LNB Wobble | A tale of two LNB's

  • I have put this in 3 different user groups as it has many applications.


    Many have noticed a frequency wobble on their 10GHhz downlink of Oscar 100.. It sounds OK but a minute later you are off frequency by a about 100Hz. Worse than that, if you listen to the CW Beacon you can detect by ear, the sliding and cyclic pitch instability.

    So what to do ?


    I assumed that as the frequency went forwards and backwards that it was not thermal or a crystal problem. Talk had been made about synthesiser chips ?


    So I took two identical Triax single LNB's 304826. I chose these because their noise figure is good and the IF gain does not fall away until below 500Mhz unlike the Octagon that at 740Mhz has already decreased significantly. The synth chip is the 2566E and a 25Mhz crystal. Sure enough they both suffered the cyclic wavering of frequency.


    I don't like the idea of GPS locking with all it's complication. So as I already had a Si Time 0.1ppm 25Mhz TCXO and grafted that in, with a 56ohm supply dropper and a 100n supply decoupler. The crystal was removed of course. The first thing to notice was the absolute frequency accuracy measured at 0.08ppm.. Was the tone of the CW better or was it "expectation bias" ? Whatever it's a perfect T9.


    So what is it like ? Perfect, absolutely pitch perfect, no waver at all and SSB stations sound way more realistic. And they stay on frequency !


    Worth the money? considering the cost of the whole station a resounding yes ! Here is the link to it https://www.digikey.co.uk/prod…X/1473-30922-1-ND/8628633.


    I suppose the question is why ? What makes it cyclic ? I really don't know, but with absolute frequency errors of up to a huge 400Khz and the awful pitch changes, it's not worth bothering with. There are other cheaper TCXO's available and YMMV.


    Regards


    Dave

    Dave Cawley | Ex G8EAO & G6ANG/T |
    (when you had to have a separate TV license !)

  • I can confirm that with ext ref the wobbling on all LNBs that i modified was gone.


    But for my beloved SW drift correction i wanted a unmodified an wobble-free LNB :)


    The old and new Octagons and the Opticum AX Black Buster are performing well.


    No luck with GM201 and HD Line Black Premium 2.

  • Hi Dave,


    Thank you for your investigation and conclusion.


    Maybe the next step is to switch over to an external placed OCXO as reference oscillator with much lower drift and better phase noise.

    It has to be synchronised for one time against a GPS reference and if kept on 24/7 it will remain within a few Hertz accurate over months. This is how I do it.


    I guess that with this approach any currently suspicious wobbling LNB will become a "pro" LNB.


    73 de Cor

  • Dave,

    0.08 ppm equals a drift of more than 800 Hertz per degree Celsius....

    So not OK for CW and digital modes and trivial for SSB.


    To me it is important to know my frequency and by using one central OCXO reference in my shack, I can have my receive and transmit frequency exactly synchronised. No separate tuning/RIT necessary.


    The LNB LO signal is simply fed thru the coaxline, less work than the TCXO modification I guess.


    But it is everyone's own choice how to do it. That is clearly visible on this forum!


    73 de Cor

  • Recently I had the same issue with one quartz oscillator at 20MHz.

    I left it powered for a few days at room temperature and monitored frequency.

    It wobbled on it's own. No pattern I could determine. At the end cheep XTAL was problem. Replaced with another and frequency was stable.

  • Dave,


    Have to apologize for not reading first the datasheet, considered it as a 'normal' TCXO. However looking at the unit-price (5 times that of my OCXO!!) there had to be a difference.


    After studying the sheet I see a drift of 2ppb per degree C, which still gives a drift of the 9750 MHz LO frequency of 195 Hz/degree.


    Taking into account long term stability, jitter and phase noise my preference stays with the SC-cut OCXO .


    Here is my system set-up:


  • Nice setup Cor, but You are using a '220 MHz. IC-375' :/ I think your are probably using a IC-475 ;) , 73 de Frank PH2M

  • Those are worst case examples and not actually as in real life. Here it is never more than 1Khz out at any temperature yet experienced. Not your accuracy I agree, but better than most will ever need. As you haven't tried one, I'm not sure how you can say your SC-cut OCXO is better ? Why not start your own thread on your SC-cut OCXO , and I'll try not to disrespect it.


    What I am reporting is real, constructive and easily achievable at moderate cost. That's why I posted that, here in my thread.


    BTW I do acknowledge your system which is clearly well designed but out of reach for many. Maybe describe that in detail in your own thread ?


    Dave

    Dave Cawley | Ex G8EAO & G6ANG/T |
    (when you had to have a separate TV license !)

    Edited once, last by G4IUG ().

    • Official Post

    Gentleman.. please keep it nice.


    Personally I have seen variations of 15 kHz and more on my unmodified LNB but these swings are very slow and do not pose and problem for me for neither SSB or CW as a 1 degC change is taking low in my unshielded setup. Other modes might be affected more.

    Some want their system connected to a central GPSDO, that is ok. I do have a couple of those (and I claim I know the internals quite well as I did some own designs as well). Still, one *can* work SSB and CW with an unmodified, unshielded LNB. Demonstrated many times. But it is true, you can do much better if you are willing to spend the time and effort (and money).

  • Agreed ! For a little while I to used an LNB "out of the box" but I could not handle the slowly changing SSB pitch !


    This a thread started by me on a specific and narrow topic where it is essential to both read "and" understand specifications, hence my annoyance of the "noise" !

    Dave Cawley | Ex G8EAO & G6ANG/T |
    (when you had to have a separate TV license !)

  • The above was written 3 months ago and the 0.1ppm still looks good. The absolute accuracy is within specification and the temperature drift specification is way better. On any day, at any time, at any temperature, a SSB signal is perfectly on frequency. And no wires, splitters, boxes, antennas etc.


    However doubts have been cast on GPS systems in general with poor phase noise.
    I have the "HP 58503B GPS (option 0004) locked frequency standard" and had to disconnect it from my HP8566B low noise analyser, because although it made it very accurate on frequency, it ruined its performance. Go figure and YMMV !

    Dave Cawley | Ex G8EAO & G6ANG/T |
    (when you had to have a separate TV license !)