AMSAT-DL Down Converter v3d

1st Official Post

    Hello to the group,


    I plan to use the two types ( OCXO /GPS) of Amsat downconverters to listen to the NB transponder in a first step ;

    for this, the LNB signal is injected in the upper left F connector and the 144 Mhz signal comes from the upper right bnc connector.

    The jumpers voltage are in position 'by default' and the 4 leds on the right side of the pcb are illuminated except the 3rd one (green) which is blinking.

    The output frequency selection is on 6 which corresponds to 144.550 Mhz.There is hiss but no audio.


    What am i doing wrong ?


    73

    tnx

    renauld on4kvi

    Ok Renauld,

    I have a similar problem here.

    If I use a short cable it works fine. If I try to work over the long distance to the dish, it did not work.

    If I use the LNB without the amsat downconverter it works well.

    Maybe a LNB 14/18V power problem....


    73´s de Robert

    • Official Post
    Quote from on4kvi

    i use a lnb with an internal clock modified by tcxo ;

    the previous amsat downconverter works like a charm !

    It should work well too...


    I'm not sure if you have the jumper's right.. LED's definitely look good.. all on and the green one blinking...

    What is your TCXO frequency in the LNB?


    You might also wwant to look here , but it's mainly in germnan:  AMSAT-DL Sende- und Empfangsmischer für QO-100 (P4-A)

    Hi Renauld,


    Running the Downconverter V3d with an unmodified LNB may cause an unspecified frequency shift, maybe out of band.

    For the calculation of the output frequency (6 ... 144,55 MHz) the converter needs to "know" the LNB frequency, which is not the case if you don't use the clock delivered from the downconverter.


    By default the downconverter operates with 24 MHz clock. Your LNB may work with a 25 MHz TCXO, so you need to jumper the downconverter to 25 MHz by putting one jumper as shown in the documentation. Then you should get a signal on 144,55 MHz, but with some shift caused by the different frequencies of your TCXO and the downconverter's clock.


    If that works you should consider to modify the LNB to get a more precise output frequency.


    vy 73, Kurt, DJ0ABR

    Tried here again, with a cable length of 45m the 24MHz level no longer seems to be sufficient for the PLL. I see the noise from the LNB but no signal. No problem with 16dB attenuator between LNB and converter.

    DD4YR

    Hi Robert,


    Extracted from the Amsat Converter Doc : "Attention: The reference clock is only applied to the WB-LNB connection."


    So I do not understand further more....


    So do we need to have the NB AND WB together connected to the LNB ? I was using only the NB part.

    73

    renauld on4kvi

    Now changed to another dish (1,2m PF) with POTY, Koax lenght abt. 16m, connecoted like the pic above als Single LNB used.

    All works well.

    Frequency offset in the NB about 1.2 kHz.

    Is there a way to correct that?


    73´s de Robert, DD4YR

    • Official Post

    I think there is a complete misunderstanding of the concept of the new V3d Down-converter.


    The above wiring is only needed when you have a Single-LNB which is modified for external Reference-Clock provided by the V3d-DownConverter!


    For this the XTAL in the LNB has to be removed and *no* TCXO installed in the LNB.. Just a few components to inject the reference prided on the WB-In of the AMSAT Converter. It is not possible to feed the Referenc-Clock to an unmodified LNB .


    If you want to use an LNB which was already modified with adding a TCXO in the LNB, the above is senseless. In that case you can continue to use the NB-IN, but use the correct Jumper 10 A/B settings..


    If you want to take advantage of all functions of the new DownConverter V3d, you should modify a Twin-LNB as described here.

    I have some questions about my amsat DL downconverter v3.

    I use the one with the GPS reciever.

    When I connect the downconvertor to my modified LNB and my reciever I still hear some drift in the RX of the satelite.

    I have connect the LCD display to the downconvertor to check if the GPS is locked.

    When the GPS is locked there is still drift on the RX side.

    When you touch te PCB or the onboard TXCO you see and hear a massive freq change.

    How is this possible when the GPS is locked.

    It looks like the GPS reference / lock does nothing to the LNB reference signal.

    Is this right ?

    No, the LNB reference is linked to the GPS signal.

    Is the converter in a housing?

    The TCXO is delicate and needs to be protected from draft.


    For experimental i have done it this way:

    Für ungeduldige,

    bei denen es die Konverterplatine während dem experimentieren noch nicht ins endgültige Gehäuse geschafft hat:

    Ein Stück leichter Schaumstoff mit Tape auf dem TCXO fixiert bewirkt Wunder was die Frequenzstabilität betrifft!


    73´s de Robert


    &thumbnail=1





    Robert - DD4YR - JN57QS

    @PA1PAS: the GPS reference frequency goes through an extremely slow PLL ( a couple of seconds). This is required because the GPS module output has a very precise frequency, but has a huge jitter.

    The PLL is too slow to compensate a short term drift caused by a temperature change of the TCXOs.

    A good enclosure helps to solve that, and additionally Robert's hint to put a piece of foam over the TCXOs also helps a lot.

    All frequencies, including the LNB reference are derived from this PLL.


    73, Kurt

    Hi,

    i watched the Amsat Downconverter V3d with GPS for quite a while and have some trouble with it.

    Firstly, It "crashes" now and then, reboots and starts to work again.

    Secondly, there is a more or less big offset to the desired frequency. This affects the reference for the lnb as well as the 10Mhz output. I'm sure it also affects the 40Mhz output. The description in the Amsat Shop gives an explanation for that, but i wonder why this could not be solved. The Downconverter essentially is a another breed of the Leo Bodnar Reference. Correct me if i'm wrong, but the Leo Bodnar reference is able to output a stable reference, low jitter AND correct frequency without offset.

    So why can't the Amsat Downconverter?

    73, Martin

    • Official Post

    About frequency offset: I don't know this particular circuit but I assume the reference frequency generated by the GPS engine is an internal clock signal which is programmable. Hence the required 1Hz ('few Hz') offset on the 10 or 40 MHz level.

    Afaik the Leo Bodnar unit has the same 'problems' and the solution to offset a tiny bit on the reference level to reduce jitter is known also there.

    It all depends on the requirement.. if you need some reference frequency for your counter to display the correct number you go for accuracy. If you need something for narrowband ops you go for stability.

    Can it be solved? Sure.. but then you need to correct the OCXO with a GPS engine and run it with really long time constants (1000's of seconds!). Then you would need to shield all of this against temperature variations (1/10 degC is noticable!) which needs a lot of volume and mass.

    Can it be solved for the price listed in the AMSAT-DL shop: no. A suitable OCXO itself starts at 100 Eur alone. The OCXO on the board would likely not be suitable for GPS-control due to its instability.

    For those who want to know more about this very interesting topic (how to build a GPS stabilised oscillator) I can recommend two sources:

    Uli Bangert DF6JB (SK): http://ulrich-bangert.de/AMSAT-Journal.pdf

    And from the former AMSAT-NA president Rick Hambly W2GPS: https://gpstime.com/files/TOW/tow-time2015.pdf


    73s Achim


    PS: and that's why I believe only very few frequency 'measurements' made of the QO-100 beacons.

    Hi Achim,

    well, if 23.2Hz is a few Hz, so be it. I left my Counter on for 3 weeks now, fed it with a GPS-controlled signal source built by DJ8LC, which i believe is accurate. The counter settled at 10.000,000 Mhz after 2 1/2 weeks now, so i believe it is accurate, too. The radout says 10.000,0232 when connected to the V3d. The 24Mhz reference for the lnb is also off, which results in an offset in the magnitude of 23khz.

    That is a bit more than the "few 100hz" stated in the manual. So if i connect my entire station to these references, which i did, everything is off. Downlink, 70cm Transverter, Upconverter, even the Transmitter itself . Sure, you can get rid of the errors by carefully adjusting the readout on your transmitter to match the tweaked readout on sdr console. Currently, i must set the LO in sdr console to 9360.017.8 to have the psk beacon on 750 and to compensate for the offset happening in the Oscillators for 70cm (102.5Mhz) , LZ5HP (10Mhz),LNB (24Mhz). That somehow is not satisfactory. Am i too picky?