Custom IC vs. a bunch of cobbled together modules.
As the LEDs are fading, likely some phase cutting is done by the driver (did you open up that box and reverse engineer the insides already?), and that is bound to throw off your multimeter. An analog one is supposedly more suited for such chopped voltages than a digital one, otherwise you can connect a scope to see what's going on.
Wiring looks like they're grouped in series. At 30V I'd expect groups of 7-8 if white, more for coloured LEDs.
Custom IC vs. a bunch of cobbled together modules.
I seriously doubt I can do soldering custom IC or order it somewhere to print for me. May be you can explain in details what modules do I need to implement this task?
did you open up that box and reverse engineer the insides already?
No. The block is monolithic, no screws at all, and I have no any tools to open such box. As well I have no enough skills to reverse engineering PCB.
ava18:
Custom IC vs. a bunch of cobbled together modules.
I seriously doubt I can do soldering custom IC or order it somewhere to print for me. May be you can explain in details what modules do I need to implement this task?
Never said you could, just that this is the explanation of how they can make it so small. Have a look at say those candle flicker LEDs, or automatic colour changing LEDs. They have a complete controller built into the LED itself! Really impressive how that's done.
An H-bridge is a good way to create AC from DC. You need one that can handle sufficient voltage for this application.
Now this sufficient voltage, that you have to figure out. The simplest way would be to connect a scope to the output of the controller while the LEDs are connected and see what you get. This gives the voltage the LEDs are driven at, and the waveform itself gives more clues on how it's done. The LEDs may have current limiting resistors built in, that'd be the easiest/cheapest way of doing this.
If you don't have the test equipment you may simply get a 35V DC power supply (36V is standard, that won't make a difference, and they can often be adjusted a few volts up or down). I expect that if you connect it to the string you get half the LEDs light up, reverse polarity and the other half lights up. If you want to be safe add a 1k resistor in series (in case the LEDs don't have current limiting resistors).
I have opened the Original driver. Details on the attached pictures. But these pictures tell me nothing about the Original driver functionality.
Also I have removed cover from the first LED. Picture in the next post. I don't see any Resistors or other elements.
Talking about L298N: in the datasheet is said that VDC manipulation is in range 5V-46V, so this module is enough to control 35V strings. May be I have made a mistake in my code I posted earlier? Writing 255 to PWM pin should give 46V on output? Could you review Uno R3 code once again?
ava18:
Talking about L298N: in the datasheet is said that VDC manipulation is in range 5V-46V, so this module is enough to control 35V strings. May be I have made a mistake in my code I posted earlier? Writing 255 to PWM pin should give 46V on output? Could you review Uno R3 code once again?
The L298 is not able to generate voltage. It's just a package full of switches that routes the power applied to it in different directions. 5-46V are the limits of what it can handle, but it can only output the voltage it is given. If you power it with 12V, you will only get 12V max on the outputs.
And it looks like 12V is not enough to power your LEDs. Since the original driver uses 35V, you need a 35V power supply of your own to power these.
Mmm... Anything on the bottom of that PCB?
I miss rectifiers/diodes, and am expecting to see a transistor on the high voltage side.
The high voltage side looks a bit like a standard switching power supply, but just not quite. AC voltage in, seemingly no filtering which is just cheaping out, then I'm expecting a rectifier. The RF1 may be a rectifier, I can't really see what's in there. Two smoothing caps (8.2uF each - using two instead of one probably to keep the profile low), and two inductors of what looks like 1.5mH each (pretty high value, and I'm not sure what they are for). So that's the rectification.
Next I'm expecting some kind of resonator - usually built around a transistor, to pulse the transformer at high frequency, high as in 10-50 kHz kind of frequency. That transistor I don't see either. I don't know how the green cap comes in play.
Then there's of course the big yellow block - the transformer. The high voltage side should have four connections, one for the primary and one for the feedback winding. The low voltage side has two connections, the secondary winding. C8 is a high voltage filtering cap.
C5 must be a smoothing cap, it's polarised which implies there should be a rectifier on the low voltage side as well. It's voltage rating is 35V, which is too low for a 35V output (should have a rating of at least 50V). This could be the manufacturer cheaping out of course.
Mode switch - I don't think you mentioned this. That implies some kind of smarts that I don't see on this schematic. Really nothing on the other side?
Talking about L298N power supply:
As far as I understood I need this item to step up my 12V power supply? It is in range up to 35V adjustable by a screw.
And afterwards when I will get this item and will write to PWM pin value 255 - will I get 35V on output?
Talking about reverse engineering of PCB:
Picture from the "low voltage side":
Big blue CY1: No text
Small blue C4: No text
Coil T1: PCB-909-B5
Capacitor C5: JWCO, VENT, LF +105Ch, 05/20 LOWESR, (unsure about this line) 20uF 220, 35V
Picture from the "high voltage side" (I have removed cover from RF1 for better view):
Capacitors C1,C2: KM 105C, 20 J(M.)G, VENT, 8.2uF 400V
Capacitor C3: JWCO, VENT, LF +105C, LOWESR 05/20, 4.7uF, 50V
Two light blue elements with colored strips L1,L2: No text
White element with colored strips (rectifier?) RF1: No text
IMHO, PCB is single layered.
Really nothing on the other side?
See next post.
ava18:
Talking about L298N power supply:As far as I understood I need this item to step up my 12V power supply? It is in range up to 35V adjustable by a screw.
That should do the job indeed.
IMHO, PCB is single layered.
Really nothing on the other side?
See next post.
Singe layer PCB with parts on both sides. That's more complex than I expected, really. Looks well designed, good separation there.
Turns out RF1 is simply a 100R resistor, which doubles as fuse.
This topic was automatically closed 120 days after the last reply. New replies are no longer allowed.
