The chip mentioned is also used on some phone battery packs.I think its worth mentioning that some batteries work best if recharged before test, especially if they haven't been used for a long time (eg electric bike batteries)For a single 18650 this is fine but larger packs may never charge due to thermal protection,
Bismuth oxide, iron oxide and zinc oxide might work in the right proportions, as bismuth has interesting magnetic properties.I have some sodium silicate here so might try the second formula mentioned.
Hi, I have also determined that some phones are better than others.A good test is to see if the LED makes things fluoresce, notably blue/green EL wire or other nearby white LEDs. If so them this will work very well.If not then some phones let you increase flash power to 100% and there's actually a hack to boost it at the cost of LED lifetime, or even just replace the diode with a dedicated blacklight microLED and white phosphor external add-on array (cough SMD soldering is REALLY difficult and needs steady hands and 63/37 solder /cough) I've since found them for sale for about 60p a unit.
Interesting failure mode I just documented on "open frame" panels with the green back, the encapsulant around the edge degrades and goes low resistance. I did find a fix though, confirmed that removing it or otherwise bypassing around it (cough laser etcher /cough) will work and the panel can then be re-sealed and used. YMMV! The front contact is a PITA though and about all you can do is etch it with mechanical means or sometimes household 3% H2O2 or blue glow stick juice will work. Thanks Jeri for the tip!
Patent pending? figured out how to use fluorescent tube starters to build a home-made magnetometer that is 100* more sensitive than anything currently on the market. Catch: needs SDR chips but those are readily obtainable.
Also relevant: I usually use JFETs as current limiters, less wasteful and you can fine tune the input voltage for lowest dissipation.