AMSAT-DL Downconverter V3D Stability ?

  • Good Evening

    Last December I purchased a V3D downconverter, GPS module and the modified LNB from the AMSAT-DL shop.

    Due to a house move I have only recently unpacked and installed the connectors as set out in the WiKi pages.

    I aligned my 80cm dish using a cheap £18 pound modified ebay LNB and then installed the V3D, GPS and LNB.

    All lights on the board and GPS came on as per the WIKI, however it was immediately noticeable that the stability was, well unstable, indeed the CW beacon was drifting wildly.

    Now I have double read the WiKi, I have double checked the connections, I have tried using the board without the GPS as set out in the WiKi, still unstable,

    I note from the Forum that others have mentioned the same issue and that a "box" is required to prevent "wind" on the TCXO, Ive also seen that others have put foam on the TCXO.....

    At this time I am just hugely disappointed with this board, at Eur 353 I didnt expect it not to work without having to then additionally purchase a box etc

    Tomorrow hopefully I can ( with some assistance ) test the board further at our local club with test equipment, incase I have made some error, but for now it is very puzzling that the board, in an unboxed state can be so unstable and completely unusable.

    Dave M0GIW

  • Dave,

    You need to look at the design concept again. This is not a typical PLL solution hanging on a GPSDO because of too much jitter. The TCXO is slowly synchronized and stabilized with the GPS sync. Everything is described in detail in the documentation. A housing has always been the default for this concept. You could have at least clamped a piece of foam on the TCXO to see the effect?


  • The reason is simple: Air draft will have an impact on the frequency any (tc/oc)xo delivers. The comparator circuit senses the gps generated frequency differs from the frequency generated by the xo , so tries to provide the correct control voltage to compensate for this.

    The xo overshoots slightly, the control voltage will be corrected and so on, until the xo frequency is the same as the gps control frequency. This is a slow process and being slow is good, otherwise you'd experience a lot of jitter.

    You can minimize this when you put the board in an enclosure (plus put some foam on the tcxo) . Or use a ocxo which already is some sort of tcxo but with an air tight enclosure and heater built in. But even then i'd not skimp on an enclosure.

    And please remember, the gps only is there for the long term accuracy, while the tcxo is there for the short term stabilty. It takes a while for the system to settle. Hours at least, days for even better accuracy. Air draft will mess this up and the play starts again.

    73, Martin

  • Hi Dave,

    You might want to look at a Euro 2,000 radio called the IC-9700. Icom put a piece of foam on the boards XCO to avoid drift and for some special applications it still has a significant drift. If you regret paying Euro 350 for a nice piece of hardware from Amsat-Dl, either design and build your own, buy something else, or please sell it. There will be enough takers.

    My 2 cents worth on the whining.... :)

  • OTOH, if you put it in a box but the cw beacon still whines even after you allowed the downconverter to settle, there might be a problem other than air draft. Can you make a screen video of your sdr software showing the cw beacon so we can see and hear what's going on? Please also list all equipment involved, like dish-size, lnb, receiver.

    As a reference how it should sound (well, I think it should sound like this) , see https://www.<mycall>.de/uncategorized/qo-100-beacon-snr/

    and https://www.<mycall>.de/wp-content/uploads/2021/10/qo-100_beacon-2021-10-04_10.12.54.mp4

    Replace <mycall>.

    BTW, place the box where vibrations are minimal, preferably on a piece of foam. I once operated it on my operator desk and even bumps from your fist on the desk can be heard as small frequency jumps when you listen to the cw beacon.

    Remember, a frequency variation of 10 milliHertz results in a 3.9Hz shift after multiplied by 390. You will hear that. The crystal in the tcxo after all is a mechanical device. It wants to sit still (and flat) for optimal work.

    73, Martin

  • Good Morning

    I e-mailed Amsat-DL about returning this board but received no reply.

    Having spent some time with this board in the Lab at Finningley ARS we have finally achieved real stability.

    No need for a case, no problem with draft and no problem with vibrations tapping on the desk.

    We have ditched the supplied GPS, and taken the TCXO out of circuit and injected a stable 40Mhz ref from a Leo Bodnar GPS.

    The Board is now stable on the test rig, however there is a huge amount of Gain on the IF so we added some Attenuation.

    The 24Mhz ref to the LNB is now exact.