Electronics > Projects, Designs, and Technical Stuff
HOW DOES THIS WORK?? Bench light circuit has me stumped...
Prehistoricman:
Oh, and since the capacitor in the dropper is basically an AC resistor, a very common failure mode for mains LED lamps is for 1 LED to break (in a string of 30 or so LEDs) and then the filter capacitor explodes. Without the load, the output voltage rises greatly and the filter cap is often not rated for 300/400V.
engrguy42:
--- Quote from: Prehistoricman on March 27, 2020, 01:04:10 pm ---These kinds of products are built down to such a price that they need every sneaky transistor/diode trick they can use. And that can make them surprisingly hard to analyse for being simple.
You say it doesn't charge the batteries if it's plugged in and the lights turned on, right? That would be an interesting thing to investigate.
Why did you use a resistor in the schematic instead of the LED array? You can use a single diode and edit its properties to make it behave like the whole array.
--- End quote ---
Yeah, as a matter of fact the other day I made the mistake of leaving the light on all night while charging and the next day the light was super dim and battery dead. The manual even says STEP 1: Turn off the light before charging.
As far as modelling the LED array, I didn't think it was worth figuring out the LED characteristics in any detail. I inserted my ammeter, read the current value, and that's all I needed. Are you thinking the LED characteristics would make an appreciable difference in any of this?
Prehistoricman:
--- Quote from: engrguy42 on March 27, 2020, 01:10:18 pm ---Are you thinking the LED characteristics would make an appreciable difference in any of this?
--- End quote ---
Not in the DC state. Yes for anything AC or transient.
I think your most recent graph would look smoother with LEDs rather than a resistor.
I don't know how LT-SPICE works but to get a single LED to behave as you wrote earlier (3.8V, 360mA), I set IS/Saturation current to 2e-19 and RS/Serial resistance to 1.7.
--- Quote from: engrguy42 on March 27, 2020, 01:10:18 pm ---Yeah, as a matter of fact the other day I made the mistake of leaving the light on all night while charging and the next day the light was super dim and battery dead. The manual even says STEP 1: Turn off the light before charging.
--- End quote ---
Ah, now it's obvious (and it was from your original post). The charger supplies less current than the LEDs consume.
engrguy42:
--- Quote from: Prehistoricman on March 27, 2020, 01:31:08 pm ---
--- Quote from: engrguy42 on March 27, 2020, 01:10:18 pm ---Are you thinking the LED characteristics would make an appreciable difference in any of this?
--- End quote ---
Not in the DC state. Yes for anything AC or transient.
I think your most recent graph would look smoother with LEDs rather than a resistor.
I don't know how LT-SPICE works but to get a single LED to behave as you wrote earlier (3.8V, 360mA), I set IS/Saturation current to 2e-19 and RS/Serial resistance to 1.7.
--- Quote from: engrguy42 on March 27, 2020, 01:10:18 pm ---Yeah, as a matter of fact the other day I made the mistake of leaving the light on all night while charging and the next day the light was super dim and battery dead. The manual even says STEP 1: Turn off the light before charging.
--- End quote ---
Ah, now it's obvious (and it was from your original post). The charger supplies less current than the LEDs consume.
--- End quote ---
DOH !!! I never connected the dots. I measured an LED consumption of 360mA, but the AC charger can only supply something like 120/3k or 40mA.
engrguy42:
Okay, so thanks again to everyone who helped on this. But I'm tenacious, and I've been trying to clean this all up and make sure I've got it correct and figured out.
I think my LTSpice diagram (attached) is now 99.99% correct (after spending a lot of eye-straining time tracing the tiny board), and I only have a couple real issues/questions left:
- I'm assuming the two unmarked SMD transistors (I'm assuming they're NPN) are both identical and marked as per the attached photograph. Is that a reasonable assumption that all SMD transistors have the same terminals (C is the single, and E & B are the double pins)?
- I also realized that while charging the LED's will come on, but as was mentioned before by the great Prehistoricman it seems the LED's draw more (360mA) than the crappy capacitor dropper can provide (I measured 200mA into the battery while charging), so the charger won't charge while the lights are on.
- I cleaned up the LTSpice diagram, and re-named components to match the markings on the board, and added the connections to the on/off switch (dashed box in LTSpice diagram). Based on that, it seems to me that when the switch is in HI mode (as shown in diagram) that D3, Q1, and all the Q1 biasing resistors can basically be erased, as they're only used in SOS mode. Does that make sense? And I have no clue what the SOS mode does...and don't really care. Doesn't seem to work anyway.
- It seems like LOW mode merely inserts an additional 1k resistor to the Q2 biasing circuit, and presumably changes the LED current.
Anyway, my LTSpice simulation pretty much matches my scope/meter measurements, so I'm think I'm pretty close.
Thanks for any help.
Navigation
[0] Message Index
[*] Previous page
Go to full version