anelok: revisiting Bluetooth, and the winner is ...

Werner Almesberger werner at almesberger.net
Fri Nov 29 20:35:39 EST 2013


Two updates to last week's list:

- the Cypress CYRF7936 [1] has also been mentioned as a potential
  candidate. It seems a bit on the inconvenient side, though,
  requiring a 2.4 supply (for which it has a built-in boost
  converter, but still ...) It has a few modulation modes that
  sound vaguely similar to what 802.15.4 uses, but I can't tell
  whether they're really the same. Also, I couldn't find any
  programming documentation.

- I dug a bit deeper into the documentation of the CC2543 and it's
  growing on me. It seems that the chip should be able to output
  a clock signal, too.

Chip		Package	Clock	RNG	Export	Unit @ 1000	Comments
---------------	-------	-------	----	-------	-----------	--------
CYRF7936	40-QFN	no?	no?	?	2.88	needs >= 2.4 V
CC2543		32-QFN	yes	yes	ok	2.52

[1] http://www.digikey.com/product-search/en?vendor=0&keywords=CYRF7936

The question whether we can do without a TRNG (True Random Number
Generator) in the transceiver is still open. A possible source of
randomness in the MCU (KL2) would be the ADC. Especially temperature
measurements are terribly noisy ... which may be a good thing on
this case :)

Since the CC2543 seems to be able to perform both "special" tasks
the AT86RF232 does in the current design, namely TRNG and providing
a crystal-based clock to the MCU, it would seem like a safer choice
than the A7105.

So I think we have a winner after all :)

- Werner



More information about the discussion mailing list


interactive