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

Re: [softrock40] USB power for SRv9? - conclusions

Expand Messages
  • Jeff Blaine
    Bob and Tony, My original tests were made using a USB and line-in tap on my docking station. Tonight I moved to a usb/line-in pair on the laptop it self and
    Message 1 of 10 , Mar 30, 2009
    • 0 Attachment
      Bob and Tony,

      My original tests were made using a USB and line-in tap on my docking
      station.

      Tonight I moved to a usb/line-in pair on the laptop it'self and the
      difference is HUGE...

      S9 from the XG2 vs. noise floor (just sliding the passband over a bit)
      results:

      Dock DIRECT Improvement
      ----------------------------------------------
      80m: 45dbm 60dbm +15dbm
      40m: 40 54 +14dbm
      20m: 43 56 +13dbm

      I am shocked to see such a huge difference.

      Based on these results, I took some quick measurements on the external
      USB vs internal, and the external line-in vs. internal. About 10dbm of
      the improvement is from the line-in, and 5dbm is from the USB.

      And that was quite a nice surprise...

      73/jeff/ac0c

      k8ki/Bob wrote:
      >
      > I don’t think there’s a lot of hope in expecting good sensitivity if
      > the supply voltage is lowered. The FST3253 needs 5v to run properly.
      > It was designed for reliable digital service, and is a fairly noisy RF
      > device. I do think that 2N3904 could instead be used as a
      > preamplifier—it’s quite capable of achieving a 5 dB noise figure
      > throughout the hf region.
      >
      > Bob/k8ki
      >
      > * From: * softrock40@yahoogro ups.com [mailto: softrock40@yahoogro
      > ups.com ] *On Behalf Of *Tony Parks
      > *Sent:* Monday, March 30, 2009 12:17 PM
      > *To:* softrock40@yahoogro ups.com
      > *Subject:* Re: [softrock40] USB power for SRv9? - conclusions
      >
      > It would be interesting to try an NPN transistor in the 5 volt USB
      > power line to the v9.0 receiver 5 volt bus. The collector of the NPN
      > transistor would connect to the USB 5 volt line. A 10 ohm resistor
      > from the USB 5 volt line would connect to the base of the NPN
      > transistor. From the base of the NPN transistor to circuit grround
      > would be a 47 uF cap and a 0.1 uF cap. The emitter of the NPN
      > transistor would then feed the 5 volt bus on the v9.0 board.
      >
      > The power loss in the NPN transistor would be less than 100 mW so a
      > transistor such as a 2N3904 could be used.
      >
      > 73,
      >
      > Tony Parks
      > kb9yig@gmail. com <mailto:kb9yig@...>
      >
      >
      >
      > On Mon, 2009-03-30 at 11:58 -0500, Jeff Blaine wrote:
      >
      >> Jeff Blaine wrote:
      >> > Conclusions and results following comments from the group and some
      >> > bench time...
      >> >
      >> > Before modification, benchmarking a S9 level (XG2) and the noise
      >> > floor, and making a simple subtraction to establish the noise floor,
      >> > my laptop internal sound card and the v9 + BPF had about a 40-44 db SN
      >> > difference.
      >> >
      >> > 1. At a 5.0V usb power input, the performance of the rig is hurt most
      >> > by the voltage drop. It's a bigger impact than the noise pickup from
      >> > the PC side.
      >> >
      >> > The diode and the regulator drop-out just kill the noise floor. To
      >> > make the best of this, I removed the diode and replaced the 7805 with
      >> > a LD1084 adj set for 4.25v terminal voltage. With 5.06v in, the rig
      >> > saw about 3.94v (usb voltage a bit low for the regulator to work
      >> > properly).
      >> >
      >> > Under this condition, the SN difference was:
      >> >
      >> > 80m: 37db
      >> > 40m: 27db
      >> > 20m: 22db
      >> >
      >> > Conclusion: The combination of lower voltage + USB noise cost about
      >> 20db!
      >> >
      >> > 2. To establish the contribution of the USB vs. the lower voltage, the
      >> > regulator input was moved over to a battery source. 6.37v in, 4.23v
      >> > out (meaning the regulator also had enough voltage to work properly).
      >> >
      >> > 80m: 35db
      >> > 40m: 31db
      >> > 20m: 35db
      >> >
      >> > Conclusion: The slightly higher voltage + battery source improved the
      >> > noise floor by 2-13db. Higher frequency, more benefit.
      >> >
      >> > 3. Resetting the regulator voltage to 5.34v, under battery power:
      >> >
      >> > 80m:: 45db
      >> > 40m:: 40db
      >> > 20m: 43db
      >> >
      >> > Conclusion: An increase in Vcc from 4.23 --> 5.34 improves the noise
      >> > floor by 5-10db. Case is similar to the original regulator & external
      >> > supply used as a baseline.
      >> >
      >> > 4. Removed regulator. Add L/C networks to USB supply lines. Liberal
      >> > use of 0.1uf chip caps. Now USB powered:
      >> >
      >> > 80m: 44db
      >> > 40m: 39db
      >> > 20m: 41db
      >> >
      >> > Conclusion: USB power with some filtering effort has brought the USB
      >> > vs. external battery performance to within a 2db noise floor increase
      >> > of battery power alone.
      >> >
      >> > Given the intention of this rig (travel use), the 2db is a reasonable
      >> > trade-off considering the advantage of not having an external power
      >> > source to contend with.
      >> >
      >> > 73/jeff/ac0c
      >> >
      >> >
      >> >
      >> > Jeff Blaine wrote:
      >> >>
      >> >> Has anyone tried to run the SRv9 powered only by the USB port?
      >> >>
      >> >> Meaning - one port to run the attiny - and another port for power only
      >> >> (bypassing the 7805).
      >> >>
      >> >> The SR is normally setup with +12v but the +12v unregulated line is
      >> >> unused presently. And the current drain is well within the 500ma per
      >> >> port USB spec max.
      >> >>
      >> >> 73/jeff/ac0c
      >> >>
      >> >>
      >> >
      >>
      >>
      >>
      > No virus found in this incoming message.
      > Checked by AVG - www.avg.com
      > Version: 8.0.238 / Virus Database: 270.11.31/2029 - Release Date:
      > 03/29/09 16:56:00
      >
      >
    • kazunori miura
      Hello, all I am making SDR with USB power supply. To reject noise from PC, it is needed comon mode filter and ripple filter. The result is here.
      Message 2 of 10 , Mar 31, 2009
      • 0 Attachment
        Hello, all

        I am making SDR with USB power supply.
        To reject noise from PC,
        it is needed comon mode filter and ripple filter.
        The result is here.

        http://zao.jp/media/blogs/b/P1010799.JPG

        But little noise is still left around center frequency.
        these noise is comming from PC via shield of USB cable.

        best 73

        JA7TDO

        Jeff Blaine ????????:
        > Bob and Tony,
        >
        > My original tests were made using a USB and line-in tap on my docking
        > station.
        >
        > Tonight I moved to a usb/line-in pair on the laptop it'self and the
        > difference is HUGE...
        >
        > S9 from the XG2 vs. noise floor (just sliding the passband over a bit)
        > results:
        >
        > Dock DIRECT Improvement
        > ----------------------------------------------
        > 80m: 45dbm 60dbm +15dbm
        > 40m: 40 54 +14dbm
        > 20m: 43 56 +13dbm
        >
        > I am shocked to see such a huge difference.
        >
        > Based on these results, I took some quick measurements on the external
        > USB vs internal, and the external line-in vs. internal. About 10dbm of
        > the improvement is from the line-in, and 5dbm is from the USB.
        >
        > And that was quite a nice surprise...
        >
        > 73/jeff/ac0c
        >
        > k8ki/Bob wrote:
        >> I don’t think there’s a lot of hope in expecting good sensitivity if
        >> the supply voltage is lowered. The FST3253 needs 5v to run properly.
        >> It was designed for reliable digital service, and is a fairly noisy RF
        >> device. I do think that 2N3904 could instead be used as a
        >> preamplifier—it’s quite capable of achieving a 5 dB noise figure
        >> throughout the hf region.
        >>
        >> Bob/k8ki
        >>
        >> * From: * softrock40@yahoogro ups.com [mailto: softrock40@yahoogro
        >> ups.com ] *On Behalf Of *Tony Parks
        >> *Sent:* Monday, March 30, 2009 12:17 PM
        >> *To:* softrock40@yahoogro ups.com
        >> *Subject:* Re: [softrock40] USB power for SRv9? - conclusions
        >>
        >> It would be interesting to try an NPN transistor in the 5 volt USB
        >> power line to the v9.0 receiver 5 volt bus. The collector of the NPN
        >> transistor would connect to the USB 5 volt line. A 10 ohm resistor
        >> from the USB 5 volt line would connect to the base of the NPN
        >> transistor. From the base of the NPN transistor to circuit grround
        >> would be a 47 uF cap and a 0.1 uF cap. The emitter of the NPN
        >> transistor would then feed the 5 volt bus on the v9.0 board.
        >>
        >> The power loss in the NPN transistor would be less than 100 mW so a
        >> transistor such as a 2N3904 could be used.
        >>
        >> 73,
        >>
        >> Tony Parks
        >> kb9yig@gmail. com <mailto:kb9yig@...>
        >>
        >>
        >>
        >> On Mon, 2009-03-30 at 11:58 -0500, Jeff Blaine wrote:
        >>
        >>> Jeff Blaine wrote:
        >>>> Conclusions and results following comments from the group and some
        >>>> bench time...
        >>>>
        >>>> Before modification, benchmarking a S9 level (XG2) and the noise
        >>>> floor, and making a simple subtraction to establish the noise floor,
        >>>> my laptop internal sound card and the v9 + BPF had about a 40-44 db SN
        >>>> difference.
        >>>>
        >>>> 1. At a 5.0V usb power input, the performance of the rig is hurt most
        >>>> by the voltage drop. It's a bigger impact than the noise pickup from
        >>>> the PC side.
        >>>>
        >>>> The diode and the regulator drop-out just kill the noise floor. To
        >>>> make the best of this, I removed the diode and replaced the 7805 with
        >>>> a LD1084 adj set for 4.25v terminal voltage. With 5.06v in, the rig
        >>>> saw about 3.94v (usb voltage a bit low for the regulator to work
        >>>> properly).
        >>>>
        >>>> Under this condition, the SN difference was:
        >>>>
        >>>> 80m: 37db
        >>>> 40m: 27db
        >>>> 20m: 22db
        >>>>
        >>>> Conclusion: The combination of lower voltage + USB noise cost about
        >>> 20db!
        >>>> 2. To establish the contribution of the USB vs. the lower voltage, the
        >>>> regulator input was moved over to a battery source. 6.37v in, 4.23v
        >>>> out (meaning the regulator also had enough voltage to work properly).
        >>>>
        >>>> 80m: 35db
        >>>> 40m: 31db
        >>>> 20m: 35db
        >>>>
        >>>> Conclusion: The slightly higher voltage + battery source improved the
        >>>> noise floor by 2-13db. Higher frequency, more benefit.
        >>>>
        >>>> 3. Resetting the regulator voltage to 5.34v, under battery power:
        >>>>
        >>>> 80m:: 45db
        >>>> 40m:: 40db
        >>>> 20m: 43db
        >>>>
        >>>> Conclusion: An increase in Vcc from 4.23 --> 5.34 improves the noise
        >>>> floor by 5-10db. Case is similar to the original regulator & external
        >>>> supply used as a baseline.
        >>>>
        >>>> 4. Removed regulator. Add L/C networks to USB supply lines. Liberal
        >>>> use of 0.1uf chip caps. Now USB powered:
        >>>>
        >>>> 80m: 44db
        >>>> 40m: 39db
        >>>> 20m: 41db
        >>>>
        >>>> Conclusion: USB power with some filtering effort has brought the USB
        >>>> vs. external battery performance to within a 2db noise floor increase
        >>>> of battery power alone.
        >>>>
        >>>> Given the intention of this rig (travel use), the 2db is a reasonable
        >>>> trade-off considering the advantage of not having an external power
        >>>> source to contend with.
        >>>>
        >>>> 73/jeff/ac0c
        >>>>
        >>>>
        >>>>
        >>>> Jeff Blaine wrote:
        >>>>> Has anyone tried to run the SRv9 powered only by the USB port?
        >>>>>
        >>>>> Meaning - one port to run the attiny - and another port for power only
        >>>>> (bypassing the 7805).
        >>>>>
        >>>>> The SR is normally setup with +12v but the +12v unregulated line is
        >>>>> unused presently. And the current drain is well within the 500ma per
        >>>>> port USB spec max.
        >>>>>
        >>>>> 73/jeff/ac0c
        >>>>>
        >>>>>
        >>>
        >>>
        >> No virus found in this incoming message.
        >> Checked by AVG - www.avg.com
        >> Version: 8.0.238 / Virus Database: 270.11.31/2029 - Release Date:
        >> 03/29/09 16:56:00
        >>
        >>
        >
        >
        >
        > ------------------------------------
        >
        > Yahoo! Groups Links
        >
        >
        >
        >
      Your message has been successfully submitted and would be delivered to recipients shortly.