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

Re: Comparing rx systems

Expand Messages
  • LA3JJ
    Ok Christos, sorry for not writing dBm, I know the diffrence very well. Error comes from me getting old I guess. I do have a calibrated signal source here so I
    Message 1 of 19 , Jan 9, 2012
    • 0 Attachment
      Ok Christos,
      sorry for not writing dBm, I know the diffrence very well. Error comes from me getting old I guess.
      I do have a calibrated signal source here so I will use that to get the absolute values in order. Then I will see what the USB2SDRs noise level really is! Will be a better starting point.
      As I understand you the noise level of the unconnected USB2SDR should then be below -130dBm? The problem I have tried to adress is that when connecting the Ensemble to the USB2SDR there is no increase in the noise (antenna not connected). I was thinking this will indicate that the USB2SDR is the limiting stage?
      Your answer indicates to me that something else is wrong giving too much noise from the USB2SDR. Therefore the next action will be to establish proof calibration of the dBm level. Again thanks for your nice comments in good ham spirit;-)

      73
      Jon

      --- In powersdr-iq@yahoogroups.com, "sv1eia" <sv1eia@...> wrote:
      >
      > Jon Ove,
      >
      > No, your assumption is wrong.
      >
      > On UHFSDR, the USB2SDR can read signals that are -130dBm, so that is by no means the limiting factor.
      >
      > If the rig can hear a weak signal and pass it to the I/Q output, then the ADC will also hear it.
      >
      > The SDR rigs are not Noiseless devices, they do have the largest portion of noise introduced in the path.
      >
      > Also, I guess you are messing dBm with dB.
      > They are different things.
      >
      > As I said, you need to see further into the matter and build a bit of knowledge on the subject.
      >
      > The dBm is an 'absolute' measure whereas dB is 'relevant'.
      >
      > The dBm can ONLY be measured from an RF device since it denotes RF signal EXACT LEVEL.
      >
      > The dB is only a measure of GAIN on any signal.
      >
      > You need to increase the sensitivity of your rig, its as simple as that.
      > Your rig itself creates noise, so you need to elevate the signals above that noise level.
      >
      > Christos SV1EIA
      >
      >
      > --- In powersdr-iq@yahoogroups.com, "LA3JJ" <la3jj@> wrote:
      > >
      > >
      > > Thanks Alan and Christos.
      > > Seems my tests are quite relevant then, and that improvements may have to be done to all stages.
      > > consentrating first on the HPADC:
      > > these data were given:
      > > • 24 bit
      > > • 192KHz sample rate
      > > • S/(N+D): 110dB
      > > • DR: 123dB
      > > • S/N: 123dB
      > >
      > > I would like to know the same data for the plain USB2SDR, may have been posted earlier but I did not find when searching. As I read my results now the evident limitation is the noise level of the USB2SDR itself. The noise from the opamp and the RF will probably also have to be improved. It remains to be seen.
      > > As I now feel I am having a good test setup, not very scientifically but still wspr tells me a lot of on the air performance. Therefore I will do some more tests and document them by posting screenshots in the photos section. I have calibrated the PSDR S-meter by comparison to the K3. Will later look up on the MDS measurement and try to get all stages in top shape..
      > >
      > > Thanks
      > >
      > > 73
      > > Jon Ove, la3jj
      > >
      > > --- In powersdr-iq@yahoogroups.com, "sv1eia" <sv1eia@> wrote:
      > > >
      > > > Hi Jon Ove,
      > > >
      > > > I'll try to give my point of view.
      > > >
      > > > All those homebrew SDR rigs do have lower MDS compared to commercial rigs, they all need pre-amplification to match the low signal performance of the older systems.
      > > > That is not needed though for low bands 160/80/40/30, yet it is needed as we move on higher bands and it is necessary for the 10m.
      > > > Although the better low noise performance of AK5394A does help, it helps though in aspects as Dynamic Range, IMD, linearity, etc.
      > > > For MDS specifically, there is a predominant role in the rig itself since the things like NF (noise figure) and Insertion Loss in rig's stages, does make a difference.
      > > >
      > > > At first, try to have the homebrew rigs level calibrated, so to have an accurate s-meter reading.
      > > > Then, measure the MDS of your rigs and do a plain read comparison with what the MDS specs of your older systems are.
      > > > (MDS is measured with 50ohm termination at antenna input, 500Hz CW, see ARRL's test procedures)
      > > >
      > > > As I've seen, based on my own experience, commercial systems have MDS more or less in the range of -130dBm (NoPreAmp) to -144 dBm (PreAmp), whereas plain homebrew SDR rigs (like my Softrocks) are at -126dBm (my Lima-SDR even with PreAmp is at -123dBm). UHFSDR is at -130dBm yet it has a very good LNA MMIC as PreAmp always enabled (18dB gain).
      > > >
      > > >
      > > > Having the best of ADC, does not mean that will remedy your rig's inability to cope with low signals, yet a bad ADC will cover with noise the low signals that a good rig does hear.
      > > > So, you need both of those components (rig + ADC) in good shape if you want to be able to cope with weak signal reception.
      > > >
      > > > Just my 2c.
      > > >
      > > > :-)
      > > >
      > > > 73,
      > > > Christos SV1EIA
      > > >
      > > >
      > > > --- In powersdr-iq@yahoogroups.com, "LA3JJ" <la3jj@> wrote:
      > > > >
      > > > > Hi Christos and all,
      > > > > I have not been on this group for a while. The reason is that I have been occupied making a completely new setup for on the air testing of SDRs with WSPR. I have now 1 Ensemble Rx II and 2 Ensemble RxTX in adition to my Lima SDR, FA-SDR, K3 and FT-450. I have also made a hybrid splitter so that I can use two receivers on one antenna.
      > > > > In addition to this I found that I had to invest in a Win7 laptop.
      > > > > The result from my first test comparing is posted in the photos section.
      > > > >
      > > > > What I have seen so far is that my SDRs will be a bout 3 dB inferiour to the "analog" setup. But as may be seen from the photo the SDR signal looks cleaner on the WSPR display and it decodes just as many, and at times even more signals!
      > > > >
      > > > > A disappointment is that when signals get really low down in the very low noise level, which I do have here at my home location, the SDRs WILL NOT DETECT the signals that are decoded by my non SDR setups (K3 and FT-450).
      > > > >
      > > > > Doing a test similar to what Christos has just published for his new high performance unit I can see that the noise level stays at abt. -120db in my case with rx plugged/unplugged, and when the antenna is connected it just rises a a few dB (when the band is quite dead). The analog rx line here tells me that there are signals to dig out in there so I think my conclusion of ordering a high performance piggiback is quite close to become a reality.
      > > > > These are just my first impressions from rather limited testruns.
      > > > > I will bring some of the setup with me to EA8 for a couple of weeks and see how it performs under more southern and noisy conditions.
      > > > > Would be nice to know if this is in line with experience of others or if I am on the wrong track?
      > > > >
      > > > > 73 and a wonderful twenty-twelve to you all!
      > > > > Jon Ove, la3jj
      > > > >
      > > >
      > >
      >
    • applewiz2000
      Hi Jon Ove, Christos, et al, Firstly what bands are you comparing on? There have been conflicting measurements presented over the years comparing traditional
      Message 2 of 19 , Jan 9, 2012
      • 0 Attachment
        Hi Jon Ove, Christos, et al,

        Firstly what bands are you comparing on?

        There have been conflicting measurements presented over the years comparing traditional rigs with SDRs. Those with large antennas, such as 4-ele beams on 10m find the sensitivity good enough. Those like me with only a wire antenna find the SDR to be deaf. But on 15m and below it is fine.

        Last weekend I put a preamp circuit into permanent use:
        http://www.m0rzf.co.uk/styled-3/

        There are those that argue the MoBo V4.3 itself makes the sensitivity much worse. Next weekend I hope to compare the "raw" SR 6.3 board with the expanded system. My gut feeling is the MoBo has a loss of <1dB, and will not make sensitivity much worse.

        ~Rob
      • sv1eia
        You re welcome jon Ove :-) Yet, it seems that you still confusing things here. Let me explain a bit further. You cannot see where the noise floor of USB2SDR
        Message 3 of 19 , Jan 9, 2012
        • 0 Attachment
          You're welcome jon Ove :-)

          Yet, it seems that you still confusing things here.

          Let me explain a bit further.

          You cannot see where the noise floor of USB2SDR (alone) in dBm..
          You can Only see that in conjuction with an SDR rig in front of USB2SDR or relative to the dBm measurement of the rig (as I'll tell you further on). And as I said, that rig Will play a major role in sensitivity.
          The rig itself has a noise floor (nothing is Noiseless :-) ) regardless of any ADC that is used at its I/Q outputs.

          The dBm that are present on the rig's antenna from a given signal, at the end of the day, are converted into electrical pulses on the rig's I/Q outputs.

          How large are these electrical pulses, and what is the rig's itself, noise floor, are making the rig's sensitivity.

          For example, for your Lima-SDR, when you connect power to it, without any signal present on its antenna, there is a tiny fluxuation on its I/Q output, a tiny noise, that is the rig's noise floor.
          if you connect a -120 dBm signal on its antenna, you will get a very small signal on Lima-SDR's I/Q outputs, it will be though above the rig's noise floor. If you connect a -132 dBm signal at its antenna, most likely you will get a signal so small, that is below the rig's noise floor, so it is not audible at all because there is noise.
          Even if you put then an ADC (better than AK5394A..) you cannot retrieve that signal because it is below the rig's tself noise floor.

          So, if the ADC's noise floor is lower than what your rig's I/Q output is, then you need to have a more sensitive rig where the antenna signals will be transfered above the rig's itself noise floor.


          I'll give you a very easy rule to see if your ADC or USB2SDR or whatever there, is performing worse or better than your rig in terms of noise floor.

          When you have performed the level calibration test with some instrument as described, then
          1. find the MDS of your rig, that can be done by placing a 50 Ohm termination at its antenna and set it to CW with 500Hz filter. Then read the S-Meter reading on its dBm and write this number. Lets say that is Rig-MDS.
          2. disconnect the I/Q cables from your ADC. Without closing the application or changing mode and filter, see the S-Meter and write its number. Lets say this is ADC-MDS.

          If Rig-MDS is the same as ADC-MDS, eg Rig-MDS:-123, ADC-MDS:-123, then your ADC is the limiting factor.

          Now, if Rig-MDS is higher than ADC-MDS, eg Rig-MDS:-123, ADC-MDS:-132, then your ADC can 'hear' lower signals, but your rig has high noise. In that case, your ADC could be hearing so much better as to what is the difference of Rig-MDS to ADC-MDS and that ADC dynamic range gets wasted because of the rig's own noise floor.

          That is the easiest way to find out quickly where you need to focus on.

          Do not take me wrong, I'm not saying that the USB2SDR included codec is better than AK5394A. Of course the new HPADC option board is way better and way more sensitive.
          I'm just trying to say that perhaps, the homebrew rigs that you mention, have a worst noise floor than the ADC that is listening to their I/Q output.

          I hope these do make sense there.

          73,
          Christos SV1EIA


          --- In powersdr-iq@yahoogroups.com, "LA3JJ" <la3jj@...> wrote:
          >
          > Ok Christos,
          > sorry for not writing dBm, I know the diffrence very well. Error comes from me getting old I guess.
          > I do have a calibrated signal source here so I will use that to get the absolute values in order. Then I will see what the USB2SDRs noise level really is! Will be a better starting point.
          > As I understand you the noise level of the unconnected USB2SDR should then be below -130dBm? The problem I have tried to adress is that when connecting the Ensemble to the USB2SDR there is no increase in the noise (antenna not connected). I was thinking this will indicate that the USB2SDR is the limiting stage?
          > Your answer indicates to me that something else is wrong giving too much noise from the USB2SDR. Therefore the next action will be to establish proof calibration of the dBm level. Again thanks for your nice comments in good ham spirit;-)
          >
          > 73
          > Jon
          >
          > --- In powersdr-iq@yahoogroups.com, "sv1eia" <sv1eia@> wrote:
          > >
          > > Jon Ove,
          > >
          > > No, your assumption is wrong.
          > >
          > > On UHFSDR, the USB2SDR can read signals that are -130dBm, so that is by no means the limiting factor.
          > >
          > > If the rig can hear a weak signal and pass it to the I/Q output, then the ADC will also hear it.
          > >
          > > The SDR rigs are not Noiseless devices, they do have the largest portion of noise introduced in the path.
          > >
          > > Also, I guess you are messing dBm with dB.
          > > They are different things.
          > >
          > > As I said, you need to see further into the matter and build a bit of knowledge on the subject.
          > >
          > > The dBm is an 'absolute' measure whereas dB is 'relevant'.
          > >
          > > The dBm can ONLY be measured from an RF device since it denotes RF signal EXACT LEVEL.
          > >
          > > The dB is only a measure of GAIN on any signal.
          > >
          > > You need to increase the sensitivity of your rig, its as simple as that.
          > > Your rig itself creates noise, so you need to elevate the signals above that noise level.
          > >
          > > Christos SV1EIA
          > >
          > >
          > > --- In powersdr-iq@yahoogroups.com, "LA3JJ" <la3jj@> wrote:
          > > >
          > > >
          > > > Thanks Alan and Christos.
          > > > Seems my tests are quite relevant then, and that improvements may have to be done to all stages.
          > > > consentrating first on the HPADC:
          > > > these data were given:
          > > > • 24 bit
          > > > • 192KHz sample rate
          > > > • S/(N+D): 110dB
          > > > • DR: 123dB
          > > > • S/N: 123dB
          > > >
          > > > I would like to know the same data for the plain USB2SDR, may have been posted earlier but I did not find when searching. As I read my results now the evident limitation is the noise level of the USB2SDR itself. The noise from the opamp and the RF will probably also have to be improved. It remains to be seen.
          > > > As I now feel I am having a good test setup, not very scientifically but still wspr tells me a lot of on the air performance. Therefore I will do some more tests and document them by posting screenshots in the photos section. I have calibrated the PSDR S-meter by comparison to the K3. Will later look up on the MDS measurement and try to get all stages in top shape..
          > > >
          > > > Thanks
          > > >
          > > > 73
          > > > Jon Ove, la3jj
          > > >
          > > > --- In powersdr-iq@yahoogroups.com, "sv1eia" <sv1eia@> wrote:
          > > > >
          > > > > Hi Jon Ove,
          > > > >
          > > > > I'll try to give my point of view.
          > > > >
          > > > > All those homebrew SDR rigs do have lower MDS compared to commercial rigs, they all need pre-amplification to match the low signal performance of the older systems.
          > > > > That is not needed though for low bands 160/80/40/30, yet it is needed as we move on higher bands and it is necessary for the 10m.
          > > > > Although the better low noise performance of AK5394A does help, it helps though in aspects as Dynamic Range, IMD, linearity, etc.
          > > > > For MDS specifically, there is a predominant role in the rig itself since the things like NF (noise figure) and Insertion Loss in rig's stages, does make a difference.
          > > > >
          > > > > At first, try to have the homebrew rigs level calibrated, so to have an accurate s-meter reading.
          > > > > Then, measure the MDS of your rigs and do a plain read comparison with what the MDS specs of your older systems are.
          > > > > (MDS is measured with 50ohm termination at antenna input, 500Hz CW, see ARRL's test procedures)
          > > > >
          > > > > As I've seen, based on my own experience, commercial systems have MDS more or less in the range of -130dBm (NoPreAmp) to -144 dBm (PreAmp), whereas plain homebrew SDR rigs (like my Softrocks) are at -126dBm (my Lima-SDR even with PreAmp is at -123dBm). UHFSDR is at -130dBm yet it has a very good LNA MMIC as PreAmp always enabled (18dB gain).
          > > > >
          > > > >
          > > > > Having the best of ADC, does not mean that will remedy your rig's inability to cope with low signals, yet a bad ADC will cover with noise the low signals that a good rig does hear.
          > > > > So, you need both of those components (rig + ADC) in good shape if you want to be able to cope with weak signal reception.
          > > > >
          > > > > Just my 2c.
          > > > >
          > > > > :-)
          > > > >
          > > > > 73,
          > > > > Christos SV1EIA
          > > > >
          > > > >
          > > > > --- In powersdr-iq@yahoogroups.com, "LA3JJ" <la3jj@> wrote:
          > > > > >
          > > > > > Hi Christos and all,
          > > > > > I have not been on this group for a while. The reason is that I have been occupied making a completely new setup for on the air testing of SDRs with WSPR. I have now 1 Ensemble Rx II and 2 Ensemble RxTX in adition to my Lima SDR, FA-SDR, K3 and FT-450. I have also made a hybrid splitter so that I can use two receivers on one antenna.
          > > > > > In addition to this I found that I had to invest in a Win7 laptop.
          > > > > > The result from my first test comparing is posted in the photos section.
          > > > > >
          > > > > > What I have seen so far is that my SDRs will be a bout 3 dB inferiour to the "analog" setup. But as may be seen from the photo the SDR signal looks cleaner on the WSPR display and it decodes just as many, and at times even more signals!
          > > > > >
          > > > > > A disappointment is that when signals get really low down in the very low noise level, which I do have here at my home location, the SDRs WILL NOT DETECT the signals that are decoded by my non SDR setups (K3 and FT-450).
          > > > > >
          > > > > > Doing a test similar to what Christos has just published for his new high performance unit I can see that the noise level stays at abt. -120db in my case with rx plugged/unplugged, and when the antenna is connected it just rises a a few dB (when the band is quite dead). The analog rx line here tells me that there are signals to dig out in there so I think my conclusion of ordering a high performance piggiback is quite close to become a reality.
          > > > > > These are just my first impressions from rather limited testruns.
          > > > > > I will bring some of the setup with me to EA8 for a couple of weeks and see how it performs under more southern and noisy conditions.
          > > > > > Would be nice to know if this is in line with experience of others or if I am on the wrong track?
          > > > > >
          > > > > > 73 and a wonderful twenty-twelve to you all!
          > > > > > Jon Ove, la3jj
          > > > > >
          > > > >
          > > >
          > >
          >
        • sv1fxo
          Hi all, When you connect the antenna to your rig and the noise on your speakers (or s-meter or panadapter etc) raises then the sensitivity of the receiver is
          Message 4 of 19 , Jan 10, 2012
          • 0 Attachment
            Hi all,
            When you connect the antenna to your rig and the noise on your speakers (or s-meter or panadapter etc) raises then the sensitivity of the receiver is sufficient. In most of the cases sensitivity is not an issue on hf. On the other hand too much sensitivity renders useless some part of receiver's dynamic range that's why in this case the solution is the use of attenuator.

            73 Christos (sv1fxo)
          • Jose Bonanca
            Christos ???? ... -- Jose (Ct1aos)
            Message 5 of 19 , Jan 10, 2012
            • 0 Attachment
              Christos ????

              On 10 January 2012 17:28, sv1fxo <sv1fxo@...> wrote:
               


              Hi all,
              When you connect the antenna to your rig and the noise on your speakers (or s-meter or panadapter etc) raises then the sensitivity of the receiver is sufficient. In most of the cases sensitivity is not an issue on hf. On the other hand too much sensitivity renders useless some part of receiver's dynamic range that's why in this case the solution is the use of attenuator.

              73 Christos (sv1fxo)




              --
              Jose (Ct1aos)
            • LA3JJ
              Christos, Alan and even Jose, Thanks for comments ;-) First I started reading, (that s always what Christos will tell us) The manual for SDR1000 etc. And more
              Message 6 of 19 , Jan 11, 2012
              • 0 Attachment
                Christos, Alan and even Jose,
                Thanks for comments ;-)

                First I started reading, (that's always what Christos will tell us)
                The manual for SDR1000 etc.
                And more important, brushing up the knowledge from scool days far behind. The following links were very useful to get a grip on newer terminilogy:
                http://www.sm5bsz.com/pcdsp/dynmeas.htm
                http://www.qsl.net/pnwvhfs/articles/K6LEW_Receiver_Measurements.pdf
                http://www.adventure-radio.org/ars/pages/lab_text/notes_mds.html

                Then started calibation and measurements. Using my Netw Analyser as signal source. It has a fairly accurate 4dbm output. Alas - , I had just 40dB of proffesional attenuators and I will have to make me another 30dB-one, but couldnt wait for that.
                So I calibrated to the -36dBm mark.

                It was imidiately evident that the sensitivity of my Ensemble TX/TX was far from what to expected, about 30 to 40dB too low. Then I went back to the nice FA-SDR (where a defect relay has just been replaced) and connected it into my USB2SDR setup. Calibrating to -36dB and Bingo!
                It compares much more favourably to the K3 results also at low signal levels.
                What probably has happened on the Ensemble is that one of the BS170 in the Tr-switching is fried? - or there may be somthing else..

                Measuring now with 500Hz bw and getting MDS figures for the FA-SDR which are in the right bin.

                The Ensemble has to be scrutinized, and there is hope for full recovery.

                Will post results in the relevant group sections later.

                73 to all you happy HAMs

                Jon Ove, la3jj

                --- In powersdr-iq@yahoogroups.com, Jose Bonanca <jabct1aos@...> wrote:
                >
                > Christos ????
                >
                > On 10 January 2012 17:28, sv1fxo <sv1fxo@...> wrote:
                >
                > > **
                > >
                > >
                > >
                > > Hi all,
                > > When you connect the antenna to your rig and the noise on your speakers
                > > (or s-meter or panadapter etc) raises then the sensitivity of the receiver
                > > is sufficient. In most of the cases sensitivity is not an issue on hf. On
                > > the other hand too much sensitivity renders useless some part of receiver's
                > > dynamic range that's why in this case the solution is the use of attenuator.
                > >
                > > 73 Christos (sv1fxo)
                > >
                > >
                > >
                >
                >
                >
                > --
                > Jose (Ct1aos)
                >
              • LA3JJ
                Sorry for this late response Rob, I am comparing apples and peaches on 20 and 30m currently. Will move to 10m later. Will do the testing with at least 3
                Message 7 of 19 , Jan 12, 2012
                • 0 Attachment
                  Sorry for this late response Rob,

                  I am comparing apples and peaches on 20 and 30m currently.
                  Will move to 10m later.

                  Will do the testing with at least 3 different SDRs.
                  Using MDS measurements and on the air WSPR receptions.
                  My indications so far is that all the homebrew SDRs will benefit from a selective frontend amplifier.
                  Still having my plans for frontend x-tal filters for all the wspr bands....
                  I think the Ensemble will need about 10dB extra gain in addition to frontend selectivity to be among the peaches.

                  73
                  Jon

                  --- In powersdr-iq@yahoogroups.com, "applewiz2000" <rob.m0rzf@...> wrote:
                  >
                  > Hi Jon Ove, Christos, et al,
                  >
                  > Firstly what bands are you comparing on?
                  >
                  > There have been conflicting measurements presented over the years comparing traditional rigs with SDRs. Those with large antennas, such as 4-ele beams on 10m find the sensitivity good enough. Those like me with only a wire antenna find the SDR to be deaf. But on 15m and below it is fine.
                  >
                  > Last weekend I put a preamp circuit into permanent use:
                  > http://www.m0rzf.co.uk/styled-3/
                  >
                  > There are those that argue the MoBo V4.3 itself makes the sensitivity much worse. Next weekend I hope to compare the "raw" SR 6.3 board with the expanded system. My gut feeling is the MoBo has a loss of <1dB, and will not make sensitivity much worse.
                  >
                  > ~Rob
                  >
                • Alan
                  ... From: LA3JJ Subject: [powersdr-iq] Re: Comparing rx systems Jon, If your Ensemble has very poor sensitivity the Softrock group can help. ... As Rob
                  Message 8 of 19 , Jan 12, 2012
                  • 0 Attachment
                    ----- Original Message -----
                    From: "LA3JJ"
                    Subject: [powersdr-iq] Re: Comparing rx systems


                    Jon,

                    If your Ensemble has very poor sensitivity the Softrock group can help.

                    > My indications so far is that all the homebrew SDRs will benefit from a
                    > selective frontend amplifier.

                    As Rob indicates the actual sensitivity you need will vary according to the
                    band, antenna and local noise. As Christos says for absolute best
                    performance all stages of the SDR need to be matched.

                    > Still having my plans for frontend x-tal filters for all the wspr
                    > bands....

                    This is interesting but debatable. I consider this would not improve things,
                    most likely cause deterioration. I think that even narrowing PSDR's
                    bandwidth adversely affects WSPR.
                    But I'm just guessing, I wait to see your results.

                    What always surprises me is that a few components in front of a computer
                    soundcard can even be compared with real, dedicated radios (OK, the USB2SDR
                    is perhaps the ultimate soundcard)

                    73 Alan G4ZFQ
                  • sv1eia
                    Jon Ove, It seems that you ve found after all what was causing your problems there. If there is a 40dB difference in noise floor between the two rigs then it
                    Message 9 of 19 , Jan 12, 2012
                    • 0 Attachment
                      Jon Ove,

                      It seems that you've found after all what was causing your problems there.

                      If there is a 40dB difference in noise floor between the two rigs then it is for sure something happening to one of them and you need to see why its behaving like this, could be something simple though like a solder bridge somewhere.

                      Now, regarding the filter you mention, in my view, it will do nothing to improve an SDR rig and there is a great chance that you will add a component that might bring in problems.

                      Actually, I'm not entirely sure what you exactly mean as a crystal filter there and where you want to place such a component. Usually that comes after an IF stage on a heterodyne, yet our SDR homebrew rigs are not utilizing anything of sort..

                      The only thing that I do understand is that you want some low noise RF preamplifier in order to get better sensitivity on higher bands (lower bands have inherited noise that can be avoided only with better antennas). And its good to have a preselector/band pass filter between the antenna and the pre-amp in order to avoid overloading of the RF preamp.
                      Yet, you should understand though, that in case you engage a preamp, then the SDR strong point which is selectivity, gets lower and becames as what is the selectivity of the preamp itself. So any good numbers in IMD DR of the SDR are no more what you will enjoy. Thats the caveat of the RF preamp on a SDR rig.
                      Not to mention of course that all RF preamps got Noise Figure which adds to your overall setup.


                      I'll give you a couple of advices, its up to you to investigate and follow them if you like.

                      1. As my friend Christos SV1FXO said, when you operate your SDR rig, observing the noise floor, when you connect your antenna if you notice an increase upwards in the noise floor, then your rig's sensitivity is enough. (In HF we do need attenuation :-) ). At that point just try to have a better antenna.

                      2. If you want to operate WSPR or any other mode through third party external to PowerSDR applications via VAC, you need to give the AGC control on the external application. For this, you should max-out the AGC-T of PowerSDR console. You can find references to that on the links section and their content (reading..).

                      3. If you want to put an RF preamp, do it only for the 10m band and only if you check the first point (1.). In my view, only 10m needs an RF preamp and that is if you happen to have a good antenna. Putting a low noise RF preamp by having a not good antenna, you just make a bad situation worst.


                      As said, these are my views only and I do not expect anyone to follow them.
                      Its up to you.

                      73,
                      Christos SV1EIA


                      --- In powersdr-iq@yahoogroups.com, "Alan" <alan4alan@...> wrote:
                      >
                      >
                      > ----- Original Message -----
                      > From: "LA3JJ"
                      > Subject: [powersdr-iq] Re: Comparing rx systems
                      >
                      >
                      > Jon,
                      >
                      > If your Ensemble has very poor sensitivity the Softrock group can help.
                      >
                      > > My indications so far is that all the homebrew SDRs will benefit from a
                      > > selective frontend amplifier.
                      >
                      > As Rob indicates the actual sensitivity you need will vary according to the
                      > band, antenna and local noise. As Christos says for absolute best
                      > performance all stages of the SDR need to be matched.
                      >
                      > > Still having my plans for frontend x-tal filters for all the wspr
                      > > bands....
                      >
                      > This is interesting but debatable. I consider this would not improve things,
                      > most likely cause deterioration. I think that even narrowing PSDR's
                      > bandwidth adversely affects WSPR.
                      > But I'm just guessing, I wait to see your results.
                      >
                      > What always surprises me is that a few components in front of a computer
                      > soundcard can even be compared with real, dedicated radios (OK, the USB2SDR
                      > is perhaps the ultimate soundcard)
                      >
                      > 73 Alan G4ZFQ
                      >
                    • LA3JJ
                      Thanks Christos for advice, I m reading and following. to your point no 1 : The fact that the noise did not increase when antenna was connected was what got me
                      Message 10 of 19 , Jan 13, 2012
                      • 0 Attachment
                        Thanks Christos for advice,
                        I'm reading and following.

                        to your point no 1 :
                        The fact that the noise did not increase when antenna was connected was what got me started. It just wasnt good enough. The advice I'm getting here from everybody is that the sensitivity should be good enough on all bands exept 12 and 10. So it is possible to get more juice out of my setup. I will consentrate on that.

                        to your point 2 :
                        I have been using AGC-T on abt 100 will set it at 120 from no on.

                        to your point 3 :

                        I will put my plans for a preselector on hold

                        As long as the wspr decoding is giving 3-6dB less wspr-SNR than the K3, and this difference increases as signals go weaker ending up with unability to decode signals that the K3 setup will decode consistently, I am not satisfied with my USB2SDR - SDR setup.
                        From what you all say it should perform better.
                        There is clearly some issue I am missing here.

                        In addition when looking at the PowrSDR-IQ display it strikes me that on your published screenshots you have a "noise baseline" of below -130dBm. On my setup it never raches -120dBm. - and it does not move acording to connected/unconnected/unpowered SDR. The level calibration did not change that problem.

                        The /setup/general/usb2sdr/HW RX I/Q gain may have to be used? I will try now to increase the dual sliders and to recalibrate after the new setting? Not seen a documentation on how this should be done. You may put me on the right track here also.

                        Even if it is Friday 13th it may be a good day for getting this right.

                        73
                        Jon Ove

                        --- In powersdr-iq@yahoogroups.com, "sv1eia" <sv1eia@...> wrote:
                        >
                        > Jon Ove,
                        >
                        > It seems that you've found after all what was causing your problems there.
                        >
                        > If there is a 40dB difference in noise floor between the two rigs then it is for sure something happening to one of them and you need to see why its behaving like this, could be something simple though like a solder bridge somewhere.
                        >
                        > Now, regarding the filter you mention, in my view, it will do nothing to improve an SDR rig and there is a great chance that you will add a component that might bring in problems.
                        >
                        > Actually, I'm not entirely sure what you exactly mean as a crystal filter there and where you want to place such a component. Usually that comes after an IF stage on a heterodyne, yet our SDR homebrew rigs are not utilizing anything of sort..
                        >
                        > The only thing that I do understand is that you want some low noise RF preamplifier in order to get better sensitivity on higher bands (lower bands have inherited noise that can be avoided only with better antennas). And its good to have a preselector/band pass filter between the antenna and the pre-amp in order to avoid overloading of the RF preamp.
                        > Yet, you should understand though, that in case you engage a preamp, then the SDR strong point which is selectivity, gets lower and becames as what is the selectivity of the preamp itself. So any good numbers in IMD DR of the SDR are no more what you will enjoy. Thats the caveat of the RF preamp on a SDR rig.
                        > Not to mention of course that all RF preamps got Noise Figure which adds to your overall setup.
                        >
                        >
                        > I'll give you a couple of advices, its up to you to investigate and follow them if you like.
                        >
                        > 1. As my friend Christos SV1FXO said, when you operate your SDR rig, observing the noise floor, when you connect your antenna if you notice an increase upwards in the noise floor, then your rig's sensitivity is enough. (In HF we do need attenuation :-) ). At that point just try to have a better antenna.
                        >
                        > 2. If you want to operate WSPR or any other mode through third party external to PowerSDR applications via VAC, you need to give the AGC control on the external application. For this, you should max-out the AGC-T of PowerSDR console. You can find references to that on the links section and their content (reading..).
                        >
                        > 3. If you want to put an RF preamp, do it only for the 10m band and only if you check the first point (1.). In my view, only 10m needs an RF preamp and that is if you happen to have a good antenna. Putting a low noise RF preamp by having a not good antenna, you just make a bad situation worst.
                        >
                        >
                        > As said, these are my views only and I do not expect anyone to follow them.
                        > Its up to you.
                        >
                        > 73,
                        > Christos SV1EIA
                        >
                        >
                        > --- In powersdr-iq@yahoogroups.com, "Alan" <alan4alan@> wrote:
                        > >
                        > >
                        > > ----- Original Message -----
                        > > From: "LA3JJ"
                        > > Subject: [powersdr-iq] Re: Comparing rx systems
                        > >
                        > >
                        > > Jon,
                        > >
                        > > If your Ensemble has very poor sensitivity the Softrock group can help.
                        > >
                        > > > My indications so far is that all the homebrew SDRs will benefit from a
                        > > > selective frontend amplifier.
                        > >
                        > > As Rob indicates the actual sensitivity you need will vary according to the
                        > > band, antenna and local noise. As Christos says for absolute best
                        > > performance all stages of the SDR need to be matched.
                        > >
                        > > > Still having my plans for frontend x-tal filters for all the wspr
                        > > > bands....
                        > >
                        > > This is interesting but debatable. I consider this would not improve things,
                        > > most likely cause deterioration. I think that even narrowing PSDR's
                        > > bandwidth adversely affects WSPR.
                        > > But I'm just guessing, I wait to see your results.
                        > >
                        > > What always surprises me is that a few components in front of a computer
                        > > soundcard can even be compared with real, dedicated radios (OK, the USB2SDR
                        > > is perhaps the ultimate soundcard)
                        > >
                        > > 73 Alan G4ZFQ
                        > >
                        >
                      • sv1eia
                        ... On which bands it did not increase? If it was on 10m, it is more or less like it on homebrew SDR rigs. ... The HW Rx I/Q gain can function as an IF
                        Message 11 of 19 , Jan 13, 2012
                        • 0 Attachment
                          --- In powersdr-iq@yahoogroups.com, "LA3JJ" <la3jj@...> wrote:
                          >
                          > Thanks Christos for advice,
                          > I'm reading and following.
                          >
                          > to your point no 1 :
                          > The fact that the noise did not increase when antenna was connected was what got me started. It just wasnt good enough. The advice I'm getting here from everybody is that the sensitivity should be good enough on all bands exept 12 and 10. So it is possible to get more juice out of my setup. I will consentrate on that.

                          On which bands it did not increase? If it was on 10m, it is more or less like it on homebrew SDR rigs.

                          >
                          > to your point 2 :
                          > I have been using AGC-T on abt 100 will set it at 120 from no on.
                          >
                          > to your point 3 :
                          >
                          > I will put my plans for a preselector on hold
                          >
                          > As long as the wspr decoding is giving 3-6dB less wspr-SNR than the K3, and this difference increases as signals go weaker ending up with unability to decode signals that the K3 setup will decode consistently, I am not satisfied with my USB2SDR - SDR setup.
                          > From what you all say it should perform better.
                          > There is clearly some issue I am missing here.
                          >
                          > In addition when looking at the PowrSDR-IQ display it strikes me that on your published screenshots you have a "noise baseline" of below -130dBm. On my setup it never raches -120dBm. - and it does not move acording to connected/unconnected/unpowered SDR. The level calibration did not change that problem.
                          >
                          > The /setup/general/usb2sdr/HW RX I/Q gain may have to be used? I will try now to increase the dual sliders and to recalibrate after the new setting? Not seen a documentation on how this should be done. You may put me on the right track here also.

                          The HW Rx I/Q gain can function as an 'IF' preamp, its not RF. Yet from what I've seen when engaged it up to 24, it gives about 3 to 6 dB better MDS and a bit more flat noise floor.

                          What I understand though from all your posts, is that you compare the K3 to your homebrew setups there..
                          Thats a bit too high on expectations isnt it?
                          Comparing the number one or two of the Sherwood Eng. Labs sort list, with something that cost just a fraction of the commercial rig you got there. And even though it gives as you said a marginal difference of 3 to 6 dB SNR.

                          If I was in your position I would really be very happy that managed to get that close !!

                          Best 73,
                          Christos SV1EIA


                          PS
                          You didnt tell me yet what happens when you unplug the I/Q Rx cable from USB2SDR and notice the noise floor then, does it get lower or is it remain the same?


                          >
                          > Even if it is Friday 13th it may be a good day for getting this right.
                          >
                          > 73
                          > Jon Ove
                          >
                          > --- In powersdr-iq@yahoogroups.com, "sv1eia" <sv1eia@> wrote:
                          > >
                          > > Jon Ove,
                          > >
                          > > It seems that you've found after all what was causing your problems there.
                          > >
                          > > If there is a 40dB difference in noise floor between the two rigs then it is for sure something happening to one of them and you need to see why its behaving like this, could be something simple though like a solder bridge somewhere.
                          > >
                          > > Now, regarding the filter you mention, in my view, it will do nothing to improve an SDR rig and there is a great chance that you will add a component that might bring in problems.
                          > >
                          > > Actually, I'm not entirely sure what you exactly mean as a crystal filter there and where you want to place such a component. Usually that comes after an IF stage on a heterodyne, yet our SDR homebrew rigs are not utilizing anything of sort..
                          > >
                          > > The only thing that I do understand is that you want some low noise RF preamplifier in order to get better sensitivity on higher bands (lower bands have inherited noise that can be avoided only with better antennas). And its good to have a preselector/band pass filter between the antenna and the pre-amp in order to avoid overloading of the RF preamp.
                          > > Yet, you should understand though, that in case you engage a preamp, then the SDR strong point which is selectivity, gets lower and becames as what is the selectivity of the preamp itself. So any good numbers in IMD DR of the SDR are no more what you will enjoy. Thats the caveat of the RF preamp on a SDR rig.
                          > > Not to mention of course that all RF preamps got Noise Figure which adds to your overall setup.
                          > >
                          > >
                          > > I'll give you a couple of advices, its up to you to investigate and follow them if you like.
                          > >
                          > > 1. As my friend Christos SV1FXO said, when you operate your SDR rig, observing the noise floor, when you connect your antenna if you notice an increase upwards in the noise floor, then your rig's sensitivity is enough. (In HF we do need attenuation :-) ). At that point just try to have a better antenna.
                          > >
                          > > 2. If you want to operate WSPR or any other mode through third party external to PowerSDR applications via VAC, you need to give the AGC control on the external application. For this, you should max-out the AGC-T of PowerSDR console. You can find references to that on the links section and their content (reading..).
                          > >
                          > > 3. If you want to put an RF preamp, do it only for the 10m band and only if you check the first point (1.). In my view, only 10m needs an RF preamp and that is if you happen to have a good antenna. Putting a low noise RF preamp by having a not good antenna, you just make a bad situation worst.
                          > >
                          > >
                          > > As said, these are my views only and I do not expect anyone to follow them.
                          > > Its up to you.
                          > >
                          > > 73,
                          > > Christos SV1EIA
                          > >
                          > >
                          > > --- In powersdr-iq@yahoogroups.com, "Alan" <alan4alan@> wrote:
                          > > >
                          > > >
                          > > > ----- Original Message -----
                          > > > From: "LA3JJ"
                          > > > Subject: [powersdr-iq] Re: Comparing rx systems
                          > > >
                          > > >
                          > > > Jon,
                          > > >
                          > > > If your Ensemble has very poor sensitivity the Softrock group can help.
                          > > >
                          > > > > My indications so far is that all the homebrew SDRs will benefit from a
                          > > > > selective frontend amplifier.
                          > > >
                          > > > As Rob indicates the actual sensitivity you need will vary according to the
                          > > > band, antenna and local noise. As Christos says for absolute best
                          > > > performance all stages of the SDR need to be matched.
                          > > >
                          > > > > Still having my plans for frontend x-tal filters for all the wspr
                          > > > > bands....
                          > > >
                          > > > This is interesting but debatable. I consider this would not improve things,
                          > > > most likely cause deterioration. I think that even narrowing PSDR's
                          > > > bandwidth adversely affects WSPR.
                          > > > But I'm just guessing, I wait to see your results.
                          > > >
                          > > > What always surprises me is that a few components in front of a computer
                          > > > soundcard can even be compared with real, dedicated radios (OK, the USB2SDR
                          > > > is perhaps the ultimate soundcard)
                          > > >
                          > > > 73 Alan G4ZFQ
                          > > >
                          > >
                          >
                        • LA3JJ
                          All my tests so far have been on 30m or 20m. I fully agree and understand that I am comparing apples with peaches. I have very good news now regarding the
                          Message 12 of 19 , Jan 13, 2012
                          • 0 Attachment
                            All my tests so far have been on 30m or 20m.
                            I fully agree and understand that I am comparing apples with peaches.

                            I have very good news now regarding the USB2SDR here.
                            I adjusted the HW IF-gain to 30 and recalibrated the level.
                            Noise floor dropped to abt -136dBm!

                            Did the 500Hz MDS measurement:
                            - open antenna: -118dBm
                            - 50 ohm at antenna: -121dBm
                            - audio disconnected: -123dBm

                            This kind of measurement was not possible before, it all showed the same value.

                            It shows that without increasing the "IF-gain" the USB2SDR is not sensitive enough. This is probably also due to the protection voltage divider in the audio line. So now I can calculate the MDS and the noise figure of the FA-SDR. Next will be to get the Ensembles in good order and do similar measurements comparing "the apples".
                            Later I can go on and see how different from the peaches they really are. Even if I know that there will be a difference it is interesting to see what it is, and it stresses me to tune hardware and software for optimum. I may be able to gain a db or two also by redesigning the protection network in the audio line. I wonder if there is a downside of increasing the HW gain even further?

                            Had a good feeling of this Friday the 13th!

                            73
                            Jon Ove

                            --- In powersdr-iq@yahoogroups.com, "sv1eia" <sv1eia@...> wrote:
                            >
                            >
                            >
                            > --- In powersdr-iq@yahoogroups.com, "LA3JJ" <la3jj@> wrote:
                            > >
                            > > Thanks Christos for advice,
                            > > I'm reading and following.
                            > >
                            > > to your point no 1 :
                            > > The fact that the noise did not increase when antenna was connected was what got me started. It just wasnt good enough. The advice I'm getting here from everybody is that the sensitivity should be good enough on all bands exept 12 and 10. So it is possible to get more juice out of my setup. I will consentrate on that.
                            >
                            > On which bands it did not increase? If it was on 10m, it is more or less like it on homebrew SDR rigs.
                            >
                            > >
                            > > to your point 2 :
                            > > I have been using AGC-T on abt 100 will set it at 120 from no on.
                            > >
                            > > to your point 3 :
                            > >
                            > > I will put my plans for a preselector on hold
                            > >
                            > > As long as the wspr decoding is giving 3-6dB less wspr-SNR than the K3, and this difference increases as signals go weaker ending up with unability to decode signals that the K3 setup will decode consistently, I am not satisfied with my USB2SDR - SDR setup.
                            > > From what you all say it should perform better.
                            > > There is clearly some issue I am missing here.
                            > >
                            > > In addition when looking at the PowrSDR-IQ display it strikes me that on your published screenshots you have a "noise baseline" of below -130dBm. On my setup it never raches -120dBm. - and it does not move acording to connected/unconnected/unpowered SDR. The level calibration did not change that problem.
                            > >
                            > > The /setup/general/usb2sdr/HW RX I/Q gain may have to be used? I will try now to increase the dual sliders and to recalibrate after the new setting? Not seen a documentation on how this should be done. You may put me on the right track here also.
                            >
                            > The HW Rx I/Q gain can function as an 'IF' preamp, its not RF. Yet from what I've seen when engaged it up to 24, it gives about 3 to 6 dB better MDS and a bit more flat noise floor.
                            >
                            > What I understand though from all your posts, is that you compare the K3 to your homebrew setups there..
                            > Thats a bit too high on expectations isnt it?
                            > Comparing the number one or two of the Sherwood Eng. Labs sort list, with something that cost just a fraction of the commercial rig you got there. And even though it gives as you said a marginal difference of 3 to 6 dB SNR.
                            >
                            > If I was in your position I would really be very happy that managed to get that close !!
                            >
                            > Best 73,
                            > Christos SV1EIA
                            >
                            >
                            > PS
                            > You didnt tell me yet what happens when you unplug the I/Q Rx cable from USB2SDR and notice the noise floor then, does it get lower or is it remain the same?
                            >
                            >
                            > >
                            > > Even if it is Friday 13th it may be a good day for getting this right.
                            > >
                            > > 73
                            > > Jon Ove
                            > >
                            > > --- In powersdr-iq@yahoogroups.com, "sv1eia" <sv1eia@> wrote:
                            > > >
                            > > > Jon Ove,
                            > > >
                            > > > It seems that you've found after all what was causing your problems there.
                            > > >
                            > > > If there is a 40dB difference in noise floor between the two rigs then it is for sure something happening to one of them and you need to see why its behaving like this, could be something simple though like a solder bridge somewhere.
                            > > >
                            > > > Now, regarding the filter you mention, in my view, it will do nothing to improve an SDR rig and there is a great chance that you will add a component that might bring in problems.
                            > > >
                            > > > Actually, I'm not entirely sure what you exactly mean as a crystal filter there and where you want to place such a component. Usually that comes after an IF stage on a heterodyne, yet our SDR homebrew rigs are not utilizing anything of sort..
                            > > >
                            > > > The only thing that I do understand is that you want some low noise RF preamplifier in order to get better sensitivity on higher bands (lower bands have inherited noise that can be avoided only with better antennas). And its good to have a preselector/band pass filter between the antenna and the pre-amp in order to avoid overloading of the RF preamp.
                            > > > Yet, you should understand though, that in case you engage a preamp, then the SDR strong point which is selectivity, gets lower and becames as what is the selectivity of the preamp itself. So any good numbers in IMD DR of the SDR are no more what you will enjoy. Thats the caveat of the RF preamp on a SDR rig.
                            > > > Not to mention of course that all RF preamps got Noise Figure which adds to your overall setup.
                            > > >
                            > > >
                            > > > I'll give you a couple of advices, its up to you to investigate and follow them if you like.
                            > > >
                            > > > 1. As my friend Christos SV1FXO said, when you operate your SDR rig, observing the noise floor, when you connect your antenna if you notice an increase upwards in the noise floor, then your rig's sensitivity is enough. (In HF we do need attenuation :-) ). At that point just try to have a better antenna.
                            > > >
                            > > > 2. If you want to operate WSPR or any other mode through third party external to PowerSDR applications via VAC, you need to give the AGC control on the external application. For this, you should max-out the AGC-T of PowerSDR console. You can find references to that on the links section and their content (reading..).
                            > > >
                            > > > 3. If you want to put an RF preamp, do it only for the 10m band and only if you check the first point (1.). In my view, only 10m needs an RF preamp and that is if you happen to have a good antenna. Putting a low noise RF preamp by having a not good antenna, you just make a bad situation worst.
                            > > >
                            > > >
                            > > > As said, these are my views only and I do not expect anyone to follow them.
                            > > > Its up to you.
                            > > >
                            > > > 73,
                            > > > Christos SV1EIA
                            > > >
                            > > >
                            > > > --- In powersdr-iq@yahoogroups.com, "Alan" <alan4alan@> wrote:
                            > > > >
                            > > > >
                            > > > > ----- Original Message -----
                            > > > > From: "LA3JJ"
                            > > > > Subject: [powersdr-iq] Re: Comparing rx systems
                            > > > >
                            > > > >
                            > > > > Jon,
                            > > > >
                            > > > > If your Ensemble has very poor sensitivity the Softrock group can help.
                            > > > >
                            > > > > > My indications so far is that all the homebrew SDRs will benefit from a
                            > > > > > selective frontend amplifier.
                            > > > >
                            > > > > As Rob indicates the actual sensitivity you need will vary according to the
                            > > > > band, antenna and local noise. As Christos says for absolute best
                            > > > > performance all stages of the SDR need to be matched.
                            > > > >
                            > > > > > Still having my plans for frontend x-tal filters for all the wspr
                            > > > > > bands....
                            > > > >
                            > > > > This is interesting but debatable. I consider this would not improve things,
                            > > > > most likely cause deterioration. I think that even narrowing PSDR's
                            > > > > bandwidth adversely affects WSPR.
                            > > > > But I'm just guessing, I wait to see your results.
                            > > > >
                            > > > > What always surprises me is that a few components in front of a computer
                            > > > > soundcard can even be compared with real, dedicated radios (OK, the USB2SDR
                            > > > > is perhaps the ultimate soundcard)
                            > > > >
                            > > > > 73 Alan G4ZFQ
                            > > > >
                            > > >
                            > >
                            >
                          • sv1eia
                            ... Ok, that means USB2SDR, alone, at this stage has lower noise floor than the rig, thats good :-) Calibrating at -36dBm is a bit high. Try to have something
                            Message 13 of 19 , Jan 13, 2012
                            • 0 Attachment
                              --- In powersdr-iq@yahoogroups.com, "LA3JJ" <la3jj@...> wrote:
                              >
                              > All my tests so far have been on 30m or 20m.
                              > I fully agree and understand that I am comparing apples with peaches.
                              >
                              > I have very good news now regarding the USB2SDR here.
                              > I adjusted the HW IF-gain to 30 and recalibrated the level.
                              > Noise floor dropped to abt -136dBm!
                              >
                              > Did the 500Hz MDS measurement:
                              > - open antenna: -118dBm
                              > - 50 ohm at antenna: -121dBm
                              > - audio disconnected: -123dBm

                              Ok, that means USB2SDR, alone, at this stage has lower noise floor than the rig, thats good :-)

                              Calibrating at -36dBm is a bit high. Try to have something in the -73dBm range as the calibrating signal. The lower, the better, its more accurate.

                              >
                              > This kind of measurement was not possible before, it all showed the same value.

                              That is strange, it shouldnt be this way. I guess the level converted/limiter there might play a role in your case.
                              As you said below, have another look at the protection divider for any unwanted attenuation. Its not really needed if the rig has plain 5V opamps like Softrock/Ensemble and FA-SDR.

                              >
                              > It shows that without increasing the "IF-gain" the USB2SDR is not sensitive enough. This is probably also due to the protection voltage divider in the audio line. So now I can calculate the MDS and the noise figure of the FA-SDR. Next will be to get the Ensembles in good order and do similar measurements comparing "the apples".
                              > Later I can go on and see how different from the peaches they really are. Even if I know that there will be a difference it is interesting to see what it is, and it stresses me to tune hardware and software for optimum. I may be able to gain a db or two also by redesigning the protection network in the audio line. I wonder if there is a downside of increasing the HW gain even further?
                              >
                              > Had a good feeling of this Friday the 13th!
                              >


                              It looks you gonna be busy for quite some time :-)

                              73,
                              Christos SV1EIA


                              > 73
                              > Jon Ove
                              >
                              > --- In powersdr-iq@yahoogroups.com, "sv1eia" <sv1eia@> wrote:
                              > >
                              > >
                              > >
                              > > --- In powersdr-iq@yahoogroups.com, "LA3JJ" <la3jj@> wrote:
                              > > >
                              > > > Thanks Christos for advice,
                              > > > I'm reading and following.
                              > > >
                              > > > to your point no 1 :
                              > > > The fact that the noise did not increase when antenna was connected was what got me started. It just wasnt good enough. The advice I'm getting here from everybody is that the sensitivity should be good enough on all bands exept 12 and 10. So it is possible to get more juice out of my setup. I will consentrate on that.
                              > >
                              > > On which bands it did not increase? If it was on 10m, it is more or less like it on homebrew SDR rigs.
                              > >
                              > > >
                              > > > to your point 2 :
                              > > > I have been using AGC-T on abt 100 will set it at 120 from no on.
                              > > >
                              > > > to your point 3 :
                              > > >
                              > > > I will put my plans for a preselector on hold
                              > > >
                              > > > As long as the wspr decoding is giving 3-6dB less wspr-SNR than the K3, and this difference increases as signals go weaker ending up with unability to decode signals that the K3 setup will decode consistently, I am not satisfied with my USB2SDR - SDR setup.
                              > > > From what you all say it should perform better.
                              > > > There is clearly some issue I am missing here.
                              > > >
                              > > > In addition when looking at the PowrSDR-IQ display it strikes me that on your published screenshots you have a "noise baseline" of below -130dBm. On my setup it never raches -120dBm. - and it does not move acording to connected/unconnected/unpowered SDR. The level calibration did not change that problem.
                              > > >
                              > > > The /setup/general/usb2sdr/HW RX I/Q gain may have to be used? I will try now to increase the dual sliders and to recalibrate after the new setting? Not seen a documentation on how this should be done. You may put me on the right track here also.
                              > >
                              > > The HW Rx I/Q gain can function as an 'IF' preamp, its not RF. Yet from what I've seen when engaged it up to 24, it gives about 3 to 6 dB better MDS and a bit more flat noise floor.
                              > >
                              > > What I understand though from all your posts, is that you compare the K3 to your homebrew setups there..
                              > > Thats a bit too high on expectations isnt it?
                              > > Comparing the number one or two of the Sherwood Eng. Labs sort list, with something that cost just a fraction of the commercial rig you got there. And even though it gives as you said a marginal difference of 3 to 6 dB SNR.
                              > >
                              > > If I was in your position I would really be very happy that managed to get that close !!
                              > >
                              > > Best 73,
                              > > Christos SV1EIA
                              > >
                              > >
                              > > PS
                              > > You didnt tell me yet what happens when you unplug the I/Q Rx cable from USB2SDR and notice the noise floor then, does it get lower or is it remain the same?
                              > >
                              > >
                              > > >
                              > > > Even if it is Friday 13th it may be a good day for getting this right.
                              > > >
                              > > > 73
                              > > > Jon Ove
                              > > >
                              > > > --- In powersdr-iq@yahoogroups.com, "sv1eia" <sv1eia@> wrote:
                              > > > >
                              > > > > Jon Ove,
                              > > > >
                              > > > > It seems that you've found after all what was causing your problems there.
                              > > > >
                              > > > > If there is a 40dB difference in noise floor between the two rigs then it is for sure something happening to one of them and you need to see why its behaving like this, could be something simple though like a solder bridge somewhere.
                              > > > >
                              > > > > Now, regarding the filter you mention, in my view, it will do nothing to improve an SDR rig and there is a great chance that you will add a component that might bring in problems.
                              > > > >
                              > > > > Actually, I'm not entirely sure what you exactly mean as a crystal filter there and where you want to place such a component. Usually that comes after an IF stage on a heterodyne, yet our SDR homebrew rigs are not utilizing anything of sort..
                              > > > >
                              > > > > The only thing that I do understand is that you want some low noise RF preamplifier in order to get better sensitivity on higher bands (lower bands have inherited noise that can be avoided only with better antennas). And its good to have a preselector/band pass filter between the antenna and the pre-amp in order to avoid overloading of the RF preamp.
                              > > > > Yet, you should understand though, that in case you engage a preamp, then the SDR strong point which is selectivity, gets lower and becames as what is the selectivity of the preamp itself. So any good numbers in IMD DR of the SDR are no more what you will enjoy. Thats the caveat of the RF preamp on a SDR rig.
                              > > > > Not to mention of course that all RF preamps got Noise Figure which adds to your overall setup.
                              > > > >
                              > > > >
                              > > > > I'll give you a couple of advices, its up to you to investigate and follow them if you like.
                              > > > >
                              > > > > 1. As my friend Christos SV1FXO said, when you operate your SDR rig, observing the noise floor, when you connect your antenna if you notice an increase upwards in the noise floor, then your rig's sensitivity is enough. (In HF we do need attenuation :-) ). At that point just try to have a better antenna.
                              > > > >
                              > > > > 2. If you want to operate WSPR or any other mode through third party external to PowerSDR applications via VAC, you need to give the AGC control on the external application. For this, you should max-out the AGC-T of PowerSDR console. You can find references to that on the links section and their content (reading..).
                              > > > >
                              > > > > 3. If you want to put an RF preamp, do it only for the 10m band and only if you check the first point (1.). In my view, only 10m needs an RF preamp and that is if you happen to have a good antenna. Putting a low noise RF preamp by having a not good antenna, you just make a bad situation worst.
                              > > > >
                              > > > >
                              > > > > As said, these are my views only and I do not expect anyone to follow them.
                              > > > > Its up to you.
                              > > > >
                              > > > > 73,
                              > > > > Christos SV1EIA
                              > > > >
                              > > > >
                              > > > > --- In powersdr-iq@yahoogroups.com, "Alan" <alan4alan@> wrote:
                              > > > > >
                              > > > > >
                              > > > > > ----- Original Message -----
                              > > > > > From: "LA3JJ"
                              > > > > > Subject: [powersdr-iq] Re: Comparing rx systems
                              > > > > >
                              > > > > >
                              > > > > > Jon,
                              > > > > >
                              > > > > > If your Ensemble has very poor sensitivity the Softrock group can help.
                              > > > > >
                              > > > > > > My indications so far is that all the homebrew SDRs will benefit from a
                              > > > > > > selective frontend amplifier.
                              > > > > >
                              > > > > > As Rob indicates the actual sensitivity you need will vary according to the
                              > > > > > band, antenna and local noise. As Christos says for absolute best
                              > > > > > performance all stages of the SDR need to be matched.
                              > > > > >
                              > > > > > > Still having my plans for frontend x-tal filters for all the wspr
                              > > > > > > bands....
                              > > > > >
                              > > > > > This is interesting but debatable. I consider this would not improve things,
                              > > > > > most likely cause deterioration. I think that even narrowing PSDR's
                              > > > > > bandwidth adversely affects WSPR.
                              > > > > > But I'm just guessing, I wait to see your results.
                              > > > > >
                              > > > > > What always surprises me is that a few components in front of a computer
                              > > > > > soundcard can even be compared with real, dedicated radios (OK, the USB2SDR
                              > > > > > is perhaps the ultimate soundcard)
                              > > > > >
                              > > > > > 73 Alan G4ZFQ
                              > > > > >
                              > > > >
                              > > >
                              > >
                              >
                            Your message has been successfully submitted and would be delivered to recipients shortly.