DaveInDenver
Rising Sun Ham Guru
Was tuning around yesterday evening and found the Pt. Reyes NOAA station was coming in nice, so I decided to let it run a while.
The Weather Service and Coast Guard have for decades offered a continuous fax service sending out charts and forecasts to ships at sea. I imagine now most everyone has satellite service so it's maybe not used as much as it once was but all it takes is an HF radio and a computer with a sound card to receive them so it serves a backup, if nothing else. It's one of several services the USCG still do,
https://www.weather.gov/marine/uscg_broadcasts
Maybe you'll hear it and want to try. See attached file "ptreyeswefax-2021aug3-4.mp3.txt" and remove the .txt extension for an MP3 of what it sounds like. I was listening to the 8682 MHz NMC station. This is the spectrum at 2141 MDT (0341Z) , good signal strength, low noise floor. Didn't need any pre-amp and AGC wasn't working very hard for fade.
This morning, not so much. It was a rough copy and based on the quality of the faxes it started really giving out about 0100 MDT (0700Z) and when I checked on it this morning it was at the limit of legibility. Completely expected for 40m HF. This was the spectrum this morning 0538 MDT (1138Z), pre-amp on, AGC working hard with lots of RF gain to dig the signal out of the noise.
Anyway, since the signal is SSB centered on 8.682.000 and you'd normally tune down 1.9 KHz using USB. But my software was configured for 1.5 KHz off center but correct 800 Hz frequency shift. That worked but requires some investigation why I had it that way.
I use Fldigi to decode the faxes, it's called WEFAX576, which means WEather FAX IOC-576. IOC-576 is a radio facsimile standard, suffice to say you can Google it if you want.
http://www.w1hkj.com/
This is what it looks like during reception, the image building at 120 line per minute. A perfect image is typically about 1000 to a max of 1300 lines. Sometimes if it's noisy two images can run together, in which case the software will just eventually truncate the receive (I think at 2500 lines, it's configurable) and start a new image.
You can see how the images slowly get worse overnight as the signal-to-noise ratio decreases.
Images have a start and stop tone, which is the black bars at the top and bottom. Fldigi resets on the stop and stores the last received image.
They slant like this because NOAA's radios and computers are highly accurate while my station's clocks are not. I was using my Icom IC-7300, which has a TXCO and is pretty accurate and stable (tunes to 1 Hz, although I suspect it's more like 0.5 ppm, so +/- more like 4 or 5 Hz in accuracy at 9 MHz).
My laptop's clock, though, drifts a bit more. It'll move several milliseconds over an hour. I sync my laptop to a NTP server every 15 minutes but there's only so much you can do.
Anyway, all the accumulated clock error is why they slant in raw form. The slant can be corrected in software but for the purposes it's not necessary.
You can cheat and see what the faxes are supposed to look like: https://www.weather.gov/marine/ptreyes
The Weather Service and Coast Guard have for decades offered a continuous fax service sending out charts and forecasts to ships at sea. I imagine now most everyone has satellite service so it's maybe not used as much as it once was but all it takes is an HF radio and a computer with a sound card to receive them so it serves a backup, if nothing else. It's one of several services the USCG still do,
https://www.weather.gov/marine/uscg_broadcasts
Maybe you'll hear it and want to try. See attached file "ptreyeswefax-2021aug3-4.mp3.txt" and remove the .txt extension for an MP3 of what it sounds like. I was listening to the 8682 MHz NMC station. This is the spectrum at 2141 MDT (0341Z) , good signal strength, low noise floor. Didn't need any pre-amp and AGC wasn't working very hard for fade.
This morning, not so much. It was a rough copy and based on the quality of the faxes it started really giving out about 0100 MDT (0700Z) and when I checked on it this morning it was at the limit of legibility. Completely expected for 40m HF. This was the spectrum this morning 0538 MDT (1138Z), pre-amp on, AGC working hard with lots of RF gain to dig the signal out of the noise.
Anyway, since the signal is SSB centered on 8.682.000 and you'd normally tune down 1.9 KHz using USB. But my software was configured for 1.5 KHz off center but correct 800 Hz frequency shift. That worked but requires some investigation why I had it that way.
I use Fldigi to decode the faxes, it's called WEFAX576, which means WEather FAX IOC-576. IOC-576 is a radio facsimile standard, suffice to say you can Google it if you want.
http://www.w1hkj.com/
This is what it looks like during reception, the image building at 120 line per minute. A perfect image is typically about 1000 to a max of 1300 lines. Sometimes if it's noisy two images can run together, in which case the software will just eventually truncate the receive (I think at 2500 lines, it's configurable) and start a new image.
You can see how the images slowly get worse overnight as the signal-to-noise ratio decreases.
Images have a start and stop tone, which is the black bars at the top and bottom. Fldigi resets on the stop and stores the last received image.
They slant like this because NOAA's radios and computers are highly accurate while my station's clocks are not. I was using my Icom IC-7300, which has a TXCO and is pretty accurate and stable (tunes to 1 Hz, although I suspect it's more like 0.5 ppm, so +/- more like 4 or 5 Hz in accuracy at 9 MHz).
My laptop's clock, though, drifts a bit more. It'll move several milliseconds over an hour. I sync my laptop to a NTP server every 15 minutes but there's only so much you can do.
Anyway, all the accumulated clock error is why they slant in raw form. The slant can be corrected in software but for the purposes it's not necessary.
You can cheat and see what the faxes are supposed to look like: https://www.weather.gov/marine/ptreyes
Last edited:
