UK Vintage Radio Repair and Restoration Discussion Forum - View Single Post

G6Tanuki · 30th Oct 2020, 7:13 pm

"TU" = "Terminal Unit" - the box that takes an IF or AF RTTY signal from the receiver and converts it into an appropriate [typically +/- 80V, or 20-Milliamp current-loop] signal to drive the teleprinter.

"Properly" demodulating RTTY signals seems complex at first: if your input is an audio signal from a classic detector as you'd expect to use for listening to AM/FM/SSB signals you're really hamstrung and results will rarely be optimal.
Tapping the IF, down-converting to something like 15KHz, and _then_ separating the high- and low-tones with a pair of discriminators was the way-to-do-it; it gave you the ability to implement a 'local AFC' on the 465KHz-to-15KHz local-conversion oscillator, for a start. That was important if your receiver's HF local-oscillator was free-running!

[A deep discussion of DC-restorers and "Computers of Average Transients" to deal with the origin-shift when presented with repeat-patterns of 0s and 1s is inappropriate here]

30th Oct 2020, 7:13 pm	#17
G6Tanuki Dekatron Join Date: Apr 2012 Location: Wiltshire, UK. Posts: 13,998	Re: History of the Reception Set R. 209 MkI "TU" = "Terminal Unit" - the box that takes an IF or AF RTTY signal from the receiver and converts it into an appropriate [typically +/- 80V, or 20-Milliamp current-loop] signal to drive the teleprinter. "Properly" demodulating RTTY signals seems complex at first: if your input is an audio signal from a classic detector as you'd expect to use for listening to AM/FM/SSB signals you're really hamstrung and results will rarely be optimal. Tapping the IF, down-converting to something like 15KHz, and _then_ separating the high- and low-tones with a pair of discriminators was the way-to-do-it; it gave you the ability to implement a 'local AFC' on the 465KHz-to-15KHz local-conversion oscillator, for a start. That was important if your receiver's HF local-oscillator was free-running! [A deep discussion of DC-restorers and "Computers of Average Transients" to deal with the origin-shift when presented with repeat-patterns of 0s and 1s is inappropriate here]