Loading ...
Sorry, an error occurred while loading the content.

Image rejection

Expand Messages
  • kc0vkn
    Hey group, I ve got an older 6.1-Lite that I ve been fussing with on 30M with CW Skimmer. I d had it up a few hours when someone notified me my spots were off.
    Message 1 of 18 , May 11, 2008
    • 0 Attachment
      Hey group,
      I've got an older 6.1-Lite that I've been fussing with on 30M with
      CW Skimmer.

      I'd had it up a few hours when someone notified me my spots were
      off. I started looking more closely (and I hadn't noticed it in
      Rocky) that I was seeing mirror signals on either side of the center...

      I've just now had time to investigate further and:

      On the Audio side, I have signal on both L/R-T/R. If I go to the
      center of the board and look for the square waves, I see them at about
      90 out of phase, but they are rather ugly looking square waves.

      Just jumping back and forth in Skimmer, I find that the "correct"
      signal is a bit louder than the image, but, both are enough for
      skimmer to pickup.

      What is the correct Q/I relationship< Q=Right, I=Left, or the other
      way around?

      I've got T/R correctly connected going out to the PC, but: Skimmer
      settings are are Q=Left I=Right in order to see things where I think
      they should be (eg: stronger sig on the low side of the center; as
      confirmed with my big rig).

      Anyone have any ideas where I should look?

      73,

      Joe, KC0VKN
    • Bill Dumke
      Use the Q/I setting in CW Skimmer. It sounds like CW Skimmer is not automatically calibrating correctly. Mine took a few hours on the air and then it worked
      Message 2 of 18 , May 11, 2008
      • 0 Attachment
        Use the Q/I setting in CW Skimmer. It sounds like CW Skimmer is not
        automatically calibrating correctly. Mine took a few hours on the air
        and then it worked fine as far as the IQ calibration went. But like you
        say, it could be due to your 6.1-Lite having something wrong with it.

        Bill WA9PWR

        kc0vkn wrote:
        > Hey group,
        > I've got an older 6.1-Lite that I've been fussing with on 30M with
        > CW Skimmer.
        >
        > I'd had it up a few hours when someone notified me my spots were
        > off. I started looking more closely (and I hadn't noticed it in
        > Rocky) that I was seeing mirror signals on either side of the center...
        >
        > I've just now had time to investigate further and:
        >
        > On the Audio side, I have signal on both L/R-T/R. If I go to the
        > center of the board and look for the square waves, I see them at about
        > 90 out of phase, but they are rather ugly looking square waves.
        >
        > Just jumping back and forth in Skimmer, I find that the "correct"
        > signal is a bit louder than the image, but, both are enough for
        > skimmer to pickup.
        >
        > What is the correct Q/I relationship< Q=Right, I=Left, or the other
        > way around?
        >
        > I've got T/R correctly connected going out to the PC, but: Skimmer
        > settings are are Q=Left I=Right in order to see things where I think
        > they should be (eg: stronger sig on the low side of the center; as
        > confirmed with my big rig).
        >
        > Anyone have any ideas where I should look?
        >
        > 73,
        >
        > Joe, KC0VKN
        >
        >
        > ------------------------------------
        >
        > Yahoo! Groups Links
        >
        >
        >
        >
        >
        >
        >
        >
      • Terry
        ... center... ... the ... about ... other ... Skimmer ... Hello Joe. I typically plug the connector in such that the red wire goes to the Q, and the white to
        Message 3 of 18 , May 11, 2008
        • 0 Attachment
          --- In softrock40@yahoogroups.com, "kc0vkn" <kc0vkn@...> wrote:
          >
          > Hey group,
          > I've got an older 6.1-Lite that I've been fussing with on 30M with
          > CW Skimmer.
          >
          > I'd had it up a few hours when someone notified me my spots were
          > off. I started looking more closely (and I hadn't noticed it in
          > Rocky) that I was seeing mirror signals on either side of the
          center...
          >
          > I've just now had time to investigate further and:
          >
          > On the Audio side, I have signal on both L/R-T/R. If I go to
          the
          > center of the board and look for the square waves, I see them at
          about
          > 90 out of phase, but they are rather ugly looking square waves.
          >
          > Just jumping back and forth in Skimmer, I find that the "correct"
          > signal is a bit louder than the image, but, both are enough for
          > skimmer to pickup.
          >
          > What is the correct Q/I relationship< Q=Right, I=Left, or the
          other
          > way around?
          >
          > I've got T/R correctly connected going out to the PC, but:
          Skimmer
          > settings are are Q=Left I=Right in order to see things where I think
          > they should be (eg: stronger sig on the low side of the center; as
          > confirmed with my big rig).
          >
          > Anyone have any ideas where I should look?
          >
          > 73,
          >
          > Joe, KC0VKN
          >

          Hello Joe.
          I typically plug the connector in such that the red wire goes to the
          Q, and the white to I. But, I doubt that a reversal is your
          problem. If the only problem was that the two wires are reversed,
          you would still get virtually no images, but the spectrum would be
          backwards. This means that as you tune a signal generator up in freq
          from the bnottom of SR coverage, you would see the signal start at
          the TOP of the display and work otself down, instead of starting at
          the bottom and going up.

          Images can be caused by a whole host of issues. What type of sound
          card are you using? Not a laptop with the internal sound card I hope.

          Several Sound Blaster type sound cards typically need an I/Q delay of
          at lease one sample.

          Make sure your sound card and software is in line-in mode, not mic
          in. Mic in is typically mono, and you will not get image rejection.

          Make sure the stereo balance on the Windows mixer is centered.

          Make sure your audio cable is actually sending stereo (I/Q) into the
          sound card. I once had a radio shack cable that was poorly made, and
          only one channel got to the sound card.

          There are other potential gotchas, depending on the sound card used.

          I'm also a little concerned about you seing a square wave on the
          audio outputs, they should be sine waves. With a fairly strong
          signal fed into the SR Lite, you should see two fairly clean sine
          waves coming out the audio, with one lagging the other by 90
          degrees. The frequency of those sine waves should be equal to the
          center frequency minus the RF carrier (or the other way around). For
          example with a center freq of 7.055MHz, and an RF freq of 7.057MHz,
          you should see an approx. 2,000 Hz audio pair, with one leading the
          other by 90 degrees.

          Good luck.
          Terry
          WB4JFI
        • kc0vkn
          I m pasting this on behalf of Bob, K6XX, who was kind enough to send me a personal reply after Yahoo gave him guff when he tried to post. Now it s in the
          Message 4 of 18 , May 12, 2008
          • 0 Attachment
            I'm pasting this on behalf of Bob, K6XX, who was kind enough to send me a personal reply
            after Yahoo gave him guff when he tried to post. Now it's in the archives.

            Thanks everyone else for replies as well. I've read them but, am now busily getting my
            work done like a good beaver so I can leave for Dayton in 2 days. See you at FDIM!

            I'm using some random SB card, but, before I leave for Dayton, I'll get a Delta44 ordered
            and maybe it'll arrive by the time I return.

            I suspect I had this problems 2 years ago when I built the thing, I had this problem, but, I
            was a different ham then, and I also was just excited to have the thing hearing signals! I
            suspect that I didn't notice it with Rocky because I wasn't looking for it.

            Joe, KC0VKN

            ----------------------
            Excellent questions, Joe.

            Hopefully a real expert will tell us both how this is best done. But
            until
            then, here's the procedure I use. That's USE in the present tense,
            since I'm
            doing this task over and over each time something changes here. That
            measly
            $10 investment in the SR hardware has incurred several hundred dollars
            of
            expense here buying a new, more powerful PC and several increasingly
            capable
            (but expensive) sound cards!

            Instead of running Skimmer, use Rocky. It has the same type of
            self-calibration routine, but its wideband spectrum analyzer display
            helps
            you see what is going on faster than the waterfall. Once Rocky is
            ready,
            we'll get back to Skimmer.

            1. Connect the SR to your sound card.

            2. Get a signal generator. I've been using an Elecraft KX1 into a
            Ethernet
            resistive termination so I can place it close to the SR antenna input.
            But a
            regular dummy load and transmitter has enough leakage to do the job. If
            you
            use the Ethernet terminator, use low power of course. Less than a half
            watt.
            (But ignore the 75 ohm value mismatch--its close enough to not hurt the
            transmitter). My transmitter system works better than my signal
            generator
            because its output waveform is much, much cleaner. Remember, the
            transmitter
            is not connected to the SR antenna input. The dummy load is merely
            placed
            alongside the SR board.

            3. Transmit a stream of dashes near the bottom of the SR tuning range.
            Do
            NOT tune it near midrange!

            4. Vary the transmit frequency. If you QSY up, does the Rocky display
            show
            the dashes moving up? If so, your I/Q is correct. If the display shows
            the
            dashes moving down, the IQ is backward. Change it in VIEW:SETTINGS.

            There is a third situation: where you see a strong image on both sides
            of
            center frequency and one moves up and the other moves down. This means
            your
            balance is way off and the SR hardware needs attention. For example, is
            one
            of the audio channels disconnected? This fortunately does not appear to
            be
            your situation.

            After this point, the SR is ready for autocalibration in either Rocky
            or
            CWSkimmer. However, I'm not patient enough to await the dozens of
            strong
            signals distributed all across the band that are required for the
            autocal to
            do its job by itself. I "accelerate" the process a bit...

            5. Move the dummy load ("probe") or increasing the signal output from
            your
            generator until you have >40dB of signal indication on Rocky. In fact,
            set
            it to 50dB or 60dB. Open the RX IQ Balance window (Tools). Sweep the
            signal
            across the band slowly. You should see red and blue dots appear and a
            approximation line or curve connect their averages. I've found a stream
            of
            dashes works better than a solid carrier, but I'm not certain about
            this
            observation.

            6. If your Phase samples are within a couple of degrees and your gain
            is
            rather flat or at least gently sloping, all the way across the band,
            your
            receiver is working fine. If the phase shows +/- 20 degrees of
            variation,
            and/or the gain or phase curve is piecewise continuous (broken into
            separate, unconnected curves), you have a hardware problem. Do not
            worry
            about a big separation at midband--this is normal. Nothing seems to
            work
            right around the SR's center frequency. That's why we started this
            procedure
            near a band edge rather than near midband.

            7. Now that Rocky is working and calibrated, close it. Open CW Skimmer
            and
            re-sweep the band using the same setup and signal levels. Although the
            same
            technique is employed with both Rocky and Skimmer, their Auto I/Q
            balance
            data is incompatible and you must re-do it for each program.
            Fortunately it
            only takes a few minutes with a local signal generator instead of the
            hours
            awaiting random, strong on-air signals.

            8. Also, this procedure must be re-done each time you change sampling
            rate
            or sound cards. That means if you have calibrated your SR and soundcard
            at
            48kHz sampling, you must repeat the process when you switch to 96kHz.
            When I
            added a preamp (to eliminate the oscillator backfeed issue), I also saw
            some
            change in the Auto I/Q balance.

            I hope this has been helpful. I developed this procedure after fussing
            with
            the autocal for many hours/days. Hopefully someone who really "knows"
            how it
            should be performed will speak up and let us all know!

            73 de Bob, K6XX
            www.k6xx.com
          • Mike Young
            Yup, on all that Bob had to say, except I found that Skimmer and Rocky both share the same calibration data. Blowing it away for Rocky also destroys the set
            Message 5 of 18 , May 12, 2008
            • 0 Attachment
              Yup, on all that Bob had to say, except I found that Skimmer and Rocky both
              share the same calibration data. Blowing it away for Rocky also destroys the
              set that CWSkimmer collected. This didn't seem a great surprise when it
              happened. Skillmer wasn't even running when I did that. (I had to reset its
              calibration because it was contaminated by the 20m Lite I had hooked up
              briefly. Now, I use WinRad on 20m during the day, and Rocky and the 40/30 in
              the evening. It might sounds hokey, but what exactly do you want for your
              $12 investment? :)

              The good news was, 40m was bouncing and hopping this evening to where you
              almost couldn't get a QSO in edgewise in the lower 50 kHz. One of those
              rare, enchanged evenings.


              ----- Original Message -----
              From: "kc0vkn" <kc0vkn@...>
              To: <softrock40@yahoogroups.com>
              Sent: Monday, May 12, 2008 7:36 AM
              Subject: [softrock40] Re: Image rejection


              > I'm pasting this on behalf of Bob, K6XX, who was kind enough to send me a
              > personal reply
              > after Yahoo gave him guff when he tried to post. Now it's in the
              > archives.
            • nels_58701
              I have the Ver 8.0 with the electronic switching BPF and it works great but I have no image rejection. I also have the Ver 7.0 and that works just fine. Any
              Message 6 of 18 , Mar 6, 2009
              • 0 Attachment
                I have the Ver 8.0 with the electronic switching BPF and it works great but I have no image rejection.
                I also have the Ver 7.0 and that works just fine.
                Any ideas where to start looking?
                I used all the same cabling to the sound card with the 7.0 so I know
                that they are ok.
                Same with Rocky and SDR-16

                Thanks
                Nels W0TUP
                nels@...
              • g8voip
                Hi Nels, If the v7.0 is working with good image rejection, then as you say all the cables etc must be ok. I cannot remember if the I and Q outputs are the same
                Message 7 of 18 , Mar 6, 2009
                • 0 Attachment
                  Hi Nels,

                  If the v7.0 is working with good image rejection, then as you say all the cables etc must be ok.

                  I cannot remember if the I and Q outputs are the same way around on the v7.0 and v8.0? Try reversing the output leads.

                  If that does not resolve the problem, then you need to start tracing the signals back from the I and Q outputs, suspect one must be missing somewhere or a problem even as far back as the mixer, clock signals or input transformer.

                  You really need to have access to an oscilloscope to trace the signals through.

                  Good Luck!

                  Bob G8VOI



                  --- In softrock40@yahoogroups.com, "nels_58701" <anelson@...> wrote:
                  >
                  > I have the Ver 8.0 with the electronic switching BPF and it works great but I have no image rejection.
                  > I also have the Ver 7.0 and that works just fine.
                  > Any ideas where to start looking?
                  > I used all the same cabling to the sound card with the 7.0 so I know
                  > that they are ok.
                  > Same with Rocky and SDR-16
                  >
                  > Thanks
                  > Nels W0TUP
                  > nels@...
                  >
                • Pete
                  Hi Nels With information provided:- a/ Previous Soft rock working OK on the same computer and soundcard. b/ Software application running OK with a previous
                  Message 8 of 18 , Mar 8, 2009
                  • 0 Attachment
                    Hi Nels

                    With information provided:-

                    a/ Previous Soft rock working OK on the same computer and soundcard.
                    b/ Software application running OK with a previous soft rock.
                    c/ The cables and connectors are the same for the previous soft rock. I.e. there are no whiskers connecting the two inputs together or a disconnected input for I or O.

                    Then the issue is with your new build.

                    Suggestions are:-

                    1/ First of all check the input transformer connections are correct. The phasing of the input signal to the QSD chip is vital. These connections must be correct.
                    2/ Check your output I/O. To do this, disconnect one output at a time at the Soft rock card and run your soft rock.
                    If the image condition exists and signals are approx the same amplitude with either one of the outputs disconnected then you know the two op amps are working and need no attention. I.e. no soldering bridges or incorrect value components. This indicates the problem originates elsewhere.
                    If there is a total loss of signal when one or other of the outputs are disconnected then your problem is in that part of the circuit. Check carefully all components are in place and no dry joints or bridges. Bridges are an issue around the pins of the SMT chips, a big help here is to use a solder wick and some free flux to remake these connections.
                    3/ Assuming all is well at 1/ and 2/ check the QSD for solder bridges by using the method suggested in 2/.
                    4/ And similarly repeat 3/ for the 74HC74.

                    In my humble experience all the issues with images have been for the above reasons. This is an incredibly Sure-Fire kit so checking and rechecking the build quality will usually result in a solution.

                    Good luck and 73 Petefmt

                    --- In softrock40@yahoogroups.com, "nels_58701" <anelson@...> wrote:
                    >
                    > I have the Ver 8.0 with the electronic switching BPF and it works great but I have no image rejection.
                    > I also have the Ver 7.0 and that works just fine.
                    > Any ideas where to start looking?
                    > I used all the same cabling to the sound card with the 7.0 so I know
                    > that they are ok.
                    > Same with Rocky and SDR-16
                    >
                    > Thanks
                    > Nels W0TUP
                    > nels@...
                    >
                  • wa1vta01452
                    Bob, I wonder how much image rejection you have with V6.3 using the Delta 44. I use PSDR V 1.12.20 and don t see much more than 20 db. Are you using shielded
                    Message 9 of 18 , Feb 28, 2010
                    • 0 Attachment
                      Bob,
                      I wonder how much image rejection you have with V6.3 using the Delta 44. I use PSDR V 1.12.20 and don't see much more than 20 db. Are you using shielded audio cable between the sound card and the RXTX? I assume you have not tied the grounds on both ends. In any case I am not sure if this would affect image rejection or only noise pickup.

                      Tom
                    • Alan
                      ... From: wa1vta01452 Subject: [softrock40] Image rejection What particular card should not make any difference. But do NOT use the Monitor/Mixer of the
                      Message 10 of 18 , Mar 1, 2010
                      • 0 Attachment
                        ----- Original Message -----
                        From: "wa1vta01452"
                        Subject: [softrock40] Image rejection



                        What particular card should not make any difference. But do NOT use the
                        "Monitor/Mixer" of the D44.
                        I do have D44s and on RX I get something like -30db BEFORE any software
                        correction, then -50dB is easily achieved with any manual software
                        adjustments. Often much more.

                        Similarly on TX, about 30dB due to hardware, much better after adjusting.

                        All audio cables should be shielded, make sure connections are good, 1/8
                        jacks can be suspect.
                        Ground loops will not affect image rejection, just create a lot of "hash"
                        spreading from the centre.

                        73 Alan G4ZFQ
                      • tom
                        Alan, what do you mean by not using the monitor/mixer? Tom ... From: Alan To: softrock40@yahoogroups.com Sent: Monday, March 01, 2010 7:20 AM Subject: Re:
                        Message 11 of 18 , Mar 1, 2010
                        • 0 Attachment
                          Alan,
                           
                          what do you mean by not using the monitor/mixer?
                           
                          Tom
                           
                           
                          ----- Original Message -----
                          From: Alan
                          Sent: Monday, March 01, 2010 7:20 AM
                          Subject: Re: [softrock40] Image rejection

                           


                          ----- Original Message -----
                          From: "wa1vta01452"
                          Subject: [softrock40] Image rejection

                          What particular card should not make any difference. But do NOT use the
                          "Monitor/Mixer" of the D44.
                          I do have D44s and on RX I get something like -30db BEFORE any software
                          correction, then -50dB is easily achieved with any manual software
                          adjustments. Often much more.

                          Similarly on TX, about 30dB due to hardware, much better after adjusting.

                          All audio cables should be shielded, make sure connections are good, 1/8
                          jacks can be suspect.
                          Ground loops will not affect image rejection, just create a lot of "hash"
                          spreading from the centre.

                          73 Alan G4ZFQ

                        • Alan
                          ... From: tom Subject: Re: [softrock40] Image rejection ... I m still using the old control panel so can t describe the new. The mixer controls should not
                          Message 12 of 18 , Mar 1, 2010
                          • 0 Attachment
                            ----- Original Message -----
                            From: "tom"
                            Subject: Re: [softrock40] Image rejection



                            >what do you mean by not using the monitor/mixer?

                            I'm still using the old control panel so can't describe the new.
                            The mixer controls should not work. If they do you may not get any TX image
                            rejection.
                            If they do there is a way to disable them, a drop-down or box..

                            73 Alan G4ZFQ
                          • Ben Hall
                            Good evening all, In the mail today arrived the final components to complete my electronically switched BPF for the my Lite+USB XTtall V9.0 receiver. While I
                            Message 13 of 18 , Mar 1, 2010
                            • 0 Attachment
                              Good evening all,

                              In the mail today arrived the final components to complete my
                              electronically switched BPF for the my Lite+USB XTtall V9.0 receiver.

                              While I have no problems winding toroid inductors - meaning that I
                              possess normal vision (albeit with fairly thick glasses) and normal
                              manually dexterity (ok, I have moments of shakiness), I started
                              wondering if anyone had substituted SMD inductors for the toroid
                              inductors on the BPF.

                              Dave WB6DHW uses SMD inductors on his successful WB6DHW
                              Band Pass Filter Board. I've built one and it seems to work just fine.
                              So I know that SMD inductors can be successfully used in RF bandpass
                              filters.

                              (Side Note: I seem to recall a section in the builders notes that state
                              that toroids may have lower loss than the SMD parts. The AADE filter
                              design software confirms this when I enter Q's of 25 from the data
                              sheets of the WB6DHW design parts into the SoftRock bandpass
                              configuration compared to an assumed infinite Q for a toroid.)

                              So here's my crazy idea - has anyone built up an electrically switched
                              SoftRock BPF substituting SMD inductors for the the wound toroids?

                              I may just give this a whirl!

                              thanks much,
                              ben
                              --
                              Ben Hall, kd5byb@... - ALWAYS OUTNUMBERED, NEVER OUTGUNNED.
                            • Jack Smith
                              In addition to the Q loss (typical SMD RF choke type inductor in the 25-50 range, powdered iron cores in the 150-300 range depending on size and other
                              Message 14 of 18 , Mar 1, 2010
                              • 0 Attachment
                                In addition to the Q loss (typical SMD RF choke type inductor in the 25-50 range, powdered iron cores in the 150-300 range depending on size and other factors), many of SMD parts are wound on ferrite cores. If you are concerned with wringing the best intermodulation performance from a filter, it's best to avoid ferrite core inductors. Even powdered iron can be problems at some levels of performance, but powdered iron is clearly to be preferred over ferrite core parts.

                                Jack K8ZOA


                                On 3/1/2010 6:28 PM, Ben Hall wrote:
                                 

                                Good evening all,

                                In the mail today arrived the final components to complete my
                                electronically switched BPF for the my Lite+USB XTtall V9.0 receiver.

                                While I have no problems winding toroid inductors - meaning that I
                                possess normal vision (albeit with fairly thick glasses) and normal
                                manually dexterity (ok, I have moments of shakiness), I started
                                wondering if anyone had substituted SMD inductors for the toroid
                                inductors on the BPF.

                                Dave WB6DHW uses SMD inductors on his successful WB6DHW
                                Band Pass Filter Board. I've built one and it seems to work just fine.
                                So I know that SMD inductors can be successfully used in RF bandpass
                                filters.

                                (Side Note: I seem to recall a section in the builders notes that state
                                that toroids may have lower loss than the SMD parts. The AADE filter
                                design software confirms this when I enter Q's of 25 from the data
                                sheets of the WB6DHW design parts into the SoftRock bandpass
                                configuration compared to an assumed infinite Q for a toroid.)

                                So here's my crazy idea - has anyone built up an electrically switched
                                SoftRock BPF substituting SMD inductors for the the wound toroids?

                                I may just give this a whirl!

                                thanks much,
                                ben
                                --
                                Ben Hall, kd5byb@gmail. com - ALWAYS OUTNUMBERED, NEVER OUTGUNNED.

                              • jorschei
                                Hi Ben, SMD coils are usefull even when the Q factor is lower than Amidon torriods. For SDR we need no narrow passbands. I use COILCRAFT surface mount
                                Message 15 of 18 , Mar 2, 2010
                                • 0 Attachment
                                  Hi Ben,

                                  SMD coils are usefull even when the Q factor is lower than Amidon torriods. For SDR we need no narrow passbands. I use COILCRAFT surface mount inductors High Q series 1812CS (4532) for the KTH-SDR wideband modulair receiver and it work well.See the KIT doc on Rob his site.
                                  http://home.kpn.nl/rw.engberts/sdr_kth_buyers.htm
                                  The have ceramic forms and no ferrite. Filter with 1.5 dB loss can be accepted.

                                  73' Joris PE1KTH



                                  > So I know that SMD inductors can be successfully used in RF

                                  --- In softrock40@yahoogroups.com, Ben Hall <kd5byb@...> wrote:
                                  >
                                  > Good evening all,
                                  >
                                  > In the mail today arrived the final components to complete my
                                  > electronically switched BPF for the my Lite+USB XTtall V9.0 receiver.
                                  >
                                  > While I have no problems winding toroid inductors - meaning that I
                                  > possess normal vision (albeit with fairly thick glasses) and normal
                                  > manually dexterity (ok, I have moments of shakiness), I started
                                  > wondering if anyone had substituted SMD inductors for the toroid
                                  > inductors on the BPF.
                                  >
                                  > Dave WB6DHW uses SMD inductors on his successful WB6DHW
                                  > Band Pass Filter Board. I've built one and it seems to work just fine.


                                  bandpass
                                  > filters.
                                  >
                                  > (Side Note: I seem to recall a section in the builders notes that state
                                  > that toroids may have lower loss than the SMD parts. The AADE filter
                                  > design software confirms this when I enter Q's of 25 from the data
                                  > sheets of the WB6DHW design parts into the SoftRock bandpass
                                  > configuration compared to an assumed infinite Q for a toroid.)
                                  >
                                  > So here's my crazy idea - has anyone built up an electrically switched
                                  > SoftRock BPF substituting SMD inductors for the the wound toroids?
                                  >
                                  > I may just give this a whirl!
                                  >
                                  > thanks much,
                                  > ben
                                  > --
                                  > Ben Hall, kd5byb@... - ALWAYS OUTNUMBERED, NEVER OUTGUNNED.
                                  >
                                • netnovel2005
                                  Hello, I use chip inductor as BPF. There is few problem. 3 pole BPF has not enough rejection of 2 times of frequency. I add stop band filter to BPF of 80m
                                  Message 16 of 18 , Mar 2, 2010
                                  • 0 Attachment
                                    Hello,

                                    I use chip inductor as BPF.

                                    There is few problem. 3 pole BPF has not enough rejection of 2 times of frequency. I add stop band filter to BPF of 80m against 40m.

                                    Please look my schematics.

                                    http://zao.jp/radio/GPS/schematics.jpg

                                    C51, L26 is stop band filter.

                                    This stop band filter is not complete. I design 5 pole BPF for next version now.

                                    73
                                    Miura
                                    JA7TDO
                                  • w0fms
                                    I m building the filter board myself right now... whereas the main board was fun to build, the filter board is a PITA.... I d take a lower Q 5 pole filter any
                                    Message 17 of 18 , Mar 3, 2010
                                    • 0 Attachment
                                      I'm building the filter board myself right now... whereas the main board was fun to build, the filter board is a PITA....

                                      I'd take a lower Q 5 pole filter any day over 71 turns x2 plus 50 turns on a core form that's really not big enough for 71 turns anyway. And do I have 69 turns on one and 72 on the other? Maybe... From now on (the last three sections) I'm bringing the wound inductors into work and measuring them on a LCR bridge to make sure we are right. Really otherwise you are guessing. Repeat that 15 times and you have the board.

                                      I do understand that going with good quality Coilcraft-type chip inductors would make the kit a little more expensive... more so if the fiter goes up to 5 poles on each band.. but reproducibility and quality of the product from both a mechanical and electrical standpoint goes up as well. I'd hate to try to run a vibe test on that filter board the way it is.

                                      I hate winding toroids.. and really, to build this correctly one needs to be too careful or have a LCR bridge of some sort. Going to prewound transformers and inductors would help the more neophyte builders which I believe is one of the goals of the Softrocks, is it not?

                                      The V9.0 receiver board converted to all SMT with the filters built in for HF and a plug in or two for VHF (Bigger AVR w/xtal and two more mux ic's.. 8 bands) would be an almost ideal project, IMHO. Yes I said SMT.. it's easier once you get used to it. Especially when the original design is a hybrid leaded/SMT design anyway. I also hate those leaded dipped ceramic caps.. they are way too easy to break.

                                      I'm not faulting Tony (or Jan or anyone else) for the current design.. it makes perfectly good sense based on the evolution of the design. But switching over to all commercial components would almost make a SDR receiver for the masses out of the current design. Yup it might go up to $90 for the kit.. it'd be worth it. My time winding those cores are worth the $30 difference.

                                      At least a new filter board would be a good idea.

                                      Fred W0FMS

                                      --- In softrock40@yahoogroups.com, "netnovel2005" <ja7tdo@...> wrote:
                                      >
                                      >
                                      >
                                      > Hello,
                                      >
                                      > I use chip inductor as BPF.
                                      >
                                      > There is few problem. 3 pole BPF has not enough rejection of 2 times of frequency. I add stop band filter to BPF of 80m against 40m.
                                      >
                                      > Please look my schematics.
                                      >
                                      > http://zao.jp/radio/GPS/schematics.jpg
                                      >
                                      > C51, L26 is stop band filter.
                                      >
                                      > This stop band filter is not complete. I design 5 pole BPF for next version now.
                                      >
                                      > 73
                                      > Miura
                                      > JA7TDO
                                      >
                                    • Peter
                                      I m using five 3rd-order filters, electronically switched, with SMT inductors, and it works fine. Also, Pulse Engineering makes a transformer which works as
                                      Message 18 of 18 , Mar 3, 2010
                                      • 0 Attachment
                                        I'm using five 3rd-order filters, electronically switched, with SMT inductors, and it works fine. Also, Pulse Engineering makes a transformer which works as well as a hand-wound toroid. It's available from mouser in single quantity for less than $3.00.

                                        --- In softrock40@yahoogroups.com, "w0fms" <fred@...> wrote:
                                        >
                                        >
                                        >
                                        >
                                        >
                                        >
                                        > I'm building the filter board myself right now... whereas the main board was fun to build, the filter board is a PITA....
                                        >
                                        > I'd take a lower Q 5 pole filter any day over 71 turns x2 plus 50 turns on a core form that's really not big enough for 71 turns anyway. And do I have 69 turns on one and 72 on the other? Maybe... From now on (the last three sections) I'm bringing the wound inductors into work and measuring them on a LCR bridge to make sure we are right. Really otherwise you are guessing. Repeat that 15 times and you have the board.
                                        >
                                        > I do understand that going with good quality Coilcraft-type chip inductors would make the kit a little more expensive... more so if the fiter goes up to 5 poles on each band.. but reproducibility and quality of the product from both a mechanical and electrical standpoint goes up as well. I'd hate to try to run a vibe test on that filter board the way it is.
                                        >
                                        > I hate winding toroids.. and really, to build this correctly one needs to be too careful or have a LCR bridge of some sort. Going to prewound transformers and inductors would help the more neophyte builders which I believe is one of the goals of the Softrocks, is it not?
                                        >
                                        > The V9.0 receiver board converted to all SMT with the filters built in for HF and a plug in or two for VHF (Bigger AVR w/xtal and two more mux ic's.. 8 bands) would be an almost ideal project, IMHO. Yes I said SMT.. it's easier once you get used to it. Especially when the original design is a hybrid leaded/SMT design anyway. I also hate those leaded dipped ceramic caps.. they are way too easy to break.
                                        >
                                        > I'm not faulting Tony (or Jan or anyone else) for the current design.. it makes perfectly good sense based on the evolution of the design. But switching over to all commercial components would almost make a SDR receiver for the masses out of the current design. Yup it might go up to $90 for the kit.. it'd be worth it. My time winding those cores are worth the $30 difference.
                                        >
                                        > At least a new filter board would be a good idea.
                                        >
                                        > Fred W0FMS
                                        >
                                        > --- In softrock40@yahoogroups.com, "netnovel2005" <ja7tdo@> wrote:
                                        > >
                                        > >
                                        > >
                                        > > Hello,
                                        > >
                                        > > I use chip inductor as BPF.
                                        > >
                                        > > There is few problem. 3 pole BPF has not enough rejection of 2 times of frequency. I add stop band filter to BPF of 80m against 40m.
                                        > >
                                        > > Please look my schematics.
                                        > >
                                        > > http://zao.jp/radio/GPS/schematics.jpg
                                        > >
                                        > > C51, L26 is stop band filter.
                                        > >
                                        > > This stop band filter is not complete. I design 5 pole BPF for next version now.
                                        > >
                                        > > 73
                                        > > Miura
                                        > > JA7TDO
                                        > >
                                        >
                                      Your message has been successfully submitted and would be delivered to recipients shortly.