Announcement

Collapse
No announcement yet.

NooElec with small antenna ---> FA Pro Stick + FA 1090 Antenna = worse results?

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #16
    Originally posted by abcd567 View Post
    A Cavity filter will sure give precise and narrow band width, but cost $100 to $150
    This cavity filter goes for €42 incl. VAT (which you won't pay if you're outside of the EU) so no, they don't all cost thee figures or substantially more than a SAW filter.

    Doubtless you can spend hundreds on them, but the same goes for nearly anything.

    Comment


    • #17
      Originally posted by abcd567 View Post
      TEST-3: Flightaware Antenna and Pro Stick/ProStick Plus with Attenuators
      Gain Setting throghout the experiment = 38.6 dB
      Adding pads (attenuators) to a system that includes amplifiers (actually two in this case, up to +70 dB or more in aggregate) is counterproductive, to say the very least.

      Signal lost in the pads can never be recovered, where injected noise is roughly proportional to the gain you use plus a bit more, therefore all you will accomplish is to decrease your SNR.

      If the problem is intermodulation noise, then use a decent, low-insertion loss bandpass filter before the first LNA rather than pads. Then set the gain to avoid overdriving anything.

      Comment


      • #18
        Thanks for all the replies!

        I do have the blue stick - connected directly to the Pi. Pi is powered by a suitable high amperage power supply. The 16" cable between the pi and the antenna is a good quality one.

        I ordered the FA 1090MHZ filter to put in front of the stick which should be here on Wed, I will update with those results. I do have quite a few Zigbee sensors around the house - 902-928mhz, perhaps those are interfering?

        Anyway, if the FA filter helps out, maybe I'll invest in a better one.

        Comment


        • #19
          Originally posted by klinquist View Post
          I ordered the FA 1090MHZ filter to put in front of the stick which should be here on Wed, I will update with those results.
          That's a bit of a pity because the bandwidth of that filter seems too wide to exclude GSM signals, which means a negative result won't necessarily prove much. I know they're more expensive, but a SAW or better still a cavity filter would probably have been a better choice. If you have the option of returning the filter and ordering a SAW filter (from China, if you must), that might be a better idea.

          On the other hand, if the BP filter in the Pro+ is after the LNA, it might help a bit to filter pre-LNA, even if the bandwidth is too wide.

          Originally posted by klinquist View Post
          I do have quite a few Zigbee sensors around the house - 902-928mhz, perhaps those are interfering?
          Hard to say. I'm unsure what ZigBee's maximum power output is, but it's probably in in the 50-100 mW range. That's tiny compared with GSM base stations, OTOH, your ZigBee transmitters are that much closer and inside your house, at which point the inverse square law comes into play.

          Comment


          • #20
            Ok, I just *also* ordered a SAW filter from Hong Kong .

            Comment


            • #21
              Originally posted by klinquist View Post
              Ok, I just *also* ordered a SAW filter from Hong Kong .
              Excellent! Can you post a link to its specs? It'll be interesting to compare the two (you should be able to post URLs now, I think).

              Based on what abcd567 said, the FA BP filter is an LC network, and LC networks suck, both in terms of insertion loss and linearity and probably also in terms of phase/group velocity distortion. Not saying SAW or cavity filters are perfect, but both are probably way better than LC networks.

              If the Pro+'s BP filter is also an LC network that might explain why the filter is after the LNA — after all, one of the two principal reasons for using a bandpass filter is to avoid intermodulation in amplifiers — but from this image, I would guess that the Pro+ uses a SAW filter.

              Comment


              • #22
                heh apparently I can't post links in the forum until I have at least 4 posts... so it's ebay item ID 272374693052

                Note I *also* ordered the regular Pro stick w/o the filter.... for even more comparisons.

                Already got too much money on this project .

                Comment


                • #23
                  Originally posted by Strix technica View Post
                  Excellent! Can you post a link to its specs? It'll be interesting to compare the two (you should be able to post URLs now, I think).
                  From ebay page http://www.ebay.com/itm/272374693052

                  Features:
                  The filter bandwidth 8MHz, Cutoff characteristics steep. 1090MHz performance for the ADS-B receiver Optimization,Can effectively filter out-of-band signals,Reduce noise,Increase the dynamic range

                  Modell: FBP-1090s
                  Impedanz: 50 Ohm
                  Mittenfrequenz : 1090 MHz
                  Passband : 1086 ~ 1094 MHz
                  In-Band-Einfügungsverlust : <3dB
                  Maximale Lastleistung : 10 mW (10 dBm)
                  Schnittstelle: SMA-M, SMA-F


                  If the Pro+'s BP filter is also an LC network that might explain why the filter is after the LNA — after all, one of the two principal reasons for using a bandpass filter is to avoid intermodulation in amplifiers — but from this image, I would guess that the Pro+ uses a SAW filter.
                  YES, the ProStick+ (blue) uses a SAW filter, but at wrong location. The SAW filter chip is located between pre-amp and tuner. If it was located between the antenna connector and pre-amp, an external filter will not be required with ProStick Plus (Blue one).
                  Last edited by abcd567; 2017-07-17, 18:10.

                  Comment


                  • #24
                    Originally posted by klinquist View Post
                    Ok, I just *also* ordered a SAW filter from Hong Kong .
                    I have found this one couple of months ago while browsing ebay. I was very much tempted to order it, but it's price prevented me. At US $28.99, it is almost double the price of Flightaware filter (US $15.50).

                    Comment


                    • #25
                      Originally posted by klinquist View Post
                      heh apparently I can't post links in the forum until I have at least 4 posts...
                      D'oh! Thought it was 3, but they may have changed that.

                      Originally posted by klinquist View Post
                      Note I *also* ordered the regular Pro stick w/o the filter.... for even more comparisons.
                      I look forward to hearing your results! Money is tight right now, but I'm awfully tempted to order one of those SAW filters because I'm nearly certain I've got intermod problems.

                      Originally posted by abcd567 View Post
                      The filter bandwidth 8MHz, Cutoff characteristics steep.
                      Thanks, abcd! Looks like a pretty neat piece of kit, pretty decent insertion loss for a passive filter (< 3dB). If'n I had the money, I'd get one of them and the cavity filter I linked the other day and compare them.

                      Originally posted by abcd567 View Post
                      YES, the ProStick+ (blue) uses a SAW filter, but at wrong location. The SAW filter chip is located between pre-amp and tuner.
                      Alas... What were they thinking, aside from "ooh, there's a largely empty spot on the PCB where we can glom on a SAW filter for extra mark-up!"?

                      Originally posted by abcd567 View Post
                      I have found this one couple of months ago while browsing ebay. I was very much tempted to order it, but it's price prevented me. At US $28.99, it is almost double the price of Flightaware filter (US $15.50).
                      You get what you pay for. The FA stand-alone filter, as you pointed out upthread, is a passive LC filter, ie crap.

                      Comment


                      • #26
                        Tonight I installed the "regular" orange FA stick and the FA 1090mhz filter....aaaaaand similar results to the blue stick. Still nowhere near as good as the little antenna with the generic stick! I'm thinking my next move might be to order an adapter to hook the generic stick to the FA antenna to see if the issue is with the antenna or not.

                        Here's the setup (note the "better" 1090mhz filter has not yet arrived)

                        IMG_9636.jpg
                        Last edited by klinquist; 2017-07-20, 03:59.

                        Comment


                        • #27
                          The test I suggested earlier is to try the FlightAware Pro Sticks with the antenna that cam- e with the generic stick and then the generic stick with the FA antenna and cable.

                          What spec cable are you using with the FA external antenna? Did it come with pre-installed connectors?
                          Mike


                          www.radarspotting.com

                          Radarspotting since 2005

                          Comment


                          • #28
                            Yeah, I think trying the old antenna with the FA stick and visa versa is definitely worth doing, as is giving some thought to the coax feeder. Golden rule of experimentation: change only one thing at a time.

                            Golden rule of debugging something: start with what once worked, and reintroduce elements until it breaks, don't try to go from what's broken to what works because if there's more than one thing wrong, a problem fixed can remain masked by others and you won't even know that you did, in fact, fix a problem.

                            Unless your coax run is very short, then consider that a third variable. Remember that different coax has different amounts of loss per metre, but I'd be surprised if the supplied antenna was fed with anything other than RG58 (or maybe RG59) which is about the worst coax commonly found. But even then, not all RG58 is created equal either. Some of it's rubbish, some of it's okay (depending on application).

                            In theory, impedance mismatches will make things worse even for rx-only, but there isn't much you can do about that except not make it worse. At least in the generic SDR case, the receiver is designed for a 75Ω feed, yet the connectors used are typically 50Ω. (In practice, that impedance mismatch loss is only about -1 dB, but you don't want to switch between impedances more often than you have to.) In the FA case, the LNA might actually be 50Ω in which case mismatch loss hardly matters because the signal will be plenty strong enough by the time it reaches the 75Ω R820T2 RF mixer, and low-loss 50Ω RG-8 to a 50Ω antenna would be ideal.

                            Your antenna has a characteristic impedance also. Whips range between 36–50Ω, dipoles are about 72Ω. In practice, dielectric loss is going to be more of a problem that mismatch loss, but it's as well to bear it in mind. Just don't constantly switch back and forth between impedances any more than you have to.

                            Originally posted by klinquist
                            Tonight I installed the "regular" orange FA stick and the FA 1090mhz filter....aaaaaand similar results to the blue stick.
                            Which is no surprise if the problem is intermod noise. As I said, the FA BP filter's bandwidth is too wide to exclude all of GSM, especially in North America and the in-line filter is a crappy LC network rather than a SAW or cavity filter.

                            If you're running dump1090-mutability or equivalent, have a look at the range of signal strengths it reports. My Munin plugins are good for this, but most of those only work with -mutability. If your peak signal stays stuck at > -5 dBFS all day and night, and your mean signal is around -10 dBFS, your gain is probably too high.

                            Comment


                            • #29
                              Try Gain 30 dB.
                              Users having FA Antenna and Pro Stick /Pro Stick Plus, generally use gain between 25 dB and 35 dB.
                              I use 28 dB with one receiver, and 30 dB with other one.

                              Python Script to optimize gain is a help, but dont give any concrete value. Finally it is trial and error to find best gain.

                              Code:
                              pi@raspberrypi:~ $ sudo nano /etc/default/dump1090-mutability
                              Above command will open file in nano editor
                              Press Ctrl+W (i.e. Find), then type GAIN= and press Enter Key. The cursor will jump to GAIN=
                              Change the gain value (shown in red)
                              Save file (Ctrl+o), then close the file and exit editor (Ctrl+x)

                              .....................
                              .....................
                              .....................
                              # Receiver options
                              #

                              # RTLSDR device index or serial number to use
                              # If set to "none", dump1090 will be started in --net-only mode
                              DEVICE=""

                              # RTLSDR gain in dB.
                              # If set to "max" (the default) the maximum supported gain is used.
                              # If set to "agc", the tuner AGC is used to set the gain.
                              GAIN=30

                              # RTLSDR frequency correction in PPM
                              PPM="0"
                              .....................
                              .....................
                              .....................

                              Now restart dump1090 for new value of gain to take affect.
                              Code:
                              sudo systemctl restart dump1090-mutability
                              Now check the status. It will show the gain value in use. This value will be retained on reboots and restarts, unless you change it.
                              Code:
                              pi@raspberrypi:~ $ sudo systemctl status dump1090-mutability -l
                              Last edited by abcd567; 2017-07-20, 15:30.

                              Comment


                              • #30
                                Originally posted by Anmer View Post
                                What spec cable are you using with the FA external antenna? Did it come with pre-installed connectors?
                                I'm using this one:



                                So last night I connected the yellow FA stick directly to the antenna with NO filter and it seems that I'm getting significantly better results - even saw one aircraft 160nm away. This was using dump1090 with completely default settings.
                                I'm running the gain python test now to figure out the ideal gain and see how much better I can do.... but it seems that any of the filters I've tried so far (the one built into the blue stick and the external one) are making things worse. Odd, I know!

                                Comment

                                Working...
                                X