[hpsdr] Interesting email on SDR transmit noise

[Shirley Márquez Dúlcey]

SDR transmitters usually only have low-pass filters. The LPF will
limit the amount of broadband noise above the band of operation, but
it will do nothing about broadband noise in or below the operating
band.

There are two flaws in an SDR that can cause noise emission. Phase
noise in the local oscillator is usually the most important one in
non-SDR designs. In most current radios this is either a PLL-based
synthesizer or a DDS IC, In a DDS synthesizer, noise will depend both
on the DDS chip used and in the phase noise level of the reference
oscillator. Some DDS chips have an on-chip PLL clock multiplier; using
the multiplier should usually be avoided for best performance. Phase
noise is a problem in both SDR and non-SDR transmitters.

The second cause is noise and distortion in the audio signal used to
modulate the transmitter. That is an issue with an SDR transmitter in
any mode, and can cause noise much farther away from the transmit
frequency than in a typical non-SDR design. In a conventional
filter-based transmitter, most noise products are limited by the
passband of the SSB filter used in the transmitter. The linked article
suggests that the problem in the FlexRadio designs is in the audio
section rather than the LO.

I haven't seen similar testing of other SDR designs. Noise performance
of a design that uses an audio interface and a quadrature sampling
modulator, like most FlexRadio designs and the Softrock transmitters,
will depend heavily on the quality of the sound card or IC used for
transmit, as well as the quality of the clock signals to both the
sound interface and the modulator. FlexRadio clearly got some part of
that wrong in the 1500 and 5000, and it probably isn't quality of the
audio modulating signal because the bad noise sidebands remain even if
there is no signal. (It could, however, be the result of a noisy audio
amplifier that sends significant amounts of broadband noise to the
modulator even when nothing is being transmitted.) The article also
shows that they do power reduction by reducing the level of audio
drive to the modulator rather than turning down the gain of the
amplifier. Noise performance of a transmitter that uses a radio
frequency DAC depends on the cleanliness of the signal fed to the DAC,
the performance of the DAC, and the quality of the clock signal used
to drive the DAC.

On Fri, Oct 18, 2013 at 11:42 PM, KA5FPT Paul <ka5fpt at tx.rr.com> wrote:
> ***** High Performance Software Defined Radio Discussion List *****
>
> I saw the following on the Flexradio mail list. I am having some problems
> understanding how a SDR transmitter can cause wide band noise. I would have
> thought that with proper filtering it would be controlled.
> -----------------------------------------------------------------------------------------------
>
> Date: Wed, 16 Oct 2013 14:59:37 -0400
> From: "Gedas"<w8bya at mchsi.com>
> To:<flexradio at flex-radio.biz>
> Subject: [Flexradio] Noise Sidebands, Article by SM5BSZ
> Message-ID: <399ADD464530473F88C1CBF0659768F5 at biggy>
> Content-Type: text/plain;       charset="utf-8"
>
> I came across this interesting article by SM5BSZ who discusses some possible
> transmitter issues with many radios, including the Flex.  It has me a bit
> concerned.  I am wondering if anyone has seen similar data for the Flex
> 5000? The article may be found here:
>
> http://www.sm5bsz.com/dynrange/dubus313.pdf
>
> Gedas, W8BYA
> Gallery athttp://w8bya.com
> Light travels faster than sound....
> This is why some people appear bright until you hear them speak.
>
> --------------------------------------------------------------------------------------------------
>
> 73,
> Paul Cecil
> KA5FPT
> _______________________________________________
> HPSDR Discussion List
> To post msg: hpsdr at openhpsdr.org
> Subscription help:
> http://lists.openhpsdr.org/listinfo.cgi/hpsdr-openhpsdr.org
> HPSDR web page: http://openhpsdr.org
> Archives: http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/

 1382160106.0