[Cryptech Tech] User auditable hardware entropy source/random number generator

Bernd Paysan bernd at net2o.de
Thu Jul 10 22:33:47 UTC 2014


Am Donnerstag, 10. Juli 2014, 22:00:57 schrieb Benedikt Stockebrand:
> Hi folks,
> 
> Bernd Paysan <bernd at net2o.de> writes:
> > Am Donnerstag, 10. Juli 2014, 22:44:43 schrieb Leif Johansson:
> >> Yeah there is no use for _low_ noise zener diodes ;-) Next they'll go
> >> after amplifiers ... the bastards !
> 
> and the vendors are all in cahoots with them---you won't find any
> statement of *minimum* noise levels in any datasheets.  It's a global
> conspiracy:-)

Indeed. Looking up e.g. NXP's low-voltage avalanche diodes (PLVA6xxA) reveals 
that they have been replaced with zener diodes (1/10 of the original noise! 
Announced as feature of course...), and the original ones aren't available 
anymore...

What you can still do is to look at the temperature coefficient.  If it's 
negative, it's zener effect dominating, if it's positive, it's avalance effect 
dominating.  The trip point (no temperature gradient) is usually around 5V, 
i.e. it starts getting interesting above 5V.  You also see that the current 
dependance goes down when you reach the avalanche dominated breakdown:

http://pdf.datasheetcatalog.com/datasheet/good-ark/BZX55C30.pdf

page 3.  The good ones for noise are the ones with a very steep plot.

With an FPGA, I would do the following:  Use the internal input pin buffer as 
amplifier, i.e. all you need is to generate a constant current throuth the 
diode (resistor from 12V), a capacitor for level shifting, invert the input 
inside the FPGA, and output it through another pin, attached to the lower side 
of the level shifting capacity via a big resistor (RC time constant 100 times 
below noise signal frequency).  This will give you a pretty good 50/50 rates 
of 1s and 0s, and except the diode itself, and the FPGA, it's all cheap 
passives.  Altera doesn't have Schmitt triggers in the input path (only MAX II 
and V have optional ones), so you can use the input buffer as amplifier.

-- 
Bernd Paysan
"If you want it done right, you have to do it yourself"
http://bernd-paysan.de/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 198 bytes
Desc: This is a digitally signed message part.
URL: <https://lists.cryptech.is/archives/tech/attachments/20140711/d1442e71/attachment.sig>


More information about the Tech mailing list