VOA Radiogram program 104 this week was an experiment in reception of different digital modes. The same news story, “Battle for African Internet Users Stirs Freedom Fears,” was broadcast five times in succession using increasingly faster modes. The modes used were:
- MFSK32 (with image)
- MFSK64 (with image)
- 8PSK125F
- MFSK128 (with image)
- 8PSK250F
There was also a MFSK32 image at the end of the broadcast.
MFSK32 is used regularly on VOA Radiogram and MFSK64 and MFSK128 are faster modes that have been used less often. But the 8PSK125F and 8PSK250F modes were new to VOA Radiogram. 8PSK250F is over five times as fast as MFSK32.
Results for 5910 kHz
I decoded three of the VOA Radiogram broadcasts this weekend. I decoded the 5910 kHz broadcast at 0930 UTC on Saturday (5:30 AM EDT Saturday). The 5910 kHz signal was quite strong here. Unfortunately, I neglected to set the recording level correctly. This meant that the audio level was much too loud and the high audio levels were clipped.
I used Fldigi to decode the audio anyway and came up with these results.
- MFSK32 had two one word errors.
- MFSK64 had two one word errors.
- 8PSK125F had six one word errors and two unreadable paragraphs.
- MFSK128 had two one word errors.
- 8PSK250F had only three intact paragraphs.
This wasn’t really a fair test because of the clipped audio, but it would seem to indicate that the MFSK modes deal better with clipping than the 8PSK modes. I was surprised by the good results for MFSK128. It is four times faster than MFSK32.
Here are the images I decoded from the broadcast:
The clipping caused the first two images to be garbled, but the final two images were fine. I have no idea why they weren’t affected by the clipping.
Results for 17860 kHz
I also decoded the 17860 kHz broadcast at 1600 UTC on Saturday at 1600 UTC (12:00 PM EDT Saturday). The signal strength of 17860 kHz was poor again here this week, but it was a bit stronger than last week. Unfortunately, the signal weakened as the broadcast continued.
I came up with these decoding results:
- MFSK32 had no errors.
- MFSK64 had no errors.
- 8PSK125F had only one readable paragraph.
- MFSK128 didn’t decode.
- 8PSK250F didn’t decode.
This doesn’t necessarily say much about MFSK128 or 8PSK250F because my reception had weakened considerably by that time in the broadcast. None of the images decoded in this broadcast.
Results for 15670 kHz
I also decoded the 15670 kHz broadcast at 1930 UTC on Sunday (3:30 PM EDT Sunday), which came in stronger this week. The interference that affected the 15670 kHz broadcast for the previous two weeks wasn’t a problem this time. If it was still there, it was at levels that didn’t overpower VOA Radiogram.
Kim Andrew Elliott pointed out that the interference I described had similarities to shortwave jamming. Since one of the purposes of VOA Radiogram is to show how information could be broadcast into areas affected by jamming, my results might actually have significance.
I came up with these decoding results:
- MFSK32 had no errors.
- MFSK64 had no errors.
- 8PSK125F had only three intact paragraphs.
- MFSK128 had at least a one word error in all but one paragraph.
- 8PSK250F had only a handful of intact words.
Here are the images I decoded:
My results suggest that 8PSK125F and 8PSK250F, although quite fast, didn’t work very well at my location. I will be interested to see if they decoded better for people in other parts of the world.
Results from WRMI in 9955 kHz
WRMI also transmitted some digital images this weekend on 9955 kHz. Here is one of their MFSK32 ID’s I decoded around 1014 UTC on Saturday (6:14 AM EDT Saturday):
