Transponders 1 & 2
Some comparisons of the two transponders due to be in use by EUMETCast. Both transponders are on the EUTELSAT 10A satellite, and as of February 2017 transponder 1 is in operational use for the Basic Service (BS) and High Volume Service (HVS-1), transponder 2 is undergoing tests prior to its transition to the second High Volume Service (HVS-2) later in the year.
Using Airspy SDR receiver and SpectrumSpy software
A - Normal operation with the variation of signal level apparent as the different colours of the waterfall plot beneath the spectrum.
B - shows the level variation during a sweep of the spectrum analyser software. Note that the level in the first part of the scan appears higher than that during most of the scan.
C - showing the apparent energy below the nominal bandwidth. Analyser artefact, effect of mode switching or another transponder spreading?
Transponder 2 shows a well-defined ~15 MHz bandwidth, but with the tail of the next LF transponder around 11362 MHz (you need to add 9750 to the frequencies on the graph). Es/No is around 12-13 dB.
Using CrazyScan tool
First, a plot of the RF level covering the two transponders, using the CrazyScan software and the TBS 220.127.116.11 driver under Windows-10/32. We see that the RF level of transponder 2 is rather down on transponder 1, and that if appears to have a nearby signal which would prevent it's use at full bandwidth at the moment. Obviously this will be resolved before the transponder comes into full service.
Note that the polarity switching was not available on this RF feed, so only the horizontally polarised signals are measured.
Also note that I don't believe the RF level values (and will take this up with CrazyCat). The SNR on transponder 2 shows as 13.0 dB, but looks more like 5.5 dB.
Transponder 1 constellation
More interesting are the constellation plots. That from transponder 1 quite clearly shows the 8PSK MODCOD as clusters around 12 o'clock, 1:30, 3 o'clock and so forth. This has the 16APSK data superimposed visible as less well-defined clusters towards the centre of the plot. The rate is 33,000 kSps.
Transponder 2 constellation
At the time of testing, transponder 2 was running 8PSK 5/6 at a 15,000 kSps rate. The clusters are well-defined.
Errors on T2
Like others, with the present configuration I do see more errors on transponder 2 versus transponder 1. At the moment my transponder 1 isn't error-free likely due to local effects (e.g. wind bursts), but here's my comparison:
CrazyCat also published a comparison, although that seems to have been removed.
Here's what I see during a file transfer on the test transponder 2 configuration. Whether it's an artefact of the interaction between the transponder's resting state (and what "fill" DVB-S2 pattern), or whether it's genuine I don't know, but here are the three plots, showing during a file transfer, during a gap, and after the gap showing the duration of the resting state.
During the transfer
During the gap
After the gap