.  You can still buy 20mA current
loop interfaces commercially, so I didn't bother supporting that.  An
external 120VDC supply is required.  I've bought 120VDC 200mA open frame
linear supplies commercially.  I included a motor control relay circuit;
I use a solid-state power relay for motor control.

     I don't sell parts, boards, kits, or assembled units.  Just send
the board design to PCBexpress and the parts list to DigiKey.

     For connecting to a computer, I use the Silicon Labs CP2102, a common
USB to serial converter part which can be configured for 45 baud.  See

	http://www.aetherltd.com/connectingusb.html

This runs with Windows, Linux, desktops, and laptops.  I normally use
an EeePC 2G Surf, a tiny subnotebook, to run a Teletype Model 15.

     This is packaged in a brass and wood case.

	http://www.aetherltd.com/aesthetic.html

(I'm into steampunk, not amateur radio.  I only do nostalgia for stuff
older than I am.)

      For software, I've written "BaudotRSS", which connects a Baudot
Teletype to RSS feeds and SMS messaging (via Google Voice).  That
software is available at

	http://www.aetherltd.com/communicating.html

with the source on SourceForge.

      The heat dissipation in the ballast resistor remains
annoying.  I have a simulated design for a solution to that at

	http://www.aetherltd.com/public/flashsupply6.asc

which needs the free version of LTSpice to run.

	http://www.linear.com/designtools/software/

This is an attempt to design a 60mA 120V current loop supply to drive one
Model 15 selector magnet from a USB port, without external power.  You
really only need 120V for the first millisecond after a SPACE to MARK
transition.  The steady-state voltage needed, once the 4H inductance
has been overcome, is only 55 ohms * 60 mA = 3.3V.  So this design
uses a photoflash IC as a switching power supply, to step up the 4.8V from
the USB port to 120V and charge a capacitor.  When the input changes to
MARK, the capacitor dumps into the selector magnet.  There's also a 3.3V
regulator to provide the sustain current, which takes over when the
capacitor has discharged.  Only 250mA is required from the USB port,
and you can take up to 500mA from a USB port.

This is just a proof of concept simulation, to demonstrate that such
a device could be driven from a USB port.  The energy budget works.
(At least at 60WPM; at 100WPM, there isn't enough SPACE time to
charge the capacitor.)

So there are some more modern designs.

				John Nagle