Posts by PA3FYM

  • Es'hail-2 transverter PA3FYM-approach

    VU2UUU OK Kaustav, my idea was to use 432 MHz as IF frequency for both RX and TX. Of course you can use 28 or 50 MHz for TX-IF but I recommend to do that in an additional step (to avoid the LO signal and mirror image bleed through / will not be supressed enough).


    If you've only 28 or 50 MHz (SSB) you can think of the following (50 MHz example given):


    2400 - 50 = 2350 / 2 = 1175 MHz LO (from cheap ADF4351 board)


    So, the first conversion is 50 + 1175 = 1225 MHz , filter this (search for SAW filters for this frequency...) and feed this into another mixer with the same LO (1175 MHz) --> 1225 + 1175 = 2400 MHz.


    To clean up the resulting signal use cheap 'WiFi filters', designed for 2400 - 2500 MHz (e.g. I use CER0347B filters, but there are also from TDK (DEA252400BT-2030A1) but these are very small, only 2 x 2.5 mm !).


    As mixer I use a HMC213 (Ebay). ADF4351 signal is amplified with a SPF5189z board (also Ebay) and fed into the mixer LO-port. If you want to make the 'double conversion' approach you (probably) need another SPF5189z to feed the other HMC213 mixer. These cheap ADF4351 boards have two outputs, so that makes life easy!


    In the picture below I drew two SPF5189z's. Perhaps one is enough to feed two mixers. The ADF4351 boards I got from Ebay do not deliver +5 dBm (3.2 mW), but around 1.5 mW @1967.5 MHz. Using a SPF5189z board I got ~14 dBm, sufficient to LO the HMC213. But ... perhaps the SPF5189z board has more gain @1175 MHz, so one SPF5189z may be sufficient.


  • UPLINK - PATCH VS HELIX - what's better

    Depends on how the patch is designed. Standard broadcast offset dishes have f/D of 0.6 (otherwise each dish needs his own LNB ; -) , which equals to 90° -10 dB beam width (width of the offset dish taken as 'size'). Patch in the link has -10 dB = ca. 105°. So it over illuminates a bit, but that does not matter as it's used to transmit only.


    To illuminate a dish with a helix properly a rule of thumb is that you need one (1) winding per f/D-unit. So f/D = 0.6 needs 6 windings. With smaller f/D's less windings are needed, however, with less windings the axial ratio becomes worse.


    I made a dual band feed with a helix, it works. However, from a mechanical (stiffness, size) perspective the patch is my favourite.

  • Octagon Oslo PLL LNB with external Reference

    DD0KP  DL5MGD  i2NDT


    OK, this morning the postman arrived bringing three samples of 'my' chip PL902190USY. It works .. in a sense that it generates 10 MHz from the Ublox 1 MHz clock. Since I made the phase detector with XOR's running at 1 MHz I used the 2nd part of the HC390 to divide the output of the jitter blocker by 10.


    The PLL locks and it seems that the jitter is gone. However, the phase noise is very high. When I listen to the CW-beacon it sounds like Aurora (!) So, it's difficult to detect whether the jitter is completely gone. Also, peculiarly, I have to tune ca 900 Hz higher ... This means that either the output of the Ublox is not exactly 1 MHz, or the jitter blocker generates a 10 MHz 'average signal' from the jittery Ublox 1 MHz.


    Now I have to find where this phase noise issue originates from. Is it intrinsic (i.e. the jitter blocker is not able to perfectly remove all jitter), the loop filter or something else?

  • UPLINK - PATCH VS HELIX - what's better

    There are many roads to Rome. It's important that your uplink antenna illuminates the dish properly and when you use circular polarization it's LHCP (in front of a dish).


    I made helices, and they work fine, but from a mechanical perspective and simplicity I prefer the patch. For a variant of the K3TZ patch look here.

  • Dual Feed For Uplink / Downlink

    DL5MGD Relative to the patch dimensions the reflector plate is not that critical. So 104mm diameter (I reckon it fits inside your DN110 pipe then?) will do fine. Perhaps you loose a milliBell gain (i.e. the radiation pattern) ; -)