[pskmail] Re: New test server available for testing (Fred)

  • From: Fred Reiselt <freiselt@xxxxxxxxxxxxx>
  • To: pskmail@xxxxxxxxxxxxx
  • Date: Sat, 19 Dec 2009 10:03:33 -0600

Hi Russell,

You are correct. Below 28mHz, 300 baud is the limit. I'll dig through the regulations and verify, but don't think it has changed recently.


Russell Blair wrote:
Fred about the PSK500 I was only repeating what I was told, before Rein goes to 
all that trouble please check and see for your self about PSK500 mode in the 
U.S. or maybe some one in the group can shead some light on this subject ?.

 1- Whoever said nothing is impossible never tried slamming a revolving door!
2- A government big enough to give you everything you want, is strong enough to take everything you have. - Thomas Jefferson


Russell Blair (NC5O)
Hell Field #300
DRCC #55
30m Dig-group #693

----- Original Message ----
From: Rein Couperus <rein@xxxxxxxxxxxx>
To: pskmail@xxxxxxxxxxxxx
Sent: Sat, December 19, 2009 9:30:37 AM
Subject: [pskmail] Re: New test server available for testing

Hi Fred,

this version can not be locked for PSK500, the problem of course is that RSID switching is in fact done within fldigi, not in pskmail.

I can make a version for US which looks at the server call, and when the call starts with a "W, K, or N" I can make sure the server does not switch fldigi to PSK500, but to PSKR500, which is 2x slower, has a baud rate of 250 Bd and is more robust. You will understand that the rest of the world does not want to be subject to the completely ludicrous 300 Bd limit...

That does of course not prevent fldigi from switching to PSK500, but the server can reset it to PSK500R every minute when it is scanning idle. We use RSID only for connected sessions, i.e. the clients only send an RSID before a connect request frame. The server sends an RSID before a mode change and before a poll (the client may have been switched to another mode). But during idle the server has RXID rx on, so the client can connect in a robust mode.

So if you want I can install a "WKN-trap" in the server for US stations...

To be absolutely sure you would have to patch fldigi. Changing the RSID number of the PSK500 mode will be enough... BTW, the 'WKN-trap' could also be inside fldigi, that is also software :)

In the mean time, what you can do is patch the mode table in arq.pm to look like:

2213  if ($Backoff_modeindex > 0 && $Backoff_modeindex < 10) {
        if ($Backoff_modeindex == 9) {
            $MODE[1] = "PSK500R";
            $MODE[0] = "PSK500R";
        } elsif ($Backoff_modeindex == 8) {
            $MODE[1] = "PSK500R";
            $MODE[0] = "PSK250R";
        } elsif ($Backoff_modeindex == 7) {
            $MODE[1] = "PSK500R";
            $MODE[0] = "MFSK32";
        } elsif ($Backoff_modeindex == 6) {
            $MODE[1] = "PSK500R";
            $MODE[0] = "THOR22";
        } elsif ($Backoff_modeindex == 5) {
            $MODE[1] = "PSK250";
            $MODE[0] = "PSK250R";
        } elsif ($Backoff_modeindex == 4) {
            $MODE[1] = "PSK500R";
            $MODE[0] = "MFSK32";
        } elsif ($Backoff_modeindex == 3) {
            $MODE[1] = "PSK500R";
            $MODE[0] = "THOR22";
        } elsif ($Backoff_modeindex == 2) {
            $MODE[1] = "PSK250R";
            $MODE[0] = "PSK125R";
        } elsif ($Backoff_modeindex == 1) {
            $MODE[1] = "MFSK32";
            $MODE[0] = "MFSK16";
        } elsif ($Backoff_modeindex == 0) {
            $MODE[1] = "PSK500R";
            $MODE[0] = "PSK500R";
That will prevent any client from switching a session to PSK500.

Be sure to chage the modes in your freqs.txt file, so every minute the mode will be reset to the default...

Do you want the next server version to include a 'WKN-trap'? I don't like to discriminate anyone :)


Rein PA0R

Wow... This is quite an upgrade!

I would like to install the server and test this weekend but I'm concerned, again, about the BPSK500 mode in the USA.

Will the setting you gave me for 0.9.27 in pskmailrc.pl still prevent switching to BPSK500? Or, is there another method needed with the new server?


Rein Couperus wrote:
The new robust pskmail server (0.9.29) is now ready for alpha testing. Together 
with the jpskmail-0.4 client and fldigi-3.13BG it builds a system which is 
extremely robust, and can be used under a broad range of band conditions.

What is new?

The server uses RSID technology to change modes. The client and server have 
been changed to handle RSID in such a way that they work together efficiently.
Transmission modes are changed automatically depending on channel quality in 
such a way that maximum speed is combined with optimum robustness. This is 
achieved by a mode switcher which constantly monitors the link, and changes to 
a more robust mode when the channel stalls, either because of qrm,qrn or qsb. 
The server offers various mode profiles which can be set from the client. Each 
profile has 3 modes, and the server will automatically switch between them, and 
also adjust the block length by looking at the number of errors on the link.

The following 9 profiles are available:
9: PSK500 PSK500R PSKR125
8: PSK500 PSK250R PSKR125
7: PSK500 MFSK32 PSKR125
6: PSK500 THOR22 PSKR125
5: PSK250 PSK250R PSKR125
4: PSK500R MFSK32 PSKR125
3: PSK500R THOR22 PSKR125
2: PSK250R PSK125R PSKR125
1: MFSK32 MFSK16 PSKR125
0: Use server default
The client can choose the profile by setting the Blocklength variable through 
the preferences dialog.
This mode table is preliminary, lot of testing will be needed to fix the most 
effective combinations. The ones we use now are the ones I prefer :)

The server can be connected in any mode (PSKR125, THOR22, PSK500... etc.)

The new test software is available at 

Fldigi and client packages are available for windows and linux.

At this moment only PI4TUE is running the new test server, but i hope there 
will be more soon. Using the new server I can now work PI4TUE with 1 Watts (20 
km, 18105 kHz), on 30 meters I need 10 Watts.

I hope some of you can do some tersting, and let me know of any quirks....


Rein PA0R

Other related posts: