OK, back to my capacitative reactance experiments...
First, I hadn't actually fried my LED tent lamps. All that toasted was the chip resistors; the LEDs are fine. I soldered in axial resistors on one, and it's working again.
Real grid AC works completely different than the pseudo-AC from my Black and Decker 750W inverter. The inverter's waveform shows on the scope as a 160V square wave, pulsed DC, at about 60Hz. How that actually functions as 120VAC I can't figure out. But it does, with most things. When I put my scope probe on every point in the LED circuit, I see the exact same pulsed 160V. But somehow the voltmeter shows only 2.5VCD across the lamp, with a single .22uF mylar cap in series. Adding more caps in parallel increase the brightness a little, but the voltage rises only a few tents of a volt each time. Adding a 200uF cap across the DC terminals of the bridge rectifier increases the brightness tremendously. With a 1.72uF cap, I get roughly the same brightness as with 6VDC. But it needs the filter cap across the DC terminals, otherwise it fries the resistors.
Now, with grid AC, the filter cap isn't necessary, presumably because the bridge rectifier gives you both halves of the sine wave. But if you're going to make these AC LED lamps for resale, it's probably recommendable to include the filter cap to avoid warranty returns from customers using inverters.
A note about using an oscilloscope: while the circuit under test may not care a whit which is power and which is neutral, the scope has its own ideas about those things, so if you don't have the circuit's "power" leg the same as the scope's, you won't see what you're expecting anywhere in the circuit, since the circuit's power is the scope's ground, and vice versa. Keep that in mind. Also, to see what's really going on at the bridge rectifier, you need probes on both the positive and negative leg; it's the difference between the two that matters, now matter how far the waveforms might swing above or below "earth" ground.
Back to blog or home page
last updated 2013-01-10 21:20:45. served from tektonic.jcomeau.com