[GreenKeys] German WX service via SDR & Websdr...

[Jeffrey Angus]

On 8/3/17 9:21 PM, Richard Knoppow wrote:
> A clear explanation. Its obvious that the discriminator type detector 
> is exactly
> identical to the FM detector used for voice or music modulation. It 
> should be
> optimized for the deviation of the transmitter, which is the same as 
> the amount
> of mark to space shift for teleprinter. It seem to me that a universal 
> converter
> should have two or three discriminator widths for various shifts. 
And the modulation bandwidth.
The "bandwidth" minimum would be 45.45 Hz for 60 WPM TTY.
In reality, it's actually 7 times that because it's a square wave and 
the extra
band width is required to recover a decent approximation of the square wave.
45.45 x 7 = 318 Hz. 850 + 318 = 1168 Hz.
This is why "RTTY" filters in receivers at the time were 1200 Hz wide.
The problem with discriminator detectors is there's a whole chunk of 
band width
in the middle just begging for noise to interfere with things. Roughly 
530 Hz.
It's a bit better with 170 Hz shift.
170 + 318 = 488 So a 500 Hz filter would be fine.
And here is why things aren't narrower than 170 Hz "normally."
318 / 2 = 159 Hz. This puts a portion of the spectral energy of the Mark 
Signal
in the Space bandwidth and vice versa.

Now you can see why 10 Hz shift is a problem.
You still need a 328 Hz bandwidth.
And the frequency stability of both ends become very critical.

The HAL ST-6000
<https://www.radioexperimenter.us/rm-1979-05/images/1614_900_229-hal-st6000.jpg>
Took all this into consideration and adjusted things accordingly.
It was still important though to have the receiver bandwidth set 
appropriately.



-- 
Jeff-1.0
www.foxsmercantile.com