Guru
Joined: 21/06/2018
Location: AustraliaPosts: 1206
| Posted: 07:51am 09 Mar 2024 |
Ok I forgot to tell you all a little tale of Nano woes that date back a quite a while.
When I first went to order a Nano I sent an image of the one I found on Ali to Poida and he said yep that there is the beast you want go for it, little did I know....
Recently when Poida was sending me code updates, quite often at times (he can program and improve faster than I can test). I had the normally deathly quiet improved zero cross version (perfect symmetry or typoed prefect symmetry) inverter code I could hear the hum varying and it grunted a bit more on startup. So I complained about the poor quality code I was being sent for free lol and Poida just glared and said it works just fine, it can't be the code.
So I had in my possession 2 whole nanos one reference unit that was dead quiet and the other that was getting all the updates to the code with varying hum. Then I got a third nano built and now it gets more interesting, randomly with new codes some would have this "hunting" hum other versions were much better.
So I got the two codes - one dead quiet and the other with the hunting hum and emailed them to Poida with notes, here is the proof, take that! My next email from Poida said yeah I tested them they are both perfect as expected.
About this time I thought Poida can't be seeing what I am seeing so I loaded all 3 of my Nanos with the exact same code and tested them, the first two perfect, the third - yep you guessed it wandering hum. It was the f$%#ing Nano mega chip which incidentally was a QFN pack (smaller than the typical gull wing types) & stamped (etched) with an Atmel logo that didnt look like an Atmel logo and clearly marked as "China".
I am positive that this was the very same Nano that had me going in circles trying to fix my inverter grunt. You can tell from all the wires soldered to it for attaching CRO probes and meters. ~$3 worth of tens of wasted hours and woes.
Moral be sure you get the right Nanos, the faulty one does basically work but I sense it emulates the mega chip or something, no doubt small port timing issues of nano seconds that sometimes all can randomly add up to inbalanced core flux.