TNF vs RTTY question

DrTeeth

When I create a TNF at the default width in the middle of the two RTTY tones, why does it seem to almost fully quieten the RTTY signal even though most of the two tones are outside the filter's bandwidth?

Larry Loen WO7R

I don't know what theory allows but I usually have a little added bandwidth on my rtty sigs including between mark and space. Sideband even if faint seem to matter.

Walt

Say Guy, I sarted to play with it since you mentioned it and I see what you mean - the TNF is not the 'brick-wall' like the others.  Best I could get was to place the filter in the deepest notch and narrowest bandwidth to not reduce the tones of the wanted signal.

I also get the nonrecoverable error after I have created and deleted a few of the TNFs - enough to where I had to crawl under the desk each time to re-boot the radio in order to get SDR to find it.

I will hope for some twin-peaking RTTY filters for Christmas . . . not sure if I have been a good enough lad to ask for FSK. .

 

Walt

Messed up on error picture - lets try this again for the Flex guys.   Maybe it will help

k3Tim

How wide are you making the TNF filter?

I have tried this and at about 60Hz wide it's fine.  Above that it may start to cut into the RTTY tones.
The band width for a 170 Hz shift 45.45 Baud signal is :

BW = 1.2 * 170 + 45.45   = 249 and change.  Let's call it 250.   
So...   250 - 170 = 80 Hz.  Each mark / space  tone has a **** of 80 cycles wide.  That would be 40 hz on each side of the M/S tone.  So the no man's land between M/S tones is 170 - 80 = 90 hz.  In theory you should be able to make the filter 90Hz wide.  I tried that in CW mode, with the CW fitler set at 240 Hz.  The audio was reduced somewhat but the signal going into FLDiGi and Spectrum Lab was same magnitude.  I changed the TNF BW to 60 Hz and took some screen shots.  Here is a montage of the panafall, Spectrum lab that shows the area between the m/s tones is filtered.  The signal is copied well under harsh conditions.  Of course, setting the TNF manually is takes the touch of a safe ****.  

I tried creating / deleting a bunch of TNFs  without any crashes.  This worked flawlessly.

Does anyone have a link / info as to the BER decrease by using this extra filtering?

_..--
k3Tim / MSee / PE / PPSEL-IFR 

Walt

Thanks for the feedback, K3TIM - I will try some other things and see what I am doing to make it fail - perhaps an interaction with some other program that is not behaving properly.

Steve-N5AC

Couple of things.  First, the TNF is a different kind of filter from the bandpass filters we use and they are generally sharp and deep.  The more you widen the filter, though, the less depth it has.  This is why there is an option to add additional depth.  Also, we do not have the precise control over the start/stop frequency so what you see in the panadapter is an approximate rendering of the effects of the TNF that you can use for guidance.  The TNF is governed by the center frequency, the Q (adjusted with the width) and the depth (one of three levels selected with the pop-up).

k3Tim

Hi Steve,

Thanks for the details on TNF...

Do you have an opinion on the benefit of the twin-peaking filtering for RTTY? 

Tim

DrTeeth

Thanks for that Steve. It is actually quite handy to plonk a default-sized TNF in the notch of a RTTY signal and **** the sound completely.

73 Guy