Uplink frequency drift correction using downlink of own signal?

  • Hi,
    I am currently building an groundstation for QO-100 and thought about replacing the GPSDO with a piece of software:

    There are gnuradio flowgraphs like this on github which lock the received signal to the nb transponders beacons using a costas loop and pipe the stabilized downlink spectrum to gqrx or similar.
    My idea is as follows: The same flowgraph will be expanded to transmit the uplink signal. The receiving part then also tracks my uplink signal received from the satellite and corrects the transmit frequency if the signal is drifting. Is this practical and or even possible? It's harder to track a changing ssb signal than a constant carrier I presume.

    Regards,
    Joshua,
    DL3JOP

  • Hi, I try to rephrase my question:

    I lock my recieving software of one of the beacons of QO-100 which eleminates signal drift effectively becuase the received signal will be as stable as the beacon itself.

    Now comes the transmitting part.
    The LimeSdr Mini as 4ppm drift when running for a longer time which results in a maximal drift of 9.6kHz at 2400Mhz which is way to high.
    The usual way would be to stabilize the LO using an GPSDO or other frequency standard.
    My approach instead would be:
    Since the received signal of the transponder is drift-corrected as mentioned above you can see your actual transmitting frequency when transmitting. So as a consequence you can precisly see how your transmitted signal drifts around on the downlink due to the 4ppm drift of the LO. As a next step you start tracking your own signal on the downlink and its drift and correct the drift by actively changing the transmitting frequency.

    Because the whole system is locked on the Beacons frequency, there is no cat-bites-tail problem.


  • Perfectly logical and it would work if your signal was a tone. I think it would be very hard to do with SSB but it might be possible. Perhaps sub audible tone on the SSB might help, or even a DC carrier.


    Mike

  • I want to built portable a station with as less components as possible. Second I wanted to know if someone already had the same idea?

    I think the best compromise will probably be to replace the GPSDO with a TCXO. It is simple, cheap and stable. At least the TCXO is not exposed to high temperature fluctuations. So do not position the TCXO in, on or near the PA ;-)


    The tx chain of my own simple station is also stabilized "only" by a TCXO. I am happy with that.

  • Hi,
    thank you all for your answers. The last days I worked on a gnu radio flowgraph which is capable of generating LSB/USB, now I'm experimenting how I could lock my loop onto the signal. Using an sub-audible tone is hard (at least for me) due to the small bandwidth.
    Otherwise I will go with the TCXO/OCXO approach.

    Regards and 73,
    Joshua,
    DL3JOP

  • Hallo,

    the LimeSDR Mini has three thermal sources. To reduce thermal drift, the board is mounted upside down on a big plate of aluminium. The plate has the same size as the Laptop I use for portable operation. All the other components like USB-Hub, sound card, power supply, USB-Relay, PTT-Interface, PA and some connectors are mounted on this plate too. The thermal capacity of the plate helps to avoid fast change of the LimeSDRs TCXO. Between the Lime and the Aliminium plate there are a little milled peace of aluminium and silicon thermal pads. This is for levelling the different hights of the Lime-Chips and thermal coupling.
    At least, the TX-Frequency changes a bit with the environmental temperature beeing outside. Adjusting this once at the beginning of a session with XIT of the Console will give sufficient stability.



    The upper cover, normal carrying the Laptop is removed. The lower floor gives room for extensions. At present there is a 12V-power supply for the Laptop. The complete radio is running with 12V too.



    Here is the LimeSDR Mini mounted upside down. No need for a fan. The hight of the board is fitting with the USB-Hub left.

    73s
    Andreas