开云体育

  1. Direwolf
  2. Topics

alsaloop for multimode

 

did anybody try to divide the audio capture of the radio audio card into alsaloop and get some more outputs, which are fed into several direwolf instances with different configuration of speed (AFSK 1k2, QPSK 2k4,? 8-PSK 4k8)?

Michael
 

I didn't use alsaloop. I just made a Y mdin6 cable. This specifically was to run 1200 & 9600 baud packet on a single channel. Note that the PTT for both channels is tied together which for my case works fine.

I have this configuration working for Winlink email, Linux RMS Gateway & APRS Gateway using a Kenwood TM-V71a radio.

/Basil n7nix
 

开云体育


Hello Basil,

Few questions if I may:

?? - With your setup, is this doing *simultaneous* decodes of 1200bps AFSK and 9600bps FSK signals on the same frequency?? It looks like it might be able to RX this way but not transmit.? Not clear to me.

?? - Does this solution mean you have the radio (looks like your using a Kenwood v71) always in "9600baud" mode where you send all TX audio directly into the discriminator for transmission?? If so, this means that all 1200bps modem tones won't have the expected pre-emphassis applied to the transmission and many remote legacy TNCs might not decode the packets due to the "twist" issue between the two tones having different signal levels.

--David
KI6ZHD



On 10/08/2021 08:22 AM, Basil Gunn wrote:

I didn't use alsaloop. I just made a Y mdin6 cable. This specifically was to run 1200 & 9600 baud packet on a single channel. Note that the PTT for both channels is tied together which for my case works fine.

I have this configuration working for Winlink email, Linux RMS Gateway & APRS Gateway using a Kenwood TM-V71a radio.

/Basil n7nix
 

  • With your setup, is this doing simultaneous decodes of 1200bps AFSK and 9600bps FSK signals on the same frequency? It looks like it might be able to RX this way but not transmit. Not clear to me.

I've tested this a lot and contrary to other posters doing 1200 baud packet with your radio set to 9600 baud works quite well.

  • Does this solution mean you have the radio (looks like your using a Kenwood v71) always in "9600baud" mode

Yes

where you send all TX audio directly into the discriminator for transmission?

Does not appear to matter. I run all my packet gateways this way.

If so, this means that all 1200bps modem tones won't have the expected pre-emphassis applied to the transmission and many remote legacy TNCs might not decode the packets due to the "twist" issue between the two tones having different signal levels.

I have NOT seen this as an issue. It certainly is NOT an issue for direwolf. I test with a KPC3+ a couple of times a year. That's a bench test and the "twist" issue could possibly be a problem with more marginal signals (or not as I have yet to experience that)

My primary goal was to transition my gateways to 9600 baud packet while still offering 1200 baud service. I've gone through a few iterations of solving this problem and have been testing and using the TM-V71a discriminator with 1200 baud packet for over a year now. Just thought I would mention my experience with PulseAudio and what did in fact work for me. (See previous github link).
 

FYI, the 6-pin miniDIN 'data jack' does not have a separate TX audio pin
1??? Data In (Mic)????????4???? 9600bps RX Audio out (from radio)
2??? Ground???????????????? 5 ????1200bps RX Audio out (from Radio)
3??? PTT????????????????????? 6 ????Sql

So, either both modes send data directly into the discriminator, or neither does.
????? The TX TWIST concern should be the same for either.

Robert Giuliano
KB8RCO



On Friday, October 8, 2021, 04:57:53 PM EDT, Basil Gunn <basil@...> wrote:


  • With your setup, is this doing simultaneous decodes of 1200bps AFSK and 9600bps FSK signals on the same frequency? It looks like it might be able to RX this way but not transmit. Not clear to me.

I've tested this a lot and contrary to other posters doing 1200 baud packet with your radio set to 9600 baud works quite well.

  • Does this solution mean you have the radio (looks like your using a Kenwood v71) always in "9600baud" mode

Yes

where you send all TX audio directly into the discriminator for transmission?

Does not appear to matter. I run all my packet gateways this way.

If so, this means that all 1200bps modem tones won't have the expected pre-emphassis applied to the transmission and many remote legacy TNCs might not decode the packets due to the "twist" issue between the two tones having different signal levels.

I have NOT seen this as an issue. It certainly is NOT an issue for direwolf. I test with a KPC3+ a couple of times a year. That's a bench test and the "twist" issue could possibly be a problem with more marginal signals (or not as I have yet to experience that)

My primary goal was to transition my gateways to 9600 baud packet while still offering 1200 baud service. I've gone through a few iterations of solving this problem and have been testing and using the TM-V71a discriminator with 1200 baud packet for over a year now. Just thought I would mention my experience with PulseAudio and what did in fact work for me. (See previous github link).
 

开云体育

Yes, unless the radio "knows" you're in 1k2 vs 9k6 mode, and adjusts the audio path internally as a result.

My own Yaesu FT-847 has such a mode.? The user guide doesn't talk much about what it does other than to say that if one is trying to run 2400 baud "you may have to experiment with this Menu selection".? That implies it does affect the transmit audio path, if not the receive as well.

Greg? KO6TH


Rob Giuliano via groups.io wrote:

FYI, the 6-pin miniDIN 'data jack' does not have a separate TX audio pin
1??? Data In (Mic)????????4???? 9600bps RX Audio out (from radio)
2??? Ground???????????????? 5 ????1200bps RX Audio out (from Radio)
3??? PTT????????????????????? 6 ????Sql

So, either both modes send data directly into the discriminator, or neither does.
????? The TX TWIST concern should be the same for either.

Robert Giuliano
KB8RCO



On Friday, October 8, 2021, 04:57:53 PM EDT, Basil Gunn <basil@...> wrote:


  • With your setup, is this doing simultaneous decodes of 1200bps AFSK and 9600bps FSK signals on the same frequency? It looks like it might be able to RX this way but not transmit. Not clear to me.

I've tested this a lot and contrary to other posters doing 1200 baud packet with your radio set to 9600 baud works quite well.

  • Does this solution mean you have the radio (looks like your using a Kenwood v71) always in "9600baud" mode

Yes

where you send all TX audio directly into the discriminator for transmission?

Does not appear to matter. I run all my packet gateways this way.

If so, this means that all 1200bps modem tones won't have the expected pre-emphassis applied to the transmission and many remote legacy TNCs might not decode the packets due to the "twist" issue between the two tones having different signal levels.

I have NOT seen this as an issue. It certainly is NOT an issue for direwolf. I test with a KPC3+ a couple of times a year. That's a bench test and the "twist" issue could possibly be a problem with more marginal signals (or not as I have yet to experience that)

My primary goal was to transition my gateways to 9600 baud packet while still offering 1200 baud service. I've gone through a few iterations of solving this problem and have been testing and using the TM-V71a discriminator with 1200 baud packet for over a year now. Just thought I would mention my experience with PulseAudio and what did in fact work for me. (See previous github link).
 

I remember asking the Yaesu rep when I bought my FT7100 (years ago now) .
He told me that radio only changed the receive pin.? That was all that was needed.

Rob KB8RCO


On Fri, Oct 8, 2021 at 18:04, Greg D
<ko6th.greg@...> wrote:
Yes, unless the radio "knows" you're in 1k2 vs 9k6 mode, and adjusts the audio path internally as a result.

My own Yaesu FT-847 has such a mode.? The user guide doesn't talk much about what it does other than to say that if one is trying to run 2400 baud "you may have to experiment with this Menu selection".? That implies it does affect the transmit audio path, if not the receive as well.

Greg? KO6TH


Rob Giuliano via groups.io wrote:
FYI, the 6-pin miniDIN 'data jack' does not have a separate TX audio pin
1??? Data In (Mic)????????4???? 9600bps RX Audio out (from radio)
2??? Ground???????????????? 5 ????1200bps RX Audio out (from Radio)
3??? PTT????????????????????? 6 ????Sql

So, either both modes send data directly into the discriminator, or neither does.
????? The TX TWIST concern should be the same for either.

Robert Giuliano
KB8RCO



On Friday, October 8, 2021, 04:57:53 PM EDT, Basil Gunn <basil@...> wrote:


  • With your setup, is this doing simultaneous decodes of 1200bps AFSK and 9600bps FSK signals on the same frequency? It looks like it might be able to RX this way but not transmit. Not clear to me.

I've tested this a lot and contrary to other posters doing 1200 baud packet with your radio set to 9600 baud works quite well.

  • Does this solution mean you have the radio (looks like your using a Kenwood v71) always in "9600baud" mode

Yes

where you send all TX audio directly into the discriminator for transmission?

Does not appear to matter. I run all my packet gateways this way.

If so, this means that all 1200bps modem tones won't have the expected pre-emphassis applied to the transmission and many remote legacy TNCs might not decode the packets due to the "twist" issue between the two tones having different signal levels.

I have NOT seen this as an issue. It certainly is NOT an issue for direwolf. I test with a KPC3+ a couple of times a year. That's a bench test and the "twist" issue could possibly be a problem with more marginal signals (or not as I have yet to experience that)

My primary goal was to transition my gateways to 9600 baud packet while still offering 1200 baud service. I've gone through a few iterations of solving this problem and have been testing and using the TM-V71a discriminator with 1200 baud packet for over a year now. Just thought I would mention my experience with PulseAudio and what did in fact work for me. (See previous github link).
 

开云体育


There is a setting in the Kenwood V71 to enable TX pre-emphasis and narrower (aka "1200") or disable TX pre-emphasis (aka "9600"):

?? --> PDF page 83 --> Menu 518 (DAT.SPD)


Per Basil's email, he acknowledged he's always sending without pre-emphasis.? I agree that Direwolf and some modern TNCs will cope with this but many legacy TNCs (MFJ TNC2-design and many basic PIC-based TNCs like TNC-X can have issues with the resulting signal after the local "classic radio" applies the standard de-emphasis circuit.? This is well described here:

??

On this very topic, I'm personally torn to keep "legacy compatibility" or say "screw the old hardware, let's move on with newer stuff".? It would be slick if Direwolf could apply an emulated pre-emphasis to only the 1200bps AFSK signal.? If that was there, I think we could have the best of both worlds here.

--David
KI6ZHD


On 10/08/2021 02:51 PM, Rob Giuliano via groups.io wrote:

FYI, the 6-pin miniDIN 'data jack' does not have a separate TX audio pin
1??? Data In (Mic)????????4???? 9600bps RX Audio out (from radio)
2??? Ground???????????????? 5 ????1200bps RX Audio out (from Radio)
3??? PTT????????????????????? 6 ????Sql

So, either both modes send data directly into the discriminator, or neither does.
????? The TX TWIST concern should be the same for either.

Robert Giuliano
KB8RCO



On Friday, October 8, 2021, 04:57:53 PM EDT, Basil Gunn <basil@...> wrote:


  • With your setup, is this doing simultaneous decodes of 1200bps AFSK and 9600bps FSK signals on the same frequency? It looks like it might be able to RX this way but not transmit. Not clear to me.

I've tested this a lot and contrary to other posters doing 1200 baud packet with your radio set to 9600 baud works quite well.

  • Does this solution mean you have the radio (looks like your using a Kenwood v71) always in "9600baud" mode

Yes

where you send all TX audio directly into the discriminator for transmission?

Does not appear to matter. I run all my packet gateways this way.

If so, this means that all 1200bps modem tones won't have the expected pre-emphassis applied to the transmission and many remote legacy TNCs might not decode the packets due to the "twist" issue between the two tones having different signal levels.

I have NOT seen this as an issue. It certainly is NOT an issue for direwolf. I test with a KPC3+ a couple of times a year. That's a bench test and the "twist" issue could possibly be a problem with more marginal signals (or not as I have yet to experience that)

My primary goal was to transition my gateways to 9600 baud packet while still offering 1200 baud service. I've gone through a few iterations of solving this problem and have been testing and using the TM-V71a discriminator with 1200 baud packet for over a year now. Just thought I would mention my experience with PulseAudio and what did in fact work for me. (See previous github link).
 

开云体育

For anyone who wants reasonably comprehensive information on how 9600 v 1200 is handled in the V-71, take a look at page 9 of the Kenwood service manual for a complete description, including levels.?



The information builds substantially on the basic information contained in the instruction manual.


On 9/10/21 11:11 am, David Ranch wrote:

There is a setting in the Kenwood V71 to enable TX pre-emphasis and narrower (aka "1200") or disable TX pre-emphasis (aka "9600"):

?? --> PDF page 83 --> Menu 518 (DAT.SPD)


Per Basil's email, he acknowledged he's always sending without pre-emphasis.? I agree that Direwolf and some modern TNCs will cope with this but many legacy TNCs (MFJ TNC2-design and many basic PIC-based TNCs like TNC-X can have issues with the resulting signal after the local "classic radio" applies the standard de-emphasis circuit.? This is well described here:

??
+1

On this very topic, I'm personally torn to keep "legacy compatibility" or say "screw the old hardware, let's move on with newer stuff".? It would be slick if Direwolf could apply an emulated pre-emphasis to only the 1200bps AFSK signal.? If that was there, I think we could have the best of both worlds here.
It's probably easy for members of a Direwolf list to think old hardware has disappeared but I strongly suspect it exists in large numbers "in the wild". I think that all users should stick with long-standing convention with regard to pre-emphasis for 1k2 packet to minimise network connection issues with hardware based sites.

--David
KI6ZHD
Ray vk2tv


On 10/08/2021 02:51 PM, Rob Giuliano via groups.io wrote:
FYI, the 6-pin miniDIN 'data jack' does not have a separate TX audio pin
1??? Data In (Mic)????????4???? 9600bps RX Audio out (from radio)
2??? Ground???????????????? 5 ????1200bps RX Audio out (from Radio)
3??? PTT????????????????????? 6 ????Sql

So, either both modes send data directly into the discriminator, or neither does.
????? The TX TWIST concern should be the same for either.

Robert Giuliano
KB8RCO



On Friday, October 8, 2021, 04:57:53 PM EDT, Basil Gunn <basil@...> wrote:


  • With your setup, is this doing simultaneous decodes of 1200bps AFSK and 9600bps FSK signals on the same frequency? It looks like it might be able to RX this way but not transmit. Not clear to me.

I've tested this a lot and contrary to other posters doing 1200 baud packet with your radio set to 9600 baud works quite well.

  • Does this solution mean you have the radio (looks like your using a Kenwood v71) always in "9600baud" mode

Yes

where you send all TX audio directly into the discriminator for transmission?

Does not appear to matter. I run all my packet gateways this way.

If so, this means that all 1200bps modem tones won't have the expected pre-emphassis applied to the transmission and many remote legacy TNCs might not decode the packets due to the "twist" issue between the two tones having different signal levels.

I have NOT seen this as an issue. It certainly is NOT an issue for direwolf. I test with a KPC3+ a couple of times a year. That's a bench test and the "twist" issue could possibly be a problem with more marginal signals (or not as I have yet to experience that)

My primary goal was to transition my gateways to 9600 baud packet while still offering 1200 baud service. I've gone through a few iterations of solving this problem and have been testing and using the TM-V71a discriminator with 1200 baud packet for over a year now. Just thought I would mention my experience with PulseAudio and what did in fact work for me. (See previous github link).

1 - 9 of 9