Clean 600hz signal, full dsp decoupling possible?

[PA3CW]

Hello forum friends,
Not being an expert in the world of DSP, but being an experienced CW operator i often asked myself this question:  Why is there no full decoupling between the received signal and the tone in the headset or speaker? In other words is it possible that the dsp discrimiates enough between signal and noise and steers a clean LF oscillator making a clean 600hz or so tone?  Is such a setting possible or do i overlook something here? Just a little out of the box thinking...
Dick PA3CW

[Geoffrey Mackenzie-Kennedy]

Hello Dick,

I am not a DSP expert either, but a tracking filter keying an audio
oscillator will do exactly what you want. In its simplest form a tracking
filter is a phase locked loop, and ICs which can be used at audio for this
purpose have been available since the 1960s and are inexpensive. Also a
tracking filter working at audio can be connected to a receiver's low level
audio output if one is not already in the receiver. I believe that the DSP
wizards have already designed something that does the same job.

Tracking filters are very useful when searching for signals below the
receiver's noise floor, at VHF for example, but listening to a keyed audio
oscillator without noise can be quite boring.

73,
Geoff
GM4ESD



----- Original Message -----
From: "PA3CW" <[hidden email]>
To: <[hidden email]>
Sent: Saturday, October 17, 2009 8:57 PM
Subject: [Elecraft] Clean 600hz signal, full dsp decoupling possible?

>
> Hello forum friends,
> Not being an expert in the world of DSP, but being an experienced CW
> operator i often asked myself this question:  Why is there no full
> decoupling between the received signal and the tone in the headset or
> speaker? In other words is it possible that the dsp discrimiates enough
> between signal and noise and steers a clean LF oscillator making a clean
> 600hz or so tone?  Is such a setting possible or do i overlook something
> here? Just a little out of the box thinking...
> Dick PA3CW

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[PA3CW]

Hi Geoff,

Thanks for your reaction. I am very curious whether the modern DSP technology can discriminate signal from noice (in CW) better than the human ear and I dont know if it would be boring to just hear a clean tone whilst the signal is at a minimum level. I think it will be an amazing thing to experience as you do expect having to put in a lot of effort to hear the signal through a lot of noise. The second step is if the DSP of the K3 is able to include such a setting or external 'tracking filter' is necessary.

Best,
Dick ... . . .._

Geoffrey Mackenzie-Kennedy wrote

Hello Dick,

I am not a DSP expert either, but a tracking filter keying an audio
oscillator will do exactly what you want. In its simplest form a tracking
filter is a phase locked loop, and ICs which can be used at audio for this
purpose have been available since the 1960s and are inexpensive. Also a
tracking filter working at audio can be connected to a receiver's low level
audio output if one is not already in the receiver. I believe that the DSP
wizards have already designed something that does the same job.

Tracking filters are very useful when searching for signals below the
receiver's noise floor, at VHF for example, but listening to a keyed audio
oscillator without noise can be quite boring.

73,
Geoff
GM4ESD



----- Original Message -----
From: "PA3CW" <[e-mail]>
To: <[e-mail]>
Sent: Saturday, October 17, 2009 8:57 PM
Subject: [Elecraft] Clean 600hz signal, full dsp decoupling possible?


>
> Hello forum friends,
> Not being an expert in the world of DSP, but being an experienced CW
> operator i often asked myself this question:  Why is there no full
> decoupling between the received signal and the tone in the headset or
> speaker? In other words is it possible that the dsp discrimiates enough
> between signal and noise and steers a clean LF oscillator making a clean
> 600hz or so tone?  Is such a setting possible or do i overlook something
> here? Just a little out of the box thinking...
> Dick PA3CW


______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[e-mail]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[pd0psb]

Hello Dick!

An SDR program called Winrad is capable of exactly what you thought.
It is a function called "despread" aimed for weak-signal copying (EME and such)

http://www.sdrham.com/winrad/index.html
It's freeware working with several SDR receivers but you can also feed K3 audio into it.

There's a good audio example of "re-synthesized CW" on this page.

73'
Paul
PD0PSB

PA3CW wrote

Hi Geoff,

Thanks for your reaction. I am very curious whether the modern DSP technology can discriminate signal from noice (in CW) better than the human ear and I dont know if it would be boring to just hear a clean tone whilst the signal is at a minimum level. I think it will be an amazing thing to experience as you do expect having to put in a lot of effort to hear the signal through a lot of noise. The second step is if the DSP of the K3 is able to include such a setting or external 'tracking filter' is necessary.

Best,
Dick ... . . .._

Geoffrey Mackenzie-Kennedy wrote

Hello Dick,

I am not a DSP expert either, but a tracking filter keying an audio
oscillator will do exactly what you want. In its simplest form a tracking
filter is a phase locked loop, and ICs which can be used at audio for this
purpose have been available since the 1960s and are inexpensive. Also a
tracking filter working at audio can be connected to a receiver's low level
audio output if one is not already in the receiver. I believe that the DSP
wizards have already designed something that does the same job.

Tracking filters are very useful when searching for signals below the
receiver's noise floor, at VHF for example, but listening to a keyed audio
oscillator without noise can be quite boring.

73,
Geoff
GM4ESD



----- Original Message -----
From: "PA3CW" <[e-mail]>
To: <[e-mail]>
Sent: Saturday, October 17, 2009 8:57 PM
Subject: [Elecraft] Clean 600hz signal, full dsp decoupling possible?


>
> Hello forum friends,
> Not being an expert in the world of DSP, but being an experienced CW
> operator i often asked myself this question:  Why is there no full
> decoupling between the received signal and the tone in the headset or
> speaker? In other words is it possible that the dsp discrimiates enough
> between signal and noise and steers a clean LF oscillator making a clean
> 600hz or so tone?  Is such a setting possible or do i overlook something
> here? Just a little out of the box thinking...
> Dick PA3CW


______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[e-mail]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[David Woolley (E.L)]

PA3CW wrote:

> speaker? In other words is it possible that the dsp discrimiates enough
> between signal and noise and steers a clean LF oscillator making a clean
> 600hz or so tone?  Is such a setting possible or do i overlook something
> here? Just a little out of the box thinking...

This is easy to do for a fade-free, low noise signal, but you wouldn't
need it for that.  For a real world weak signal, or even quite a strong
one, subject to fading, you would either have to accept highly variable
mark lengths, including missing and false positives, or you would need
to delay the output by several characters, to allow the decoder to get
an optimum solution based on sufficient lookahead.

It's almost certain that a trained ear will be able to compensate for
noise and fade much better than any current digital slicing logic.  Even
then, I suspect the trained ear does rely on lookahead.

This is why morse decoding software is still easily beaten by a human ear.

It would be easier to do for synchronous CW, with all units the same
length, and machine sent code may approximate this, as the transition
points would be predictable, Even then, without lookahead, weak signals
would risk having dah changed into di di and v.v. and etc.
--
David Woolley
"we do not overly restrict the subject matter on the list, and we
encourage postings on a wide range of amateur radio related topics"
List Guidelines <http://www.elecraft.com/elecraft_list_guidelines.htm>
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

[Julian, G4ILO]

Geoffrey Mackenzie-Kennedy wrote

Tracking filters are very useful when searching for signals below the
receiver's noise floor, at VHF for example, but listening to a keyed audio
oscillator without noise can be quite boring.

Listening to the awful noise I have to put up with here is very tiring. In fact, I pretty well use digimodes exclusively now so I don't have to listen to the audio at all. I would love a feature like this. I assume that you could inject some "white noise" into the background to make it easier to listen to; you could even vary the S/N ratio so that a loud signal sounded louder.

Whether the DSP has enough "cycles" left for this to ever be more than a dream is another matter...

[pd0psb]

Hello Julian,

In november 2008 CPU was hitting just over 50%.
http://w7brs.com/k3/wkshop/
I don't think the additions over the last year used the other 50%.

If so, there may be room for economizing the code, it's quite common to do that after new developments,additions & tests. (Which have been put out in large numbers lately!)
 
Time for users to economize on requests for the time being ;-)

73'
Paul
PD0PSB



Whether the DSP has enough "cycles" left for this to ever be more than a dream is another matter...

[Guy, K2AV]

"I don't think the additions over the last year used the other 50%."

Actually with various continuing features on (NR, NB, CWT, etc), the current
code has been punching the top.  While all you say (cleaning up, improving
process, etc) is true, the idea of a vast unlimited code execution space is
not.

As inevitably happens with hardware code in a fixed hardware environment
(nature abhors a vacuum), ideas (even genius level ideas) will exhaust
resources and trade-offs begin.  Fast, compact, and resource conserving code
always comes AFTER proven concepts, and is the hardest code to write.

It is quite interesting to watch the Aptos crowd work on this stuff. I hope
people understand the rarity of this publicly aired on-going accomplishment.

73, Guy.
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html