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

Re: [SpectrumLabUsers] UDP standard def

Expand Messages
  • wolf_dl4yhf
    Hello Henry, Looks like a very advanced project for homebrew - and even though I have some experience with handcrafted SMD boards, I wouldn t try my luck at
    Message 1 of 5 , Feb 20, 2011
      Hello Henry,

      Looks like a very advanced project for 'homebrew' - and even though I have some experience with handcrafted SMD boards, I wouldn't try my luck at soldering an FPGA. That's why I will stick to the three SDRs I have, and the various soundcards which give me up to 192 kSamples/second.
      As a 'real' radio I still prefer something which operates without a computer, but can be controlled with a nice big VFO know, a friendly digital frequency readout, and a few pushbuttons on the front to select the mode ;-)

      About the Ethernet interface: With the lower sampling rates described on James' website, SL could be work. At half the netto troughput of a 100 MBit Ethernet port: NO CHANCE. I don't know how Quisk processes the data at 960000 samples / second, with 48 bits per sample, it's an awfully large CPU load which calls for an extremely optimized code to handle this amount of data. It may be ok for a dedicated software defined radio, but for the versatile processing blocks in Spectrum Lab this is a bit too much. Continuous(!) processing such a stream with 50 percent overlapping, large, FFTs would be asking too much. I will not try this myself.

      All the best,
        Wolf .


      Am 20.02.2011 18:09, schrieb ehydra:
       

      Hi Wolf -

      I'm interested in this SDR project:
      http://james.ahlstrom.name/transceiver/index.html

      Here is a german link:
      http://www.mikrocontroller.net/topic/197747#1938792

      In the first link is a proposed and realized UDP definition.

      Is it possible to follow this?

      I'm still not final decided if I make my own ADC board or just use one
      of the SDR projects.

      - Henry

      --
      ehydra.dyndns.info


    • ehydra
      Hello Wolf - Thanks for responding! Well, my job was btw. to solder FPGAs by hand for repair. Can be done will a little experience. Or let it done by
      Message 2 of 5 , Feb 20, 2011
        Hello Wolf -

        Thanks for responding!

        Well, my <older> job was btw. to solder FPGAs by hand for repair. Can be
        done will a little experience. Or let it done by others.

        My question is simply, if you're willing to add software support for
        this UDP frames?

        As of the bandwidth: This board uses a FPGA decimator to reduce
        bandwidth. As you know: Dividing bandwidth by two adds just one bit to
        the data width.

        What do you think about it?
        A board soldering for you for the software addition?


        - Henry


        wolf_dl4yhf schrieb:
        > Hello Henry,
        >
        > Looks like a very advanced project for 'homebrew' - and even though I
        > have some experience with handcrafted SMD boards, I wouldn't try my luck
        > at soldering an FPGA. That's why I will stick to the three SDRs I have,
        > and the various soundcards which give me up to 192 kSamples/second.
        > As a 'real' radio I still prefer something which operates without a
        > computer, but can be controlled with a nice big VFO know, a friendly
        > digital frequency readout, and a few pushbuttons on the front to select
        > the mode ;-)
        >
        > About the Ethernet interface: With the lower sampling rates described on
        > James' website, SL could be work. At half the netto troughput of a 100
        > MBit Ethernet port: NO CHANCE. I don't know how Quisk processes the data
        > at 960000 samples / second, with 48 bits per sample, it's an awfully
        > large CPU load which calls for an extremely optimized code to handle
        > this amount of data. It may be ok for a dedicated software defined
        > radio, but for the versatile processing blocks in Spectrum Lab this is a
        > bit too much. Continuous(!) processing such a stream with 50 percent
        > overlapping, large, FFTs would be asking too much. I will not try this
        > myself.
        >
        > All the best,
        > Wolf .
        >
        >
        > Am 20.02.2011 18:09, schrieb ehydra:
        >>
        >> Hi Wolf -
        >>
        >> I'm interested in this SDR project:
        >> http://james.ahlstrom.name/transceiver/index.html
        >>
        >> Here is a german link:
        >> http://www.mikrocontroller.net/topic/197747#1938792
        >>
        >> In the first link is a proposed and realized UDP definition.
        >>
        >> Is it possible to follow this?
        >>
        >> I'm still not final decided if I make my own ADC board or just use one
        >> of the SDR projects.
        >>
        >> - Henry
        >>
        >> --
        >> ehydra.dyndns.info
        >>
        >>
        >
        >

        --
        ehydra.dyndns.info
      • wolf_dl4yhf
        Hello Henry, Thanks for the offer, but I don t need yet another SDR... even if that project may be interesting. I could add support for this special UDP frame,
        Message 3 of 5 , Feb 20, 2011
          Hello Henry,

          Thanks for the offer, but I don't need yet another SDR... even if that project may be interesting.
          I could add support for this special UDP frame, but, to be honest, I don't want to, because if I started doing that, I wouldn't do much else in my rare sparetime besides adding support for other radios.
          I have done this for SDR-IQ, Perseus, and that's enough.
          What I can do is provide a simple skeleton for a "driver DLL" written in C, which receives the UDP frames, converts the data into a structure suitable for Spectrum Lab.

          Sorry, there's not much more I can do, because I know how time-consuming it *can* be to add support for some exotic hardware... especially when it's at an early development step.

          Bye,
            Wolf .


          Am 20.02.2011 19:05, schrieb ehydra:
           

          Hello Wolf -

          Thanks for responding!

          Well, my <older> job was btw. to solder FPGAs by hand for repair. Can be
          done will a little experience. Or let it done by others.

          My question is simply, if you're willing to add software support for
          this UDP frames?

          As of the bandwidth: This board uses a FPGA decimator to reduce
          bandwidth. As you know: Dividing bandwidth by two adds just one bit to
          the data width.

          What do you think about it?
          A board soldering for you for the software addition?

          - Henry

          wolf_dl4yhf schrieb:
          > Hello Henry,
          >
          > Looks like a very advanced project for 'homebrew' - and even though I
          > have some experience with handcrafted SMD boards, I wouldn't try my luck
          > at soldering an FPGA. That's why I will stick to the three SDRs I have,
          > and the various soundcards which give me up to 192 kSamples/second.
          > As a 'real' radio I still prefer something which operates without a
          > computer, but can be controlled with a nice big VFO know, a friendly
          > digital frequency readout, and a few pushbuttons on the front to select
          > the mode ;-)
          >
          > About the Ethernet interface: With the lower sampling rates described on
          > James' website, SL could be work. At half the netto troughput of a 100
          > MBit Ethernet port: NO CHANCE. I don't know how Quisk processes the data
          > at 960000 samples / second, with 48 bits per sample, it's an awfully
          > large CPU load which calls for an extremely optimized code to handle
          > this amount of data. It may be ok for a dedicated software defined
          > radio, but for the versatile processing blocks in Spectrum Lab this is a
          > bit too much. Continuous(!) processing such a stream with 50 percent
          > overlapping, large, FFTs would be asking too much. I will not try this
          > myself.
          >
          > All the best,
          > Wolf .
          >
          >
          > Am 20.02.2011 18:09, schrieb ehydra:
          >>
          >> Hi Wolf -
          >>
          >> I'm interested in this SDR project:
          >> http://james.ahlstrom.name/transceiver/index.html
          >>
          >> Here is a german link:
          >> http://www.mikrocontroller.net/topic/197747#1938792
          >>
          >> In the first link is a proposed and realized UDP definition.
          >>
          >> Is it possible to follow this?
          >>
          >> I'm still not final decided if I make my own ADC board or just use one
          >> of the SDR projects.
          >>
          >> - Henry
          >>
          >> --
          >> ehydra.dyndns.info
          >>
          >>
          >
          >

          --
          ehydra.dyndns.info


        • ehydra
          Hello Wolf - Yes, I understand. Let us freeze this discussion at the moment. Writing PC software is not what I want. When, then I write software for embedded
          Message 4 of 5 , Feb 20, 2011
            Hello Wolf -

            Yes, I understand.

            Let us freeze this discussion at the moment.

            Writing PC software is not what I want. When, then I write software for
            embedded devices on a small scale.



            BTW:
            Did you built this low-input impedance antenna amplifier by Northolt? I
            thought about the circuit but to me it is not clear where the benefit
            pops up.
            Another problem is loop-stability. Not every interesting concept will
            work in reality.


            cheers -
            Henry



            --
            ehydra.dyndns.info



            wolf_dl4yhf schrieb:
            > Hello Henry,
            >
            > Thanks for the offer, but I don't need yet another SDR... even if that
            > project may be interesting.
            > I could add support for this special UDP frame, but, to be honest, I
            > don't want to, because if I started doing that, I wouldn't do much else
            > in my rare sparetime besides adding support for other radios.
            > I have done this for SDR-IQ, Perseus, and that's enough.
            > What I can do is provide a simple skeleton for a "driver DLL" written in
            > C, which receives the UDP frames, converts the data into a structure
            > suitable for Spectrum Lab.
            >
            > Sorry, there's not much more I can do, because I know how time-consuming
            > it *can* be to add support for some exotic hardware... especially when
            > it's at an early development step.
            >
            > Bye,
            > Wolf .
          Your message has been successfully submitted and would be delivered to recipients shortly.