[onerng talk] OneRNG blinking and no data

[Paul Campbell]

On Friday, 6 July 2018 8:53:28 PM NZST andreas.livet wrote:
> I don't know why I didn't think to test the RF noise before ! It's working
> with cmd6 and 7 so that's great.
> 
> I don't recall any research made with RF noise in parapsychology (my field
> of study), but I might give it a try with this source of entropy and after
> test with a new OneRNG with avalanche noise.

great - the one possible issue is that unlike the avalanche source the RF 
noise source is slightly bursty - it moves around looking for a random 
channels to sample (these are zigbee channels in and around the wifi bands, so 
many more than there are wifi channels).

It looks for places where the incoming data PLL won't lock and samples the 
LSBs of the RF DACs for a random time then moves on (seconds not minutes) - 
when it changes freq we have to wait for local oscillator PLL (the tuning PLL) 
to lock to the new freq - that means it stops generating noise ..... you'll 
only notice this if you're sucking data from the OneRNG at full speed, 
otherwise the internal entropy buffer will take up the slack

	Paul