• Work the World on 10W - An Introduction to JT65A

    Tonight I was sitting at the station fiddling around with things (like I usually do), and thought I'd fire up WSJT and give JT65A a spin on 30m. For those of you unfamiliar with JT65A it is the ultimate weak signal mode. It was originally designed for EME and troposcatter operations, but also works very well on HF for long haul operations in less than ideal band conditions.

    JT65A QSO's are unique in that each QSO takes exactly six minutes to complete. Being an EME mode, JT65A follows the EME convention of each station transmitting for just under a minute, then listening for a minute. The computer being used for this digital mode must have a system clock that is synced to an external reference time standard. My computer utilizes NTP (Network Time Protocol) to connect to one of a multitude of time servers available on the Internet which are all referenced to an atomic clock. Synchronizing the computer's system time is of critical importance because all stations working JT65A must transmit/listen in synchronized windows for a successful QSO.

    The exchange in a JT65A QSO is in a very standardized format. This is due in part to the fact that each 1 minute transmission only transmits a maximum of 13 characters. This mode is not for ragchewing! It's purpose is to exchange the basic information required to complete a legitimate QSO reliably over extremely weak signal paths. The JT65A software can often decode signals below the noise floor.

    A basic QSO begins with a standard CQ transmission.

    1) Station A sends: CQ KE7KUS DM33

    2) Station B sends: KE7KUS W1AW FN31

    3) Station A sends: W1AW KE7KUS -15

    4) Station B sends: KE7KUS W1AW R -12

    5) Station A sends: W1AW KE7KUS RRR

    6) Station B sends: KE7KUS W1AW 73

    7) Station A sends: W1AW KE7KUS 73

    At this point, the JT65A QSO is complete and KE7KUS begins transmitting CQ again. Let's dig into what's going on during this QSO a little deeper.

    1) The CQ call is fairly explanatory. The sending station grid follows the callsign - remember...this mode was originally designed for EME.
    2) The responding station also sends their grid following the station callsign. The astute observer will notice these exchanges are more than 13 characters (spaces count as a character). If calls fit the standard format above, JT65A software has a special algorithm which encodes/decodes more than 13 characters. If the transmitted message is non-standard in format, the 13 character limit is strictly applied.
    3) This transmission is the first of the two exchanged signal reports. Unlike a traditional RST signal report, the JT65A signal report is given in dB. In this case, KE7KUS is receiving W1AW at -15dB.
    4) W1AW responds with his signal report. R -12 mean "Roger, your signal is -12dB."
    5) If the signal report is received successfully by KE7KUS, he responds RRR, which means "Roger, Roger, Roger." Successful receipt of transmissions is where many JT65A newcomers get confused. After the other station responds to your CQ call, the exchange should proceed exactly as outlined above. You may find times where a static crash or QSB takes out part of the signal you are receiving which may cause it not to be decoded properly. Any time an expected transmission does not decode properly, you should resend your previous transmission again. This lets the other station know that you did not receive his last transmission and indicates to him that he should resend. Example:

    In the above exchange, say I did not receive the transmission KE7KUS W1AW R -12, for whatever reason. Since that was what I was expecting to get, if I don't receive it I would retransmit W1AW KE7KUS -15. W1AW would get this transmission and say to himself "Hmm...he already sent me that...it must mean he didn't get my last transmission...I will send it again." W1AW would then send me KE7KUS W1AW R -12 again and the QSO would proceed to the next step (assuming it decoded properly the second time around.)

    6/7) The 73 exchange lets both stations know the QSO was successful. Instead of sending the standard 73 message, some stations use this transmission to send their transmit power and antenna type. In our example, such an exchange might look something like this:

    Station B sends: 20W DP 73
    Station A sends: 10W VERT 73

    This would indicate that Station B is running 20W into a dipole and Station A is using 10W into a vertical antenna. If you choose to do this, you should be aware of two things: a) since the transmission will be nonstandard, the 13 character limit is strictly enforced, and b) that usually prevents you from sending your callsign to close the QSO (a Part 97 requirement). Some JT65A software programs get around this by quickly sending the station's callsign in CW after the JT65A part is sent. If you opt to send your power/antenna, please ensure this is set up correctly.

    Much like PSK-31, JT65A QSO's by convention are usually limited to 50W. Even with that accepted convention, many stations seldom operate above 25W due to the efficiency of the mode. Unlike PSK-31, linearity of the transmitted signal is not as critical as with PSK-31; however, I seem to have the best results when running low/no ALC with JT65A.

    Detailed instructions on how to operate the WSJT software, and more details on how to conduct a JT65A QSO are available in the WSJT operating manual: http://physics.princeton.edu/pulsar/...T_User_600.pdf

    During my session on 30m tonight I made two contacts into the South Pacific from my home here in Phoenix. The first was ZL1ATB in Auckland. New Zealand (6763.4 miles), the second was VK3AMA near Melbourne, Australia (8200.2 miles). Both contacts were made running 10W into my S9 Antennas 43' vertical. The attached pictures show screenshots of both the WSJT software and a snapshot of PSKReporter, which shows the stations which are receiving my signal (each green bubble is a station that has received my 10W signal with the time the last signal was received inside the bubble.) As you can see, the ability of this mode to get out is quite remarkable!





    You will find JT65A around the .076's for most bands (7.076, 14.076, and 21.076 MHz). On 17m, JT65'ers seems to hang out just above 18.100 and on 30m they seem to hang out just below 10.140. In general, if you've worked PSK in the past, JT65'ers won't be too far away from those frequencies. Just listen for the very distinctive, slow, musical tones of JT65A and you've hit the jackpot. Although JT65A isn't for everyone, if QRP DX or racking up DX entities with modest station/antenna equipment interests you, this mode might just be what you're looking for. Last year after several weeks of JT65A ops I decided to take a break and go back PSK-31 for a couple of days, only to find myself quite disappointed that I had to ratchet up the power to 50W and still had less success with DX QSO's. This mode will spoil you when it comes to getting out with low power or poor band conditions. Nothing else on the air quite compares. Have fun with JT65A and hope to see you on the air!

    73,
    Kurt
    KE7KUS
    This article was originally published in forum thread: Work the World on 10W - An Introduction to JT65A started by KE7KUS View original post
    Comments 8 Comments
    1. K6OZY's Avatar
      K6OZY -
      JT65A is a ton of fun and I love it when I work people and I can't even hear them on the radio. -23dB 4tw!
    1. K6OZY's Avatar
      K6OZY -
      Quote Originally Posted by AC0VH View Post
      I did not see it mentioned other than in your statement about path efficiency (which is a very true statement). Be aware of your transmitter's duty cycle rating. JT65 is 100% duty cycle for about 2 minutes at a stretch, so you should generally never run beyond about half of your rig's max rating and even then be careful that air flow is good, etc. It's safest to run 25% or even 10% of rated, more so with backpack and mobile radios. Very tight, relatively poor air flow, small or no fans... Even a real base radio or amp will still not really like a high power, sustained TX. For example, I would not run a FT-817 at 5W in JT65 even with a fan blowing across it. Just not designed for that.
      I keep my TX power at 100%, but turn down my audio so that the actual power output is 50% or less for PSK31, even less for JT65a. This seems to guarantee a very good quality audio sound.

      Go to this website, type in the power you used and the signal report from the other side. It'll tell you how much less you could have used to make the same contact. Pretty cool stuff.

      http://www.k7ek.net/viewpage.php?page_id=15
    1. AC0VH's Avatar
      AC0VH -
      Quote Originally Posted by K6OZY View Post
      I keep my TX power at 100%, but turn down my audio so that the actual power output is 50% or less for PSK31, even less for JT65a. This seems to guarantee a very good quality audio sound.
      JT65 is an AFSK modulation with tones like ~1KHz on a continuous wave, so during TX it's essentially a sine wave (e.g. dead key CW) for ~47 seconds (372ms x 126 intervals). Personally, I think the guideline should be run the transmitter at the AM rating and set your audio just below clipping. This guarantees the best SNR and a reasonable expectation of the power amp. It's designed for path loss in the hundreds of dB, so using 5W is anything but low power and 25W would be the equivalent of a legal limit or more on CW or SSB.
    1. AC0VH's Avatar
      AC0VH -
      If anyone is wondering, JT65A/B/C was originally conceived for earth-moon-earth (e.g. moon bounce) contacts. The path loss on 2m for an EME contact is around 255dB. That's 250,000 miles, so you need lots of power and gain to have any sort of signal to receive, which is the reason the protocol is designed to work with such low SNR. If you dump 1kW into 25dB of antenna gain, you are transmitting roughly a 55dB signal, so the receiving antenna might have 25dB of gain, which means the radio needs to hear way down into roughly -130dBm signal strength. That will be in the noise, just as long as it's above the radio's noise figure it'll do it's thing.

      To compare to HF propagation, 5W is equivalent to 7dB which on a 5,000km path on 20m using unity gain (1/4 wavelength) antennas will produce a received signal of about -122dBm with a path loss of maybe 125dB. This is truly very low but as long as the signal is higher than the radio's noise it will decode it. The real key to JT65 is that your receiver is quiet and has a very low noise figure. Use the band where your receiver has the best sensitivity and dynamic range, which is often 40/20/10 for most radios.
    1. KE7KUS's Avatar
      KE7KUS -
      I live inside the metro area here in Phoenix where the ambient noise levels are ridiculously high. In addition, I have to run a vertical antenna, as I don't have the supports or height available to run any type of horizontal antenna. As such, most nights on 40m I see S8-S9 noise levels, which equates to 48-54dB. Although it wasn't designed to fight through urban noise, it's one of the reasons I love the mode...it lets me play DX when I would otherwise be out of the picture.
    1. AC0VH's Avatar
      AC0VH -
      Quote Originally Posted by KE7KUS View Post
      I live inside the metro area here in Phoenix where the ambient noise levels are ridiculously high. In addition, I have to run a vertical antenna, as I don't have the supports or height available to run any type of horizontal antenna. As such, most nights on 40m I see S8-S9 noise levels, which equates to 48-54dB. Although it wasn't designed to fight through urban noise, it's one of the reasons I love the mode...it lets me play DX when I would otherwise be out of the picture.
      As long as your noise is spectrally flat, JT65 is pretty good at handling it. Even -30dB SNR is not difficult to achieve WSJT copy with a decent radio (around -24dB is almost a guarantee copy) which equates to picking out a S4 or S5 signal out of S9 noise. Assuming you don't have someone next door who runs his transmitter horribly non-linear (Goodness help you if you have an illegal CB'er in the neighborhood) and are careful about turning off local periodic sources, computers, switching supplies, TVs, washing machines you can copy pretty weak signals even in the city.
    1. K6OZY's Avatar
      K6OZY -
      Quote Originally Posted by AC0VH View Post
      JT65 is an AFSK modulation with tones like ~1KHz on a continuous wave, so during TX it's essentially a sine wave (e.g. dead key CW) for ~47 seconds (372ms x 126 intervals). Personally, I think the guideline should be run the transmitter at the AM rating and set your audio just below clipping. This guarantees the best SNR and a reasonable expectation of the power amp. It's designed for path loss in the hundreds of dB, so using 5W is anything but low power and 25W would be the equivalent of a legal limit or more on CW or SSB.
      Great feedback. I'll try dropping the TX to 25% and tune the audio better on my Flex-3000. With my Flex-1500, I run 100% (5W) without worry because they told me that the PA on that unit is horribly under utilized and I could run it at 100% duty cycle for hours on end without any issue. I did drop it to 1W to see how far it would go and was quite amazed how many stations picked me up on Psk31 reporter.

      Some PSK31 work at 2.5W, and I can see some JT65a going on up at .076 too!
    1. AC0VH's Avatar
      AC0VH -
      Just worked VK2DAG on 15m with 10W from home with my terribly inefficient 31' vertical. :-)