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

Re: [softrock40] Re: Interesting SDR project funded by DARPA

Expand Messages
  • Sid Boyce
    That s getting expensive of course, like HiQSDR or Hermes with an embedded backend processor, then we are looking at high end SDR s. I have not looked at how
    Message 1 of 36 , Oct 24, 2012
    View Source
    • 0 Attachment
      That's getting expensive of course, like HiQSDR or Hermes with an embedded backend processor, then we are looking at high end SDR's.

      I have not looked at how much could be programmed into the FPGA, needs further study.

      With the simple SDR, it's a limited range receiver with a relatively slow clock. When I looked at it in QEX I wondered if the range couldn't be extended with a 25MHz clock and also if the 24 UDB's of the CY8C38 would be sufficient to also accommodate TX code. I can't remember which version of the CY8C38 was used in the Simple SDR.

      I have only dealt with the Altera Cyclone III FPGA and had a read of the specs of their SoC FPGA's which look good for a future DDC/DUC standalone SDR.
      73 ... Sid.
       
      On 25/10/12 00:45, Rick Simpson wrote:
       

      Sid,
       
      The Adapteva seems to focus on being an ultra-high-speed processor with some programmable logic. For SDR I would think you would want a chip with ultra-high-speed ADC, moderate-speed DAC, some programmable-gain analog amplifiers, plenty of processing power, one or more programmable high-speed oscillators, and logic for the mixer. A start towards that is the Cypress CY8C38 used in the "Simple SDR Receiver" from Simple Circuits Inc.
       
      In other words, we need a chip with both analog and digital circuitry specifically aimed at SDR, not a general purpose digital-only chip.
       
      Rick
       
      ----- Original Message -----
      From: Sid Boyce
      Sent: Wednesday, October 24, 2012 3:32 PM
      Subject: Re: [softrock40] Re: Interesting SDR project funded by DARPA

       

      Hi Rick,
      You can start off on this page.
      http://www.kickstarter.com/projects/adapteva/parallella-a-supercomputer-for-everyone

      The $99.00 promise is the 16-core version, higher subscriptions will get you a 64-core board.
      I hope they reach their funding target though the date is getting very close.
      They haven't mentioned anything about extending the target date which would be a good idea as it's been running only a short time.

      I am always on the lookout for a small form factor with lots of processing power for SDR use. The ODROID-X performs well.

      Another board with lots of I/O ports that's not yet released but expected to be priced around $200.00
      http://www.cnx-software.com/2012/02/24/low-cost-freescale-imx6-quad-sabre-lite-development-board/
      73 ... Sid.

      On 24/10/12 19:01, Rick Simpson wrote:
       
      Sid, just to let you know I enjoy your additions to this forum. You seem to stay on the cutting edge of SDR, a location I am most interested in, even if just as a spectator. What does your $99 buy you and what is the URL that explains this?
       
      Rick
      K3IND

       
      ----- Original Message -----
      From: Sid Boyce
      Sent: Wednesday, October 24, 2012 11:17 AM
      Subject: Re: [softrock40] Re: Interesting SDR project funded by DARPA

       
      Another interesting project is parallela.
      I subscribed $99.00 to the project last week. Details on adapteva.com.
      http://www.adapteva.com/white-papers/sdr-and-cognitive-radio-on-parallella/
      73 ... Sid.

      On 23/10/12 23:07, m0fmt wrote:
       

      Hi all

      The 8 bit ADC seems to work ok on RX applications using the DVB-T USB stick with HDSDR application, subjectively. It is on the cusp with clarity of the waterfall presentation of cw but audibly it sounds fine. I up convert to HF (probably missing the point at this stage)and to see the whole HF spectrum as an endless ribon is a joy but with a 1meg screens worth of signals is impossible so you have to ride the zoom control. We are not talking TX here only RX.

      Will follow this thread with interest.

      73 Pete m0fmt

      --- In softrock40@yahoogroups.com, Sylvain AZARIAN <sylvain.azarian@...> wrote:
      >
      > Hi,
      >
      > Decimation sounds to be a good option but in fact it requires a huge amount
      > of processing to remove with enough attenuation unwanted signals (low-pass
      > filtering + decimation). It means having a long set of coefficients for
      > DSP, and this processing must be done at full speed (at the rate of the
      > incoming samples). This typically requires a carefully written code + a
      > multi core machine...
      >
      > The good thing is to be able to see the full band but... apart from the 350
      > Khz we have in the 20 meter band... (Ok, this could be useful for VHF/UHF
      > where the bands are not that busy (at least in france); but from my
      > personal experience, not very easy to identify a QSO on a 2MHz wide
      > waterfall...)
      >
      > 73
      > F4GKR
      >
      > 2012/10/23 David Turnbull <dturnbull@...>
      >
      > > **
      > >
      > >
      > > Alex,
      > >
      > > I asked if hams would be interested and what I posted is the answer I got.
      > > The image is a problem in transmit because it's a spurious emission and in
      > > receive because, for example, a CW contest becomes a maddening problem. The
      > > dynamic range is also a problem as you describe but you can't get more with
      > > decimation trickery. Maybe someone will figure out some hardware hacks
      > > though.
      > >
      > > 73 David AE9RB
      > > http://AE9RB.com/
      > >
      > >
      > > On Mon, Oct 22, 2012 at 4:45 PM, Alex Turcu <talex004@...> wrote:
      > >
      > >>
      > >>
      > >> David,
      > >> Wouldn't oversampling help in this case? Given the ADC can do
      > >> 8bit/20msps, wouldn't that be roughly equivalent to 16bit/78ksps ? Also,
      > >> shouldn't the problem be due to dynamic range, instead of image
      > >> rejection (stronger stations would make weaker stations inaudible,
      > >> regardless of where they are with respect to the center frequency)? I just
      > >> want to make sure I understand all these notions right.
      > >> Thank you,
      > >> Alex
      > >>
      > >>
      > >> On Mon, Oct 22, 2012 at 12:17 PM, David Turnbull <dturnbull@...>wrote:
      > >>
      > >>> **
      > >>>
      > >>>
      > >>> I know Jared from this project and will be seeing him tonight.
      > >>> Unfortunately, he's told me this radio won't be good for hams. The ADC
      > >>> is 8-bit which doesn't offer much image rejection. Since it has huge
      > >>> bandwidth, this really isn't a problem for research or ISM use.
      > >>>
      > >>> 73 David AE9RB
      > >>> http://AE9RB.com/
      > >>>
      > >>>
      > >>> On Mon, Oct 22, 2012 at 8:37 AM, shaynal01 <steve@...>
      > >>> wrote:
      > >>> >
      > >>> http://www.forbes.com/sites/andygreenberg/2012/10/19/darpa-funded-radio-hackrf-aims-to-be-a-300-wireless-swiss-army-knife-for-hackers/
      > >>> >
      > >>> > http://ossmann.blogspot.com/2012/06/introducing-hackrf.html
      > >>> >
      > >>> >
      > >>>
      > >>>
      > >>
      > >>
      > >>
      > >
      > >
      >





      -- 
      Sid Boyce ... Hamradio License G3VBV, Licensed Private Pilot
      Emeritus IBM/Amdahl Mainframes and Sun/Fujitsu Servers Tech Support
      Senior Staff Specialist, Cricket Coach
      Microsoft Windows Free Zone - Linux used for all Computing Tasks
      
  • Sid Boyce
    They must get a real good deal as the Origen board starts at $199.
    Message 36 of 36 , Oct 29, 2012
    View Source
    • 0 Attachment
      They must get a real good deal as the Origen board starts at $199.
      http://www.samsung.com/global/business/semiconductor/minisite/Exynos/platform_partners_platform.html#Arndale-Board
      73 ... Sid.

      On 29/10/12 22:02, Alex Turcu wrote:
       

      ODROID-X is indeed an impressive platform. However at $129 it seems subsidized (the CPU module itself sells for $119 and is out of stock), but as with the BeagleBone, I like the modularity!

      Alex

      On Mon, Oct 29, 2012 at 5:37 PM, Sid Boyce <sboyce@...> wrote:
       

      I gave up on the Pi largely because of the USB load problems and the tussles I had with trying unsuccessfully to get sound working.

      I have replaced it with the ODROID-X which is running hiqsdr-server, dspserver and QtRadio happily.
      73 ... Sid.

       
      On 29/10/12 20:25, Shirley Márquez Dúlcey wrote:
       

      > Cheap as it is, I dislike the RPi for its closed SoC, and I seriously doubt it has the horsepower for SDR (last time I checked I had no confirmation it would work). While it does have a GPU that may have enough horsepower, > it lacks OpenCL support and Broadcom don't seem to eager to implement it anytime soon (nor release specs for the open-source community to implement)


      The closed-source issue is no longer true as of a few days ago: http://www.raspberrypi.org/archives/2221

      The lack of computing power remains. I don't know whether the GPU in the Pi has enough flexibility to implement meaningful OpenCL support... though with the driver source now available it's something the community can investigate.

      Broadcom has historically been a problem for the open-source community, notably its WiFi interfaces. Kudos to them for opening up the source for the SoC in the Pi.



      -- 
      Sid Boyce ... Hamradio License G3VBV, Licensed Private Pilot
      Emeritus IBM/Amdahl Mainframes and Sun/Fujitsu Servers Tech Support
      Senior Staff Specialist, Cricket Coach
      Microsoft Windows Free Zone - Linux used for all Computing Tasks
      



      -- 
      Sid Boyce ... Hamradio License G3VBV, Licensed Private Pilot
      Emeritus IBM/Amdahl Mainframes and Sun/Fujitsu Servers Tech Support
      Senior Staff Specialist, Cricket Coach
      Microsoft Windows Free Zone - Linux used for all Computing Tasks
      
    • Your message has been successfully submitted and would be delivered to recipients shortly.