LED flashing error code

Hans hans.ridder at gmail.com
Mon Sep 4 17:16:35 BST 2017


Thanks for the rapid response Paul. It's taken me a couple days to get access to a scope and time...

C15 has about 18V on it. I'm getting about 40mV of noise at C17, and then a few hundred mV on the collector of T2 (the input to the CC2531). The highest noise peaks are <1V, not anywhere near the 2.5V logic threshold specified in the CC2531 datasheet which has me suspicious that T2 (the avalanche junction) has "gone flat".

I don't have any spare SOT-23 transistors of any kind to try swapping, but I tried breadboarding the noise generator with some 3904's I have in a TO-92 package. That circuit seems to generate a couple hundred millivolts at the coupling capacitor (what would be C17 in the OneRNG), so perhaps I'm on the right track. That said, the circuit seems quite sensitive to component selection, values, and even voltage so I'm not convinced. Your comments appreciated.

I also stumbled across this page http://holdenc.altervista.org/avalanche/ in which the author uses a similar noise source and seems to be getting higher noise voltages at what appears to be much lower currents through the avalanche junction (of a 3904). Your circuit is limited at around 4mA (20V and 4k7 ohm), his is around 100uA (13V and 150k ohm). But as I say, the whole thing seems very sensitive to the exact components and values so I'm impressed you've made it work well in production.

In the meantime I've ordered some garden variety (i.e. ebay) SOT-23 3904's, which I should have around anyway. I've also ordered a replacement OneRNG, but I would like to fix this one.


More information about the Discuss mailing list