[time-nuts] Near-perfect chip for Loran-C frequency receiver
magnus at rubidium.dyndns.org
Thu Jul 3 17:03:45 EDT 2008
From: "Poul-Henning Kamp" <phk at phk.freebsd.dk>
Subject: [time-nuts] Near-perfect chip for Loran-C frequency receiver
Date: Thu, 03 Jul 2008 19:53:41 +0000
Message-ID: <14904.1215114821 at critter.freebsd.dk>
> ST microelectronics have made a new ARM chip: ST32M* and it almost
> fits the bill for a single-chip Loran-C frequency receiver: it has
> a 1 million samples per second A/D converter.
> Olimex has an eval-board costing EUR 46:
> (Check sparkfun for US pricing)
> Unfortunately, the clock-circuit in the chip prevents use of the
> USB port at 1Msps A/D sampling rate, or if your input clock is
> 10MHz, but it has a serial port as well.
> I have played with various low-cpu/low-ram reception scenarios as
> part of my SDR experiements, and I think we can fit it comfortably
> into the available flash/ram of the chip.
> I wrote 280 lines of simulation code in C yesterday, and using my
> spectrum recording from the 2006-01-01 leapsecond, I was able to
> detect the existence of signals as far away as GRI 8830, 7980 and
> 7270 in just 2 minutes integration time and get good clean signals
> from 7499, 6731, 9007, 8000 and 7001.
> Anybody interested in doing a little group-project to complement
> all the disciplined GPS oscillators we've see recently ?
Hmm. The GNSS Sampler uses a very nice USB chip which would allow a very
minimal wireup from a A/D chip and then you have a pipe of samples straight
into the computer. I think they picked that idea up from the Software Receiver
project (somebody please correct me on the details and links, I'm in low
bitrate country, the summerhouse).
It would be nice to have LORAN-C, MSF and DCF-77 in the same solution if
possible. I am sure some of the US signals can be included.
More information about the time-nuts