.:madworm:.'s photos with the keyword: current

Testing the power supply that came with my LED strip. Made with gnuplot.

I couldn't get hold of a precision shunt resistor, so I had to do with this 10% 5W one here. It didn't turn out too bad at all. This one doesn't affect the current displayed on the power supply's display (the display measures the current itself), but the current set-point in the opamp feedback loop. I have a schematic, but it is not quite matching the PCB revision I have, so I couldn't instantly figure out which pot to tweak to compensate for small deviations. Reverse engineering a schematic from a given PCB sucks. I compared the current set on the power supply (via a keypad) and the current measured by a good multimeter and it wasn't off by more than 5% worst case. That's good enough for me right now.

I've retired this one, but kept it. Just in case ;-) It was replaced by a standard 10% 5W 390mΩ resistor. Not perfect, but it works good enough for me. The worst deviation I've measured is less than 5%.

You need ad least 1.5mm² to handle 16A safely. 0.5mm² is simply not enough! But does the actually have 0.5mm² of copper inside? Let's check!

The plug itself might handle 16A, but not the cable that was attached to it.

Hmmmm. Each wire has about 20 strands. Diameter 80µm. There's just NO WAY that adds up to 0.5mm². Depending on how I calculate the cross section (multiply strand cross section vs. measure wire diameter with air gaps), I end up with values between 0.005mm² and 0.15mm² - NOT ENOUGH.

...and much worse inside.

1A / div The desired current is 10mA, which is reached in the end. Unfortunately any sensitive device would have been killed way before that (e.g. a LED).