POTY and return loss plot

  • do1ctl Frank in the middle I have a hole for the 22mm wavegide an the I put the Venton lens on it. works in this way also with 10Ghz because the 10 GHz is not going though the endcap..:thumbup:

    Normally 10GHz should work. I used this grey water tubes since long time ago for our 10GHz ATV repeaters slot antennas as random. Never seen anay degradation. They are also UV resistent. After 15 years outside i dismounted the slot antenne from the water tube random an they looked as new. The grey tubes are perfekt..

  • Quote

    Thanks for that nice antenne - next challenge to make it water and winter restistance.

    What I currently have is a RND 455-00198 (available at Reichelt in Germany). I am unable to measure to the last dB (can someone please do these measurements?), but empirical shows no loss and it doesn't heat in the microwave.

    To install the POTY, you need a 22mm hole (10 GHz pipe), a hole for the SMA for 2.4GHz, I have not weathered those yet, but the box itself is IP65.

    And, to attach the POTY, 4 M3-screws and rings is sufficient.

    Attaching pictures for inspiration

  • Hi, I have a problem related to the original post. I have a very similar setup: I have a fairly recent POTY (i.e. with the cylindrical lens) on an Octagon Quad Green HQ in front of a 85cm offset dish. Reception using the POTY is fine and I am currently trying to tune it for transmitting. This is my patch:

    From what I've read here, the POTY can be characterised by some notch (for some around 1800MHz) from the reflector and, depending on your measurement equipment, either a double-dip slightly below and slightly above 2400MHz. or just a large dip around 2400MHz.

    Using a MiniVNA Tiny, the RL-dip was slightly too high for me in the beginning:

    After some round of gentle bending the result looks like this:

    which looks a bit better (return loss of about 19db at an SWR of about 1.3) and has some potential for optimization.

    However, my smith diagrams differ vastly from what I've seen on the BATC forum. This smith diagram corresponds to the previous RL/RP plot:

    which is similar to the results by some other people, except that the loop in the diagram is kind of blown up. When I look at the smith diagram over a larger frequency range, I get this:

    From what I see everything seems to be okay per se, but none of the other POTY measurements I've seen had multiple spiralling loops, so I may have made some mistake?

    (Btw: the loss in the last plot is slightly smaller because the patch probably bent back a little after around 30min, so I'll have to repeat the process)

  • How about marker 1? I think there is your balloon in the vincinity of Z = 1 (normalized) of the Smith Chart. Perhaps you can reduce your span to zoom in?

    Then . . . I can('t!) repeat this infinitely ... don't stare yourself blind at these measurements and resulting pictures. A few decades ago (when sex was dirty and the environment clean) having a RL of 10 dB (equals VSWR50 = 2) on the microwave bands was considered 'enough', anything more was 'nice to have'. People having worse RL's are still alive (and making contacts over QO-100)

    The 'loss' (with RL = 10 dB) is only 10% of the power, so the remaining power is 90% which, relative to RL = ∞, is 0.46 dB less (!)

    Also, related to the difference between linear and circular polarization (theoretically 3 dB), mutilation of the axial ratio will be a fraction of that.

    "Deutsche Gruendlichkeit" is not always 'the' solution for a problem ;; -)

  • PA3FYM Oh, you mean an even smaller balloon? I haven't even considered that. My guess is that it would be below the measurement noise then, but I'll have a look tomorrow.

    By the way: I am actually not too worried about the loss in general, that's why i wrote that everything looks fine per se, except for the very spirals which puzzle me. Nevertheless, as you wrote earlier in this thread: test equipment is cheaper and thus more available than ever and people may get lost in some unfocused over-optimization (I'm paraphrasing here) and I certainly agree. But I also see this as an opportunity to overcome some "street knowledge" that is traded like folklore.

    DL9SEC Ah, you're right. Of course I calibrated directly at the device. Good pointer, thanks! I'll recalibrate tomorrow!

  • Quote

    Oh, you mean an even smaller balloon? I haven't even considered that. My guess is that it would be below the measurement noise then, but I'll have a look tomorrow.

    Yes, these 'large spirals' are not uncommon, but at marker 1 there is a discontinuity which I suspect will be the 'circular polarization' balloon.

  • There is not always a loop, the inflection cusp at the marker point. Don't worry about the loops on a larger span, it looks like your calibration plane is not at the antenna but perhaps some distance away. If I had the inclination and more information I could work out how far.

    The real test is does it work? Transmit a known power, do the link budget. Is it in agreement with your measured result?

    I think it would help people a lot is AMSAT DL would come clean with the actual measured performance of the transponder and a real sample link budget. If they can't due to an NDA or do not have the technical capability, I will work with some friends to measure it as best we can.


    • Official Post

    G0MJW we do have a link budget with is in very good agreement with the observed performance. But as you feared, this has too much NDA'ed information in it and for the sake of staying clean, I will not publish any numbers myself (sorry for being paranoid).

    However, if anybody else is generating a spreadsheet which can be verified by the QO100 community, there is nothing we can do about...