[GreenKeys] RTTY Answerback Unit

[jim at k6ccc.org]

Alf Fisher said:
> One of the essential features of a machine for use on i-Telex is that it 
> has an answerback unit and I know that a workaround has been developed 
> by the designers.
<snip>

> Has anyone had experience of building an electronic answerback unit that 
> could be initiated on receipt of the code.

I did almost 40 years ago using TTL logic and an ERPOM for the answerback outgoing responses.  Sorry, I have no docs on what I did, so only my memory (which ain't that good anymore).  The triggers all started with my callsign, followed by the letter Z, followed by one more letter.  For example, WB6WMWZW would trigger a "Who are you" response.

If I remember right, the current loop passed through an opto-isolator to convert it to TTL logic levels and then a UART to convert it to 5 bits of parallel data.  That then went into a series of shift registers and gates that compared that to the desired commands.  Sorry, I can't remember the detail on this part.  Once a valid command was recognized, a binary counter was set to a specific number depending on what the command was.  That would set a start address for the EPROM.  Each byte in the EPROM had 5 bits for data, one to continue to the next character or stop, one to key the transmitter, and one to enable the motor on a Model 28 RO.  The 5 bits of data would go back through a UART, and then a buffer amp that keyed a reed relay that keyed the loop.  The transmit bit was just that - it caused the 2M transmitter to key up by way of a buffer amp and a relay.  The printer bit was similar to the transmit bit except that it controlled power to the Model 28 RO.  The continue / stop bit when it had the continue value would cause a clock to cause the binary counter to increment which would cause the next character in the EPROM to be sent.  Once the continue / stop bit changed to the stop value, the logic would reset everything back to the idle state and wait for the next command.

My normal practice was that when a valid command was received, transmit and the printer would be enabled and a few [LTRS] characters were sent, then the response.  Normally after the first line of text, the printer would be turned off.  That gave me a log of what commands were received fairly easily.  I have a vague recollection that each response either started or ended with a date / time stamp, but I have no recollection how I did that - or maybe I thought about it a ruled it out as too hard to implement in TTL logic...

> In building your own, I don’t think there is any good reason why it 
> couldn’t have more characters than the normal 20.

As I said, mine were all 10 character commands and all started with the same nine:  W B [FIGS] 6 {LTRS] W M W Z  Changing that would have made it more difficult and required more TTL.

73
-----
Jim Walls - K6CCC
jim at k6ccc.org