JT44 Test Procedure

Home

WSJT

50MHz

70MHz

144MHz

1296MHz

 

JT44 Test Message Board

Write message

Read messages


Guestbook by TheGuestBook.com

Test procedure for establishing minimum power contacts using JT44

Introduction

During a 144MHz contact between PE1AHX (JO21OS) and M5MUF (IO92JP), JT44 signals were so strong that we decided to reduce power levels. Ultimately this resulted in 2-way communication over 450km using 50mW. Subsequent tests have shown the path can be worked regularly at power levels of a few hundred milliwatts under flat band conditions. This test procedure is being developed to produce a protocol for power reduction tests without a need for talkback, and also gain useful propagation data.

Suggestions for v1.3

I shall list here any occasional ideas I may have, or that have been suggested to me, that could be included with the procedure. Please email any comments, suggestions, ideas, or contact me via ON4KST chat. Or please try the Message Board link to the left.

  • TXP and RXP should be expressed as ERP rather than TX power, as this will "equalize" results from tests with different stations.
  • ERP could, with large enough stations, reach the megawatt level: Code for this MW: ie 100kW =MW1, 5 Megawatts =5MW, etc
  • Special messages (see bottom of page): are these the best? eg to avoid confusion with single letter codes used in MS and EME QSOs. Any suggestions?
  • v1.3 (and any subsequent version) will become "active" when there is general agreement that it is ready for on-air testing.

ERP = tx power x antenna gain: download an ERP Calculator (12kb, zip-file), from W5UN‘s site.

Current Procedure v1.2 - 01/10/2003

Message format

Step1st eleven2nd eleven
Initial call:PE1AHXM5MUFTXP
Optional partial report:TXPDBTXPDB
Report: TXPDBRXPTXPDBRXP
Rogers:TXPDBRXPRRRTXPDBRXPRRR
  • TXP: my tx power, should be expressed in standard format, using two figures and the unit as a decimal marker; ie ##L, #L#, L##, hence: 200W=0K2 or K20, 50W=50W or K05, 25µW=25U etc.
  • DB: report (dB below noise in 2.5kHz) - send after initial call decoded
  • RXP: your tx power received by me - send with report (this confirms that "unknown information" has been passed)
  • RRR: send after report decoded IF RXP received = TXP sent!
  • Long callsigns: these were a problem in the original format - by limiting callsign exchange to the first stage, a total of 19 letters can be accommodated plus TXP on the initial call.
  • Optional partial report: on occasion it can be useful to send a "partial report" (TXPDB) during a contact, as this tells your partner if you have signals and no decode, for instance. It is not a necessary step in the QSO.

This is a complete qso in itself, but it is also where the power reduction procedure begins.

Once initial calls have been passed, the message gets longer each over until it is completed, so it is easy to see changed message even with poor decode.

For the report send DB as the best decode at RXP, so at each power level there is a report and confirmation of the power at both ends.

This example from a previous test shows that I received a -13dB signal from PE1AHX showing TXP=18W, and my report DB=08 (dB below noise), and RXP=05W confirms that my transmit power was also 5W. Hence a single line from the decode box will supply data for both directions of the QSO.

202900 10   -7  2.0  -38    *   M5MUFPE1AHX18W                     
203000  9  -18  2.1  -39    *   M5MUFPD1$HX18W0805V             1  
203100  6  -21  2.0  -24    *   L57U7PW6AHP17W0805V            7H  
203300  5  -22  2.0    1    *   M5MUFPF1A5X05521554.N4         F5 4
203400  9  -13  2.0    9    *   M5MUFPE1AHX18W0805WRRR         05 R

When rogers are received, go to next stage - either RRRR7373 if finished or:

Power reduction procedure

Both calls have been exchanged by this point so do not NEED to be repeated every over - however, an occasional single period of ID during a QSO is good manners, and may get you extra contacts if someone is listening. The rest of the procedure uses an 11 character (including blanks) message repeated twice, so it is possible for the receiving station to use the "fold" function for another 1.5dB rx sensitivity. Don't forget to fill the JT message boxes with both "halves" of the message.

Step1st eleven2nd eleven
New message:TXPTXP
Report:TXPDBRXPTXPDBRXP
Rogers:TXPDBRXPRRRTXPDBRXPRRR
Station ID:M5MUFPE1AHXM5MUFPE1AHX

Always clear average when sending "rogers", because this is the signal for your contact to reduce power. When rogers have been received, reduce power and return to new message with TXP changed accordingly. Keep repeating until there are no signals left!

The transition from one power level to the next will go something like this:


203900  9  -17  0.2   10    *   M5MUFPE1AHX18W0805WRRR         01 R
204000  9   -9  0.2   15    *   M5MUFPE1AHX18W0805WRRR         01 R
204100  8  -10  0.2   12    *   05W                                
204200  7  -19  0.1   13    *   05W    // //    //                 
204300 11  -10 -0.1   15    *   05W                                
204400 11  -11  0.0   13    *   05W0901W                           
204500  8  -10 -0.1   11    *   05W0901W                           

After sending rogers, and receiving mine, PE1AHX has reduced power to five watts. Within three periods he is able to send my report (-9dB S/N) for my (now reduced) one watt transmission.

And so the test continues, working through the cycle until we eventually exchange reports at TXP/RXP=100, ie 100mW in either direction:


204600  8  -13 -0.1   15    *   05W0901WR                          
204700  7  -18 -0.1   15    *   01W                                
204800  5  -11  0.1   17    *   01W                                
204900  5   -6  0.2   16    *   01W                                
205000  8  -13  0.2   13    *   01W10                              
205300  5  -14  0.2   16    *   01W15500                           
205400  6  -16  0.2   16    *   01W15500                           
205500  3  -19  0.2   16    *   01W15500RRR                        
...                                                                
...                                                                
...                                                                
214400  6  -15  0.5   11    *   10012100                           

(Note: this QSO used the version 1.0 TXP/RXP convention (100=100mW), whereas the current system would use 0W1 for 100mW)

What is a complete QSO?

A QSO is complete if both parties exchange callsigns, some unknown information (usually a signal report, but not necessarily), and confirmation of reception of that unknown information. The TXPDBRXP format has a confirmation built in - RXP can only be sent after it has been received, therefore it is a confirmation, equivalent to a R-report in MS procedure. When either station receives a TXPDBRXPRRR from the other (equivalent to RRRRRRRR in MS), the QSO is complete, as callsigns, unknown information, and confirmations, have all been sent and received in both directions.

Special messages:

In the event of sudden problems or difficult conditions the following messages may be useful (repeated for the full 22-characters, they utilise all aspects of JT44's weak signal recovery features).

I will Stop (end QSO)SSSS...
Increase your powerIIII...
Nil for TEN periods (0/10 AVG)XXXX...
I will Miss some overs (TX & RX)MMMM...
QRX I won't TX some oversQQQQ...

The M and Q messages are intended to warn of silent periods whilst fitting attenuators, rebooting computer, etc. Use Q if you will still be able to receive. These messages can be sent QRO without messing up the test!

If you are interested in joining our tests of JT44's capabilities, please email me (see Home page), use the Message Board or contact M5MUF or PE1AHX on the ON4KST 144MHz chat

Top