Søren Kjærsgaard<p>In January, I bought two 10V references, based on LM399. I was a bit skeptical on the design, following Marco Reps on YouTube does that to you 😂, but a teardown revealed quality components (Vishay precision resistors) and a decent circuitry and PCB layout.<br>I measured the standard deviation on the 10V output, using my 3458A, and I’ve done so monthly-ish ever since, leaving them powered all the time. <br>In March I added two more, unit 3 and 4.<br>Result so far: unit 1,2,3 are converging towards 1uV SDEV, unit 4 however (separate curve), started out sky-high, but it has now joined the ppm club 🤷🏼♂️😳<br>Someone mind telling me what’s going on? I mean, it’s wellknown that the noise drops over time, but why? What physical/chemical factors comes at play, causing this behavior? <br><a href="https://techhub.social/tags/voltnuts" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>voltnuts</span></a> <a href="https://techhub.social/tags/ppmgeek" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ppmgeek</span></a> <a href="https://techhub.social/tags/electronicsengineering" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>electronicsengineering</span></a> <a href="https://techhub.social/tags/testandmeasurement" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>testandmeasurement</span></a></p>