2024-08-29 09:20 UTC and 11:00 UTC: ISS Cubesat Deployment (JAXA) J-SSOD#31 - [CosmoGirl-Sat, SaganSat0, SAKURA, Wisseed Sat, Binar-2, -3, -4]

[fredy]

The latest TLE sets from space-track are now assigned:

0 OBJECT WS
1 60952U 98067WS  24249.59968314  .00148388  00000-0  24068-2 0  9992
2 60952  51.6374 276.9165 0013188 328.9205  31.1004 15.51641754  1128
0 OBJECT WT
1 60953U 98067WT  24249.53555050  .00108777  00000-0  17968-2 0  9995
2 60953  51.6378 277.2379 0014118 325.8685  34.1396 15.51287953  1119
0 OBJECT WU
1 60954U 98067WU  24249.34193350  .00130529  00000-0  21391-2 0  9991
2 60954  51.6378 278.1948 0014175 325.8848  34.1229 15.51412783  1088
0 OBJECT WV
1 60955U 98067WV  24249.33963561  .00244589  00000-0  37903-2 0  9998
2 60955  51.6376 278.1799 0013471 327.9417  32.0752 15.52538906  1089
0 OBJECT WW
1 60956U 98067WW  24249.53477633  .00147533  00000-0  23810-2 0  9994
2 60956  51.6372 277.2321 0013263 328.9922  31.0284 15.51783905  1100
0 OBJECT WX
1 60957U 98067WX  24249.34155771  .00176829  00000-0  28431-2 0  9994
2 60957  51.6373 278.1934 0013327 327.9301  32.0877 15.51771033  3068
0 OBJECT WY
1 60958U 98067WY  24249.53570885  .00104661  00000-0  17345-2 0  9990
2 60958  51.6378 277.2392 0014026 325.3742  34.6333 15.51216477  1104

[fredy]

Below is the current status update after TLE sets have been assigned by space-track. We need some more separation to identify most of the satellites.

Temporary NORAD ID	Satellite	NORAD ID to follow	Other NORAD IDs	Identified	Deployed	Reception	Last Update
RECEIVED AND DEMODULATED
-	CosmoGirl-Sat	60953	-	YES	YES	Received	Identified, here are the ikhnos results
99196	Sakura	60954	60952(unlikely)	-	YES	Received	Continue to follow 60954 , other possible 60952(unlikely) removed 60956, 60957
98808	Binar-2	60956	60957	-	YES	Received	Continue to follow 60956 , removed 60954
98807	Binar-3	60957	60956	-	YES	Received	Continue to follow 60957 , removed 60954
98806	Binar-4	60952	60954(unlikely)	-	YES	Received	Changed to follow 60952 , removed 60953, 60958
NOT RECEIVED
98810	SaganSat0	60958	all of them	-	YES	Not Received	Changed randomly to follow 60958 , all the rest unidentified objects are possible
IN DB WITHOUT FREQUENCY
98809	Wisseed Sat	60955	all of them	-	YES	No Transmitter	No change

[vk6tpw]

Hi @PE0SAT and @SA2KNG, you seem to be making some progress, thank you. please let me know if I can help any further. The other sample rate Jacob files might have been is 192000.

[VK6JAC]

@K4KDR Thank you for posting the wav decoder flowgraph. I have implemented that locally and we are starting to decode satellite data from the SatNOGS audio files (and getting a lot of CRC fails).

@PE0SAT Apologies on missing the sample rate for the files - I have only started using GNURadio this week so i’m not full bottle on all the information that is required! The CRC fail IQ file runs at 400k sample rate and requires a -22.5kHz offset to centre the beacon; the successful decode also runs at 400k and is centred at 0. I pass the signals through a rational resampler to interp and decimate down to 192k before running through the decode chain.

GNURadio_Binar this link contains all of the decoder flowgraphs I have been using for both SatNOGS and raw IQ files, and will reproduce the success/fail decode with the IQ snippets I posted earlier. They are functionally the same as what VK6TPW has posted earlier but broken out a little further for my own understanding of the signal path. I hope this can be of some help!

Jacob - VK6JAC

[PE0SAT]

The SatNOGS BINAR gr-satellites integration is working and sending frames to the network as can be seen in this obs https://network.satnogs.org/observations/10178037/

Due to the fact that the object identification isn’t permanent yet not all frames are combined under the same id in the db

Jan - PE0SAT

[VK6JAC]

Awesome news - the beacons decoded in that pass are giving us data through our decoder. The timestamp is incorrect, but we believe that to be an issue on the satellite side related to the primary flight computer and GPS as we get a matching error in our recorded IQ data. Huge thank you to everyone that has participated in the hunt to decode. A snippet of the data from @PE0SAT’s pass is below:

ID: Binar-2
Op Mode: SM
VK6BSP Binar234-Contact binarspace.com
Timestamp: 1/1/1 15:45:23

voltages
battery: 7.65664
solar_panel_di: 1.94883
solar_panel_do: 0.490153
solar_panel_xx: 2.2284
solar_panel_yy: 0.051664
v5: 4.98039

currents
3v3: 0.111158
battery: 0.215896

Once the commissioning phase is complete and we transition from safe mode into application mode, we are planning to start up our “Capture the Beacon” game - everyone can try transmitting to the satellite with their callsign to capture it. If your transmission is processed correctly the satellite will add your callsign to the beacon alongside a duration that you have owned it. Our team will post the exact details for the transmit frequency, packet format and start/end times before we start the game. I will find out if there are prizes we can send out for the winners.

[K4KDR]

Thanks again for sharing the GNU Radio flowgraph for the BINAR’s.

I added an input for UDP audio from GQRX + the SatNogs upload block and was able to get my first ‘live’ decode!

https://x.com/scott23192/status/1831993006775230717

[PE0SAT]

Will moving out of safe mode also give more output power ?

The signals are pretty weak and difficult to receive and decode.

[VK6JAC]

Unfortunately there is no provision for upping the TX power of the satellite radio after assembly, for any operating mode. The team is currently investigating why there are so many CRC fails - working theories include an issue with the antenna being only partially deployed, high transceiver/RF front end temperature and/or the lack of forward error correction. VK6TPW and I will keep everyone here informed as we work through the problem.

We have attempted a couple of transmissions to Binar-2, however the satellite hasn’t responded yet. To this end, we are also making modifications to our ground station to increase the RX and TX gain, by adding additional bandpass filtering before the preamplifiers and moving to a quad-stack yagi setup from our current dual. I will update everyone with the results from these changes once we have the parts installed.

[SA2KNG]

@VK6JAC

To this end, we are also making modifications to our ground station to increase the RX and TX gain, by adding additional bandpass filtering before the preamplifiers and moving to a quad-stack yagi setup from our current dual.

https://moonbounce.dk/hamradio/ham-radio-current-systems/70cm-eme-system.html#Upgrading-of-preamp

Filters before preamp is not optimal as you add the loss in the filters and never get that back. Make sure the actual preamp can handle big signals so it can handle the entire signal from the antennas. Or just buy one of those made for EME. The dual coax might sound like a hassle, but it means you can keep the nr of relays to a minimum in the complete system. DC shorted antennas help in limiting the out of band noise in my experience.

I don’t have high elevation passes of the satellites up in the north, so not super strong signals.
Received with a 15+15el X-yagi, max el of 10 deg, deviation guesstimated to 4800.
From #10189907, b205mini on gpsdo.

There’s no whitening, meaning the heavily biased parts will throw most demodulators off track. Disabling DC block and having precise frequency adjustment will compensate for this. It actually looks to be pretty spot on it’s frequency so only the RX needs to be calibrated in this case.

It would be nice with some GS documentation from your end (: (if possible)

[PE0SAT]

I have shared the IQ from observation 10187334 on the following download location, maybe easier to measure the deviation.

It is a full overhead pass received with a 9 elements DK7ZB (fixed position) AirSPY R2 and a Noolec wide-band LNA.

[SA2KNG]

Yeah, much better. That looks like 4500 deviation, odd one.

maaybe, choosing a less DC skewed filler at the end could alleviate it a bit…
01 01 01 01 01 01 01 01 01 01 01 01 00 03 01 ff ff
like switch to AA or 55 instead, manual whitening (:

[VK6JAC]

Current ground station RF setup. We proposed adding the filter between the power divider and the preamplifier on the mast to reduce the incoming power from local FM stations, as we are located in the middle of Perth and unfortunately get quite a bit of EMC noise across the spectrum.

To clarify my understanding - because the end of the packet containing the sequence of ‘01 01 …’ tends to hold the waveform in a low frequency transmit state, either the demodulator is losing synchronisation with the packet, or the RX hardware is erroneously blocking the signal as it appears like DC power, and by switching to a sequence of ‘AA AA …’ we more evenly spread the signal power across the bandwidth, improving the demodulation quality?

SA2KNG is correct, the deviation is 4800. The sequence of ‘01 01 …’ at the end of the packet you are seeing is a sequence of COBS encoded ‘00 00 00 …’ which would normally contain satellite data from the ADCS when we are performing operations such as detumbling. The encoder has converted that into ‘01 01 01 …’ terminated with the final ‘00’ before the ‘03 01’, which is the HWID of the satellite radio, and finally ‘FF FF’ is a placeholder for a software CRC (descoped due to time constraints).

Unfortunately this means that we aren’t able to change the sequence of ‘01 01 01 …’ without requesting an operation that would populate satellite data in those fields.

[fredy]

Identification has been completed for the satellites that we have received. Here is the latest status:

Temporary NORAD ID	Satellite	NORAD ID to follow	Other NORAD IDs	Identified	Deployed	Reception	Last Update
RECEIVED AND DEMODULATED
-	CosmoGirl-Sat	60953	-	YES	YES	Received	Identified, here are the ikhnos results
-	Sakura	60954	-	YES	YES	Received	Identified, here are the ikhnos results
-	Binar-2	60956	-	YES	YES	Received	Identified, here are the ikhnos results
-	Binar-3	60957	-	YES	YES	Received	Identified, here are the ikhnos results
-	Binar-4	60952	-	YES	YES	Received	Identified, here are the ikhnos results
NOT RECEIVED
98810	SaganSat0	60958	60955	-	YES	Not Received	No change
IN DB WITHOUT FREQUENCY
98809	Wisseed Sat	60955	60958	-	YES	No Transmitter	No change

[fredy]

I forgot in my previous post to mention that SaganSat0 has a mass and structure similar to Sakura and CosmoGirl-Sat satellites. This means that should be an object close to them. Given the mission of Wisseed Sat, it has different mass and structure that means it should be further for the rest of the group.

As we can see on the screenshot above, OBJECT WV (60955) is far away from the other satellites, while OBJECT WY (60958) is close to Sakura and CosmoGirl-Sat.

So, the currently selected NORAD IDs to follow for these unidentified satellites have a big chance to be the right ones. However, as in the past we have seen that the order of satellites can change easily and also not necessarily the same with their deployment order, we can not be 100% and identify them.

[SA2KNG]

@VK6JAC

To clarify my understanding - because the end of the packet containing the sequence of ‘01 01 …’ tends to hold the waveform in a low frequency transmit state, either the demodulator is losing synchronisation with the packet, or the RX hardware is erroneously blocking the signal as it appears like DC power, and by switching to a sequence of ‘AA AA …’ we more evenly spread the signal power across the bandwidth, improving the demodulation quality?

The DC block that is in the FSK demod uses 32 samples to find the balance and adjust accordingly. This works fine for whitened frames or ones that does not contain a lot of ones or zeroes in a row. Could you enable whitening perhaps ?

Enabling the DC block on the plot above looks like this and does not produce a crc ok frame:

As you can see, the 01 01 stuff at the end pulls the negative symbols to -0.25 instead of -1, this is also true for other parts of the symbols where the symbols leans to one side.
And as Jan pointed out, we have a method to invoke satellite specific settings using sat.cfg but is a manual station setting. I can add it to either 1 2 for stations that use my implementations.

If you want to play around with these plots, the tool I’m using is waveform_plot.py and the waveform.f32 is coming from adding --dump_path . to gr_satellites, just remember to cut it down to a few seconds worth of samples, else it will be pretty slow (:

edit: forgot to comment on station design.

How is the total phasing controlled between the antennas ?
This can become quite a problem and cause a lot of fading due to destructive interference.
Combining the antennas as close as possible and measuring the phase if the paths have more components in them, like relays and such. Then preamp as close as practically possible to the last divider/circulator.
Treating the antennas as one system one can play with filters and such later down the line, even in the operations room.
Maybe use one 4-port divider, less connectors means less losses. Not sure of the benefits to transmitting with only half the antennas.

[dl7ndr]

SAKURA is back to its initial beacon format SAKURA JS1YMY SIxxxx < data >

2024-09-16T17:48:29 sakura js1ymy si85ab a57ebc6229a10000
2024-09-16T17:50:28 sakura js1ymy si8513 a57fbb6229a10000

2024-09-16T21:01:57 sakura js1ymy si7dc4
2024-09-16T21:03:38 sakura js1ymy si763f a37abb6328a00000
2024-09-16T21:05:36 sakura js1ymy si792a

2024-09-16T22:37:27 sakura js1ymy si72dc a37bba6328a00000
2024-09-16T22:39:25 sakura js1ymy si787f a279ba642aa00000
2024-09-16T22:41:23 sakura js1ymy si8a9c a27bb96629a00000

2024-09-17T00:16:50 sakura js1ymy si73a8 a27ab96728a00000
2024-09-17T00:18:59 sakura js1ymy si6f1b

[dl7ndr]

SAKURA is back to normal beacon:

2024-09-17T16:55:04 sakura js1ymy js1yni a57ebe6328a00000
2024-09-17T16:57:00 sakura js1ymy js1yni a57ebe6328a10000

[dl7ndr]

Dashboard for CosmoGirl-Sat online. Please upload CW beacon data.

[yd9ckh]

how to upload?