Second build sold out. Message will appear here when store is ready for third build ordering.

Auto select comp/de emphasis based on mode and frequency

I've got a few people interested in the quality of AM using my kiwi at a wider audio bandwidth, pointing them to a local station.

However I have to explain to them to turn the COMP off and the de emphasis on.


Is it possible this combination could engage if AM is selected perhaps 520-1700khz. Or perhaps if I copy the URL then these settings could be copied too?

These two settings default behaviour don't seem to be entirely correct for AM broadcast reception however where excellent quality audio is possible and could really get people into AM broadcasting.


Many thanks!

Comments

  • It might be better to add a third AM mode, e.g. "AMW" (AM-AMN-AMW) that does this. What should the bandwidth be in this case? It is 9800 Hz for AM now.

  • edited January 2022

    I like the idea of a wideband AM mode.

    The old wideband AM-stereo receivers such as the Carver TX-11a, Sony ST-JX430A, etc... had AM BWs up to ±15kHz with 9/10kHz whistle notch filters. With the adoption NRSC in the US, stations are now limited to ±10.2kHz analog audio BW (Canada being ±12.5kHz).

    On my Kiwi, I typically have DX labels for local music stations set to open the BW up to ±9.6kHz (19.2kHz), which keeps out the 10kHz het and splatter from weak adjacent stations low.

    Looking at the AMAX certification requirements for AM-stereo receivers (a standard proposed and briefly pushed in the early/mid 90s), Denon TU-680NAB, Sony SRF-42...

    • The receiver has to have a minimum of ±7.5kHz BW in wideband mode for home/auto receivers (±6.5kHz for portables)

    • Follow the proper NRSC-1 audio de-emphasis curve

    • Attenuate the 10kHz het whistle, etc....

    I'd suggest the default BW for the wideband mode be set low enough that you don't hear the 9/10kHz het, so maybe ±8.75kHz (17.5kHz)? Keeping in mind many Kiwis are in use where 9kHz spacing is used. I assume the function would be disabled for Kiwis not configured for 3 channels, or default to a maximum of ±6kHz for Kiwis in 4+ channel mode?

    I have a question I've been meaning to ask: What is the current de-emphasis of the de-emphasis function?

    https://www.nrscstandards.org/standards-and-guidelines/documents/standards/nrsc-1-c.pdf

    Not the hijack the thread from the OP, I just like the suggestion.

  • There is also a difference between de-emphasis values of 50 and 75uS values which are used for broadcast services, and are frequently misquoted on the web in relation to NBFM systems.

    NBFM (e.g. amateur radio and CB on the HF bands and also commercial 2 way radio on the higher bands) uses 6dB/oct roll off above 300Hz, which would probably equate to a value of somewhere around 500uS, if it was expressed in that manner.

    Previous thread on the subject for those interested.

    http://forum.kiwisdr.com/index.php?p=/discussion/comment/6677#Comment_6677

    Regards,

    Martin

  • Hi John,

    Hope you are well.

    With the recent improvements in propagation conditions, I've been doing a lot of listening to NBFM signals on the upper HF bands.

    I've also been playing with SDRangel to control and stream I/Q from the KiWi and then use SDRangel to provide local demodulation of NBFM and other modes.

    https://www.sdrangel.org/

    Comparing the two, the locally demodulated NBFM sounds much 'smoother' than the audio produced by the KiWi.

    Part of this is due to the value of de-emphasis that has been applied, but there is still an underlying 'scratchyness' and lack of recovered audio from low deviation signals, that doesn't seem to be present in the locally demodulated signals.

    If you ever feel like taking a look at this sometime, I'd be most appreciative.

    Best wishes,

    Martin

  • Well, I would need someone to give me a new algorithm to try.

    The current NBFM demodulator was implemented many years ago. I first tried several well-known demods that should have worked. But they didn't for some reason. The demod from the CSDR package worked and is what we currently use. I have no idea how it works or why it might have limitations. That particular aspect of DSP is beyond me..

Sign In or Register to comment.