Radiofax: RBW41 (?) Murmansk, Russia

Since we’re probably on the “tail-end” of the radiofax technological lifecycle, I’ve actually been quietly hoping to receive something from Russia as on historical information, it seems they’ve been rather unique.

While most stations today have settled on 120LPM with 576IOC, with Kyodo News using 60LPM, apparently many Soviet era stations operated 90LPM with 576IOC. There were other stations which operated IOCs including 288 and 352 but it seems rather unlikely to meet them now. Indeed, the last NWS/NOAA radiofacsimile schedule document seems to claim that RBW41 and RBW48 both are transmitting 120/576 with only the schedule broadcast at 90/576.

Thankfully, a reader by the name of Vitor informed me of the whereabouts and the transmission characteristics of RBW41 – namely 6328.5kHz (center frequency) with an inverted/negative shift with charts between 1300-1500 and 2000-2100UTC. With this tip-off, I decided to go searching … and alas, I can report some success!

Charts presented represent snapshots in time and are not kept up to date. Charts are not to be used for navigation purposes. Copyright in the faxed materials belongs to the transmitting agency in Russia, the exact entity is unknown. Charts are reproduced for documentary purposes to illustrate the technical characteristics of transmission, the formatting of the faxed information and as proof of activity of the station, propagation conditions and the reception of the charts at the stated times and dates.

Faxes

The station, believed to be RBW41 of Murmansk, Russia, required a change to my normal kiwifax.py-based workflow. For one, its location meant that it was hard to find a receiver that could get a good signal, so some signal hunting was required to optimise the choice of receiver. I first waited for a fax to appear based on listening into the closest station. Once certain a fax was being sent, I looked around to find the signal at neighbouring receivers. In the end, I found four receivers that provided visibility of the signal good enough to recognise a fax was taking place – SM2BYC, RZ3DVP, ArcticSDR and Kiwi-IVA.

To most efficiently monitor the station, a number of changes to the kiwifax.py script was necessary as RBW41 does not utilise APT start/stop tones. With some modifications, it became possible to monitor more effectively and in-parallel across receivers. Unfortunately, some receivers had daily limits, so I took care to choose the most advantageous time windows to use them.

A typical semi-decent reception actually looks like the above as the signal isn’t particularly easy to receive. As RBW41 is the only station I know of at present that uses negative shift where white is -400Hz and black is +400Hz, it results in an inverted picture when decoded using regular tools. To resolve this, you can invert the received image in an editor, use the invert feature in your decoder or tune 1.9kHz above the centre frequency and demodulate using the LSB mode.

Slant optimisation was performed incrementally – through trial and error, thus the later charts show reducing levels of slant. Faxes are presented in order of reception date and time.

I first received the station at 1330UTC on 3rd February 2019. Choosing the KiwiSDR closest to Murmansk, I was able to get mostly noisy copy but with a few “strong” lines. This gave me hope that the station was alive. Prior to the start of the image and after the image, tuning tone (-400Hz) was present. It appears that images are sent with a black border around them and a white margin in the image area.

The next day (4th), I tried again, but instead trying a different KiwiSDR nearby. This time, I was able to improve the reception a little more, and Russian text was now evident. The chart is definitely a surface analysis or prognosis of sorts.

On 5th February, I decided to try an SDR that was further but within Russia – I suspected this may provide a better copy possibly as the directionality of antennas may affect reception and this could be in the “rear” of the beam. The copy improved even more, to the point that a date is now visible in the copy. Scheduling at the station appears to be somewhat loose, with this fax coming in one minute early at 1329UTC.

Of note is that the image area of this chart exhibits wobbles which suggests the scanning timebase is not consistent. Most modern radiofax stations generate the signals electronically, resulting in crystal-locked accuracy or better, this may indeed be indicative of a “cold” mechanical machine that is slowly warming up and being consistently adjusted by a feedback loop to attempt to maintain a consistent scan rate. This suggests to me that this could be a modified photofax style machine which was used with telephones in the past to “wire” photographs for press usage just like in this video.

Owing to the better signal, other charts started coming out of the noise. Although QRM affected as well, this chart was received at 1030UTC on 7th February 2019 which is a first glimpse at a different type of chart.

The surface chart came a minute later on the 7th, being broadcast at 1331UTC. A little QRM was present and the white tone post-chart seemed to be a bit shorter than usual. This suggests the charts are sent manually.

Staying on-frequency was rewarding as at 1449UTC, a different type of chart was sent. There is a logo of sorts, but it seems the lack of horizontal scanning timebase stability has corrupted all of the text. It may indeed be IOC288 or IOC384 with its poor resolution.

On 8th February 2019, at 1331UTC, another surface chart is received.

However, it was then discovered that it was retransmitted at 2000UTC and this was received much better. This is seemingly as good as it gets for clarity, which is only barely sufficiently clear to read the larger numbers “if you squint”.

On 9th February 2019, some QRM seems to have “clobbered” the beginning of the fax, but this was received at 0330UTC, so it seems the station is actually active at other times as well.

Later on the same day at 1329UTC, the standard surface chart but reception is very poor. But you should never give up … since things can change.

Finally, at 2001UTC, the chart is again received rather pristine – it seems that propagation favours this time to the KiwiSDR receivers I chose.

Conclusion

It was interesting that around-the-clock monitoring of the “suspected” RBW41 station’s frequency was able to turn up charts being sent on a regular schedule. While it was not as exotic as I expected, using 120LPM mode, the station does not utilise APT tones and uses a negative shift which makes it unique.

The charts sent are all short-form charts of a fixed size. That, along with the lack of APT tones, low clarity, sometimes wavering edges and inverted images makes me suspect this is being “scanned” by a modified photofax machine formerly used to transmit photograph prints over telephone lines. Apparently, this was common amongst “negative” printing stations.

As a result, receiving the station is tricky due to weak signals, propagation issues, lack of APT tones – even when all the stars align, the text on the charts are almost unreadable, clumping together. It is, however, still noteworthy that the station does still operate and it’s all thanks to Vitor for letting me know where and when to look for it.

About lui_gough

I'm a bit of a nut for electronics, computing, photography, radio, satellite and other technical hobbies. Click for more about me!
This entry was posted in Radio and tagged , , , , . Bookmark the permalink.

Error: Comment is Missing!