08 Mar 2016
This post is long overdue. I've been exploring Bill Cox's Infinite Noise Multiplier and I've constructed his simple design on a breadboard. It works just as intended. However, I had difficulty in that I first used a cheap (50¢) op-amp, and discovered that the board would only work at a very slow clock speed. I sprung $10 for a faster op-amp and it worked like a charm. I used this IC for the switches.
22 Jun 2015
I recently got a pogobed test kit from Sparkfun to test Z1FFERs before the headers are soldered on. It's pretty awesome, but unfortunately it looks like I got one of the last ones before they were retired. Someone in their comments, however, mentions a place called Merifix that makes cool enclosures for testing your own boards.
28 Feb 2015
I recently put together two prototypes with transistors from two different manufacturers. I was surprised they generated different results, even though their data sheets were the same.
When I test a new board, the statistic I look at the most is the ratio of ones and zeros. With the Fairchild MMBT304 I see about 0.5% deviation – that is, 49.5% ones, and 50.5% zeros. With ON Semiconductor’s MMBT304 that bias is more like 2%.
My original whitening algorithm could handle 1% deviation but not 2.5%. I’ve since strengthened the whitening algorithm (I’ll explain that in another post), but will probably go with the Fairchild for production runs.
Read the rest of this entry
16 Nov 2014
Just got the prototypes in for version 0.2.0. This has only one noise source instead of two. XOR-ing two noise sources seemed to improve bias somewhat, but not enough that it could ace the NIST tests without whitening. One source, plus some casual temporal mixing produces stellar results, so the second source seemed unecessary. As an added bonus, the circuit and code are now simpler.
04 Aug 2014
Newer, design-y-er PCBs are in. The circuitry is the same, functionally, but it's laid out to more closely resemble the circuit diagram. Also checkout the white solder mask and new logo.
25 Jun 2014
Made the first prototype PCB a while ago. Works great so far. A better laid out one is in the factory now. It'll have exciting new design features, such as a logo, version number, and (...drum roll...) white solder mask.
16 Jun 2014
I've added a performance analysis. I'm starting with just two tests, a simple one performed on a 38mb file with ENT, and then a battery of NIST tests using the same file. I intend to make the raw randomness files available as soon as I figure out how to connect S3 to this website.
10 Jun 2014
Tonight I'm giving a short talk at Eyeo. I'm using it as an opportunity to introduce Ziffer and launch OpenRandom.org.
My first priority is to get the Ziffer out there and into people's hands. If you'd like to get an email when I'm ready to take orders, sign up for the mailing list.
As for the site, I've got a lot of ideas and I'm going to add them as fast as I can. I intend to really beef up the resources section - on deck are source code, Eagle files, links to other RNGs out there, a thorough analysis of the Ziffer's output, and instructions for building your own Ziffer on a breadboard. So, please check in again soon!