[time-nuts] WWVB SDR discussion

[Mark Haun]

No more so than for a GPSDO using a microcontroller.  In this case, the
MCU would steer an external oscillator using an appropriately long time
constant, based on a phase tracking loop.  In my tentative block
diagram, the oscillator is a cheap 26 MHz OCXO which supplies the MCU
clock and (as a byproduct) the internal ADC sample rate.  The basic plan
for an STM32L4 was something like

26 MHz clock input from OCXO
78 MHz MCU clock (x3 via internal PLL)
3.12 MHz ADC sample rate (78 MHz / 25)
13x downsample to 240 kHz using DFSDM block (hardware CIC)
sign flip and deinterleave into I and Q channels (I, Q, -I, -Q, ...)
downsample again, to ~ 10 Hz or so; do SDR stuff at this rate
carrier tracking loop runs back outside the MCU to the OCXO control
input, via PWM DAC

With care, the internal ADC gives you about 11 good bits at the high
sampling rate.  Assuming white noise (questionable I suppose), this
yields something like 18 noise-free bits at the low rate where the phase
tracking and decoding happens.

Prerequisite is an analog input centered around 60 kHz with less than
3.6Vp-p amplitude.  There are no particular bandwidth requirements
(Nyquist is 1.5 MHz!) as long as it is narrow enough to avoid clipping
from strong interfering signals.  I am having a hard time with this
part, unfortunately.  My tuned loop seems still too broadband, even
after a couple more poles of op-amp filter.  I have a bunch of 60-kHz
tuning-fork crystals and wanted to try a crystal filter like the "pros"
do in cheap consumer clocks, but I can't figure out a circuit to do this
(weak on analog).  So if anyone wants to tackle that and send me a
working design, I'm happy to do the easy stuff, i.e. the DSP/SDR parts ;)

Mark

On 10-Aug-20 11:45 AM, paul swed wrote:
> Hello to the group.
> Have been looking forward to seeing how the STM32 SDR project might be
> going.
> SDR is a weak spot for me. So been reading. And believe the answer is that
> a SDR solution may work for AM code and even BPSK code to an extent. But
> doesn't the sampling destroy the quality of the incoming signal for
> establishing a locked reference?
> Regards
> Paul
> WB8TSL
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe, go to http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
> and follow the instructions there.
>