Re: #wsjt time sync #wsjt


On Sun, May 9, 2021 at 05:04 PM, <ce3bkn@...> wrote:
Hi friends
I will appreciate your help:

install the image, I have the rasp time ok with my time zone but wsjtx does not decode ......, my usb card does work. What will be the problem? Do I necessarily need an additional GPS?
Thank you
christian from chile

hi christian,

have your raspi got an internet connection? Such as WiFi hotspots and so on?

can you give more info such as screenshots of wsjt-x?


#wsjt time sync #wsjt


Hi friends
I will appreciate your help:

install the image, I have the rasp time ok with my time zone but wsjtx does not decode ......, my usb card does work. What will be the problem? Do I necessarily need an additional GPS?
Thank you
christian from chile

Re: RadioPi Hotspot


Hi Pete!

We are glad you like it!
A building in hotspot on RadioPi maybe be the next solution. We are starting to talk about it.
Thanks for your sharing!

Hope cu on air!


BG6LH, Zhiming Cao

Re: RadioPi Hotspot

Peter Silverstein

Thank you very much for the guidance!

Using some instructions from KM4ACK on YouTube, I created a hotspot on the pi that allows me to connect the Android tablet over wifi without an internet connection. This works well in remote areas where there's no cell service.
It also seems compatible with
RadioPi's built in CAPABILITY. I'm using RadioPi everyday now.

Thanks again for your help!



Re: RadioPi Hotspot


I was using an iPhone to share a wifi connection. It does not work fine every time......However, make a reboot, it will work...


Re: RadioPi Hotspot


hi, Peter!

I suggest you set up a hotspot named "radiopi" on your smartphone.
and, both your radiopi and your android tablet connect to your smartphone's hotspot.

We'd saved a wi-fi connection to hotspot "radiopi" on the radiopi's image, already.

For more information, I hope this document could give help.


RadioPi Hotspot

Peter Silverstein

Thanks for this outstanding field operating system!

I'd like to set up in the field and operate the pi headless with an Android tablet, but can't see the RadioPi hotspot listed in available networks on the tablet.

Is there something I need to enable to get it working?



Performance test of Raspberry Pi FT8 decoding


After we launched the first version of RadoPi, many OMs would have such a question: Can the performance of a small Raspberry Pi meet the FT8 QSO? For this reason, we designed a test as close as the actual scene , and used the same data to make a comparison between different versions of RPis and PCs.

Let’s start with the conclusion:

  • Fixed/remote/QRO:
    • Raspberry Pi 4B 2G is preferred.
    • It can basically guarantee normal communication with deep decoding.
    • However, when there are many parallel signals, your QSO partner may need to repeat tx multiple times.
  • Portable/QRP:
    • At least Raspberry Pi 3B should be used.
    • Without deep decoding (ignoring weak signals), it can basically communicate normally when there are less than 10 parallel signals.
    • When there are many parallel signals, your QSO partner may need to repeat tx multiple times.
  • SWL:
    • Without deep decoding, Raspberry Pi 2B can work well.
    • Open depth decoding requires 3B or above model to ensure basically complete decoding.

Test design:

{ Below is translated by Google Translate from our Chinese article }
A series of dedicated amateur radio communication protocols such as JT65 and FT8 invented by Professor Joe Taylor, through a series of complex information compression, redundancy, error correction, analysis and prediction, to achieve reliable communication on weak signals on high-noise links. Approaching the theoretical limit of Shannon’s theorem. The complex algorithm of this superior mode determines that it relies more heavily on computing performance than traditional digital modes such as RTTY/AFSK/PSK31.
{ Above is translated by Google Translate from our Chinese articale }

As you know, a full FT8 RX/TX cycle is 30 seconds, one of 15 seconds is the receiving and the other 15 seconds is the transmitting. The actual signal carrier time in each cycle is about 12 ~ 13 seconds.

In the case of normal RX/TX, the decoding starts at the end of the 12.6 seconds of carrier. There is about 2.4 seconds to decode and make preparations for the next 15 seconds to transmission.

Because FT8 decodes all signals in the 0 ~ 3kHz audio bandwidth in most cases, the more parallel signals, the slower decoding will be. Generally, signals with a high SNR will be decoded faster, while signals with a low SNR will be slower. If it fails to decode within 2.4 seconds, the decoding process will continue to run. If the entire decoding process is completed within 15 seconds, it can catch up with the next transmission cycle. If it is too long to decode, your QSO partner may lose patience and halt TX.

Therefore, to achieve a relatively normal communication, the decoding of most signals in one cycle should ideally not exceed 2.4 seconds. Weak signals require more computing, and a faster CPU help to deep decoding of weak signals.

In order to have a stable benchmark, we have selected 12 signal segments recorded by WSJT-X that are actually received and distributed at different times in a day, all of which are .wav files. The FT8 signals contained in each file range from 3 to 21 (receiving equipment is IC-7300, filter bandwidth 3.2kHz, a 22.4 meters long wire antenna, 20m band), using WSJT-X 2.2.2 built-in jt9 decoding Engine (consistent with the background call of the WSJT-X‘s GUI).

Each model of hardwares runs the following two test cases.

Case 1. Normal Decode, the average time to decode all 12 signal segments:

First, use the following command to decode all 12 signal segments.

~/ft8_samples $ jt9 -8 -d 1 *.wav

Use the command to decode 12 signal segments 
Above: Use the command to decode 12 signal segments

When decoding is done, a summary file of timer.out will be generated in the directory, and the data in the red circle is the total decoding time we need. Divide this value by 12 to get the average decoding time per segment.

View the total decoding time of 12 signal segments 
Above: View the total decoding time of 12 signal segments

Case 2. Deep Decode, the single segment with the most parallel signals

First, delete the decoding prediction file jt9_wisdom.dat generated by the previous case:

~/ft8_samples $ rm jt9_wisdom.dat

Then, deep decode the signal segment 200614_153000.wav:

~/ft8_samples $ jt9 -8 -d 3 200614_153000.wav

Use the command to decode a single signal segment 
Above: Use the command to decode a single signal segment

View the total decoding time of a single signal segment 
Above: View the total decoding time of a single signal segment

Summary of test results:

We tested the above cases in:

  • 5 Raspberry Pi.
  • 2 Thinkpad, as reference.

The following is a summary of the results (the red words “秒”, pronounce “miǎo”, is “Second” in Chinese.) . “21个” In the last column means there are 21 calls were decoded when deep decoding the single segment.

 Above table: Summary of FT8 decoding performance of each model (unit: second).

There is a Chart of the 7 computers decoding time.

  • The Bule bar means Average Decoding Time of 12 segments on each computer.
  • The Orange bar means Total Decoding Time of single segment on each computer.
  • The longer bar means slower decoding performance .

Above: Comparison of FT8 decoding performance of various models (unit: second). Note: The longer bar means more time to decode, an not suitable for FT8 decoding.

It can be seen that:

  • For normal decoding, the RPi 4B, 2G RAM version, is pretty suitable.
  • The performance of the RPi 4B, 4G RAM version, is better than Thinkpad X61.

We don’t have a Raspberry Pi 3B+ , and there is no result about it. However, we uploaded the 12 signals’ wav files to If you are interested in this test, you can download and test it by yourself.
SHA256: df20f47858ee8c28cd2c722d46c9963dc4e7d0b4fe3edff1610d5ba5ff7d2420


BI1EIH, Translated by BG6LH and Google Translate.


RadioPi, the Hamradio Raspberry Pi OS image for dummies.


RadioPi, is another Hamradio Raspberry Pi OS image, for......linux dummies.. :)
RadioPi is optimized for remote control and field operation. You can just flash it into a SD card, and run it on a Raspberry Pi computer directly.
RadioPi is desgined to:
  • Control your rig, run FT8, log QSO, and upload to LoTW.
  • Control your rig at home, when you take a walk in a park.
  • Try to make QSO by SDR and satellites.
  • Go to field operation without your laptop computer.
Features of RadioPi
  • Default user: pi
    Password: radiopi599
  • Setting Locale to en_US.UTF8. Chinese Language also was supported. fcitx fcitx-googlepinyin input method are installed
  • chornyd and gpsd for timing service
  • non-ham softwares are removed。the zipped distro IMG is only 1.28GB
  • Gnuradio and SDR libs are installed, such as hackrf, gr-osmosdr and so on. it is ready for more fun of SDR testings
  • SSH, VNC, Avahi services were enabled for field operation
  • A hotspot "radiopi" was installed for field operation
  • A 1280x1024@60Hz screen resolution was installed. it's better for iPad's VNC to remote control RadioPi.
Software on RadioPi
  • WSJT-X, 2.2.2, The extreme weak-signal communication software
  • JTDX, 2.1.0-RC150, More features than WSJT-X
  • TQSL, 2.4.3, The ARRL LoTW QSL client
  • Xlog, 2.0.14, A light weight logging program
  • Fldigi, 4.1.01, Amateur Radio in Digital Modes
  • CHIRP, daily20190104, Programming amateur radios, supporting a large number of manufacturers and models, such as Yaesu, Icom, Kenwood, Baofeng, TYT and so on.
  • QSSTV, 9.2.6, Receiving and transmitting SSTV and HAMDRM(DSSTV)
  • GNU Radio Companion,, GNU Radio is a free & open-source software development toolkit that provides signal processing blocks to implement software radios.
  • GQRX, 2.11.5, Gqrx is an open source software defined radio receiver (SDR) powered by the GNU Radio and the Qt graphical toolkit.
  • CubicSDR, 0.2.5, Cross-platform SDR application
  • Direwolf, 1.4, A software "soundcard" modem/TNC and APRS en/decoder
  • Gpredict, 2.3-33-gca42d22-1, A real-time satellite tracking and orbit prediction application
  • VNC Server, 6.7.2, To remote control your rig remotely by VNC
  • PulseAudio Preferences, 1.1, For simultaneous line-output, and transmit monitoring.
Full of RadioPi User Guide
More info about RadioPi:
Please visit:
or, email us: askom[at]

Flash a card and go to field!

1 - 9 of 9