I like LEDs – a lot.
A few days ago I found this special offer for a 5m warm-white LED strip on ebay and just couldn’t resist. It arrived today, was indeed shipped from Germany very quickly. So it seems the seller – of Chinese origin – has a local warehouse. Nice.
I tested the LED strip, all seems to be fine. The power supply is a switch-mode one, but a bit bulky. Standby power consumption is about 1.5W – not too bad, but could be a bit lower. I’ve seen worse (3W-ish) – and better (0W). A few odd things caught my eye. a) The IEC cable, supposedly rated for 16A, is suspiciously thin and flexible. b) The IEC plug doesn’t fit very well – too much wiggle. I tested it with another device – same thing. I chose to replace the cable with a known good one, chop the iffy one apart and take some measurements. I came to the conclusion that a) The IEC plug is not quite according to standard. Its dimensions are, shall we say… a ‘bit’ off. b) there is definitely something wrong with the cable itself. The plug says 16A, but the cable is stamped “0.5mm²” – that does NOT match.
The BAD IEC plug:
A good one:
Chop, chop, chop
There are about 20-ish strands in each wire, but they are a bit odd. Strand diameter is 80µm, so I calculate a copper-area of about 0.10mm² (0.005mm² per strand). Hmmmmm. If I go a different route and measure the effective OD of the wire (not compensating for the air-gaps…) I end up with about 0.15mm². A bit more, but still faaar away from 0.5mm². And I certainly don’t want 16A going through that. The 16A printed onto the plug require at least 1.5mm², not 0.5mm², not 0.15mm² and certainly not 0.005mm². CRAP.
BUT THERE’S MORE!
After cutting the wire, stripping off the insulation… I found something else. I ‘tried’ to twist the strands together, but they just didn’t want to stay twisted! They immediately sprung back. The strands are suspiciously springy… almost as if they weren’t made of copper… could it be? Total resistance of the wire (about 1m) is about 1.6 Ohm. Again I didn’t compensate for DMM leads at that point, but when the leads are shorted together I get almost 0. So the wire has significant resistance. The specific resistance of copper is about 0.0179 Ohm * mm² / m, steel is about 0.12 Ohm * mm² / m. So for pure copper I should have measured about 0.18 Ohm, for steel the theory predicts about 1.2 Ohm. The latter is much closer to what I measured.
LOAD TEST using my dummy load project (YauDL):
Gnuplot files used to create the graphs.
One thing that I noticed is that the strip gets quite warm. Although the LED-strip seems to have quite thick copper traces (apparently copper), there’s a voltage drop of about 1V from one end of the strip to the other. The current at the supply side is 1.26A.
Right side: LEDs close to the power supply, left side: LEDs at the end of the strip. It’s a bit hard to see, but the right ones are brighter.
One of the LEDs died during a “burn-in” test. Luckily I have a couple of PLCC2-3528 warm-white ones, so it will be replaced. It must be replaced, as now a whole diode-drop is missing in a segment. More current going through the remaining 2 diodes is likely to kill them later.
No more unexpected fatalities so far!
The temperature increase of the LED-strip is not an issue, as it only happens when the thing is rolled up. Wasn’t thinking straight at the time of initial testing.
There were a few ebay-messages going back and forth between me and the seller. I had to use google-translate (English <--> Chinese) to make my point (bad 240V IEC cable). They offered to send me a new power-supply, which I declined. I quite like the power-supply and replaced the cable myself. My intention was to make them check / reevaluate their source of AC cables. Was I successful? We shall never know…