K3S USB audio questions
Locked
 
|
A few questions about the K3S internal audio interface:
1. Is the internal connection between the USB audio interface and the 2nd IF analogue or digital? If digital, how is the sample rate conversion implemented? 2. It seems to support sampling rates of 32000, 44100 and 48000 Hz, but not higher or lower values. Is this correct? 3. What determines the clock frequency for this sampling? Is it a free-running crystal? If so, what is its frequency tolerance? Or is it locked to the main TXCO and/or external 10 MHz reference? 4. When listening to SSB audio at S9+, with slow AGC, the output level via this interface seems to be -22 dBFS true peak. Why is it throwing away three bits of resolution? Or is this amount of headroom necessary/useful? TIA. -- Richard Lamont G4DYA <[hidden email]> ______________________________________________________________ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:[hidden email] This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html Message delivered to [hidden email] |
|
|
Hi Richard,
Some answers, below. --matt On 11/5/2016 9:38 AM, Richard Lamont wrote: > A few questions about the K3S internal audio interface: > > 1. Is the internal connection between the USB audio interface and the > 2nd IF analogue or digital? If digital, how is the sample rate > conversion implemented? The KIO3B uses a codec (digital on its USB side, analog one the audio side (line in/out). The codec's USB D+/- nets are connected to a USB hub. > > 2. It seems to support sampling rates of 32000, 44100 and 48000 Hz, but > not higher or lower values. Is this correct? These are the sample rates the codec supports. Yes. > > 3. What determines the clock frequency for this sampling? Is it a > free-running crystal? If so, what is its frequency tolerance? Or is it > locked to the main TXCO and/or external 10 MHz reference? The clock is an XO: 12 MHz, +/- 100 ppm as rated. The XO's output is buffered and feeds the codec. > > 4. When listening to SSB audio at S9+, with slow AGC, the output level > via this interface seems to be -22 dBFS true peak. Why is it throwing > away three bits of resolution? Or is this amount of headroom > necessary/useful? You can find a copy of the data sheet here: http://www.digikey.com/product-search/en?keywords=PCM2901E%2F2K 73, matt W6NIA > > TIA. > -- Always store beer in a dark place. - R. Heinlein Matt Zilmer, W6NIA [Shiraz] ______________________________________________________________ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:[hidden email] This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html Message delivered to [hidden email] |
|
|
Hi,
On item 4: The audio level seen by, say a recording program on the PC depends on the LINE OUT setting on the K3. I left that as the default, 10, and used the Windows mixer to adjust the level. My audio is now reading approximately -6dB below clipping which is probably a little high, but I have set my AGC SLP to the flattest possible setting (highest number). If you prefer a considerable slope on the AGC curve above threshold, you will need a lower LINE OUT setting. Remember that even when throwing away 3 bits, you still have far more resolution than you can use with an HF receiver. AB2TC - Knut
|
|
|
On 05/11/16 18:13, ab2tc wrote:
[ >>> 4. When listening to SSB audio at S9+, with slow AGC, the output level >>> via this interface seems to be -22 dBFS true peak. Why is it throwing >>> away three bits of resolution? Or is this amount of headroom >>> necessary/useful? > The audio level seen by, say a recording program on the PC depends on the > LINE OUT setting on the K3. I left that as the default, 10, and used the > Windows mixer to adjust the level. My audio is now reading approximately > -6dB below clipping which is probably a little high, but I have set my AGC > SLP to the flattest possible setting (highest number). If you prefer a > considerable slope on the AGC curve above threshold, you will need a lower > LINE OUT setting. Remember that even when throwing away 3 bits, you still > have far more resolution than you can use with an HF receiver. Consider JT65+JT9. There might be 30 or more signals within a 4 kHz band, with widely varying levels. If they're all at the same level, then each signal will be about 30 dB below the sum of the signals. If the signal you want is 30 dB weaker than the others, then it will be 60 dB below the sum. The data sheet shows the ADC has an SNR of 89 dB and THD+N of 0.01% (i.e. -80 dB). If the sum of the signals is peaking to -22 dBFS true peak, than that weak JT65/JT9 signal is going to be at -82 dBFS - close to the ADC noise floor. Maybe getting at least some of those three bits back could make quite a difference. -- Richard Lamont G4DYA <[hidden email]> ______________________________________________________________ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:[hidden email] This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html Message delivered to [hidden email] |
|
|
Hi,
You make a good point for some digital modes. But you should be able to make up for most or all of these 3 bits by increasing the LINE OUT as it must surely be before the codec ADC but probably not before the K3's DSP DAC. It's a shame that they have to do it this way (back to back DAC/ADC, but everybody else is doing the same). But by adjusting LINE OUT, the USB audio interface should be no worse than the audio LINE OUT considering the nonlinearity of the transformer in that path. AB2TC - Knut
|
|
|
In reply to this post by Richard Lamont
Hi,
FWIW I've done this. I've increased the gain by 12 dB, i.e. two bits. Incidentally, the scaling on CONFIG:LIN OUT is linear - i.e. the output voltage is proportional to the number. So for +12 dB w.r.t. the default I've increased it from 10 to 40. Richard Lamont G4DYA On 06/11/16 16:48, ab2tc wrote: > Hi, > > You make a good point for some digital modes. But you should be able to make > up for most or all of these 3 bits by increasing the LINE OUT as it must > surely be before the codec ADC but probably not before the K3's DSP DAC. > It's a shame that they have to do it this way (back to back DAC/ADC, but > everybody else is doing the same). But by adjusting LINE OUT, the USB audio > interface should be no worse than the audio LINE OUT considering the > nonlinearity of the transformer in that path. > > AB2TC - Knut > > > > Richard Lamont wrote >> On 05/11/16 18:13, ab2tc wrote: >> [ >>>>> 4. When listening to SSB audio at S9+, with slow AGC, the output level >>>>> via this interface seems to be -22 dBFS true peak. Why is it throwing >>>>> away three bits of resolution? Or is this amount of headroom >>>>> necessary/useful? >> >>> The audio level seen by, say a recording program on the PC depends on the >>> LINE OUT setting on the K3. I left that as the default, 10, and used the >>> Windows mixer to adjust the level. My audio is now reading approximately >>> -6dB below clipping which is probably a little high, but I have set my >>> AGC >>> SLP to the flattest possible setting (highest number). If you prefer a >>> considerable slope on the AGC curve above threshold, you will need a >>> lower >>> LINE OUT setting. Remember that even when throwing away 3 bits, you still >>> have far more resolution than you can use with an HF receiver. >> >> Consider JT65+JT9. There might be 30 or more signals within a 4 kHz >> band, with widely varying levels. If they're all at the same level, then >> each signal will be about 30 dB below the sum of the signals. If the >> signal you want is 30 dB weaker than the others, then it will be 60 dB >> below the sum. >> >> The data sheet shows the ADC has an SNR of 89 dB and THD+N of 0.01% >> (i.e. -80 dB). >> >> If the sum of the signals is peaking to -22 dBFS true peak, than that >> weak JT65/JT9 signal is going to be at -82 dBFS - close to the ADC noise >> floor. Maybe getting at least some of those three bits back could make >> quite a difference. >> >> >> -- >> Richard Lamont G4DYA >> < > >> richard@.me > >> > >> <snip> > > > > > > -- > View this message in context: http://elecraft.365791.n2.nabble.com/K3S-USB-audio-questions-tp7623845p7623862.html > Sent from the Elecraft mailing list archive at Nabble.com. > ______________________________________________________________ > Elecraft mailing list > Home: http://mailman.qth.net/mailman/listinfo/elecraft > Help: http://mailman.qth.net/mmfaq.htm > Post: mailto:[hidden email] > > This list hosted by: http://www.qsl.net > Please help support this email list: http://www.qsl.net/donate.html > Message delivered to [hidden email] > > ______________________________________________________________ Elecraft mailing list Home: http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/mmfaq.htm Post: mailto:[hidden email] This list hosted by: http://www.qsl.net Please help support this email list: http://www.qsl.net/donate.html Message delivered to [hidden email] |
«
Return to Elecraft
|
1 view|%1 views
| Free forum by Nabble | Edit this page |