I want control LED strip using arduino nano and TLC5940NT.
I build everything but something miss, because every TLC output finally gives lover Voltage.
First gives ~9.6V, next lower, next lower....six TLC output gives only 9V....
I dont know how to draw schematic but try to explain:
TLC output -> BC556B -> C2073 -> LED strip "-"
5V to BC556B Emitter I get using 12V power supply and L7805 (dont use any resistors or capacitors)
Arduino Nano GND and VIN connect to 12V power supply.
so why after C2073 I dont get 12V? C2073 only open 12V "-". I thing C2073 is not fully open and I get only 9V or lower. Could it be that if I use more than one TLC output TLC cant fully open BC556B which cant fully open C2073 and in result I get lover and lover final voltage?
You should probably draw a schematic anyway, I don't quite understand your setup!
If you don't know how to draw it because you don't have a PC program, use pen and paper!
If you don't know how at all, look up some schematics and symbols (PNP, NPN transistors etc) online, if you're unsure about the IC's, draw a block with pin numbers.
The TLC9540 does not source power it only sinks it. This makes your design wrong.
Therefore you will need a pull up resistor on the base of each of those transistors. Mark which one is the emitter with the conventional arrow. Why are you using a 5V regulator, I can't see the point? If you must use a regulator you must use a capacitor each side of the regulator and saying you don't is not acceptable.
Does this show two power supplies? If so, you need a very large capacitor on the one driving the LED strip.
How many LEDs are there on this strip and what strip is it? Link please.
First off, C2073 is not a MOSFET transistor although you drew it like one. It's a regular NPN transistor and you could have driven it directly off the TLC5940.
BC556, it looks weird the way you put it in the schematic and also you don't have any current limiting resistor on the base! Fortunately for you I think you connected in a way that you didn't burn any components.
Again, what is your idea of having both the BC 556 and the C2073?
Why the 7805?
You could try doing TLC output, base resistor, C2073, LED strip. However BJT transistors like the C2073 are not really meant for PWM, so you might have to find a MOSFET, but I guess you can try it out at least, see if it works!
EDIT: Didn't check that TLC only sinks current, read what mike said
Grumpy_Mike:
The TLC9540 does not source power it only sinks it. This makes your design wrong.
Therefore you will need a pull up resistor on the base of each of those transistors. Mark which one is the emitter with the conventional arrow. Why are you using a 5V regulator, I can't see the point? If you must use a regulator you must use a capacitor each side of the regulator and saying you don't is not acceptable.
Does this show two power supplies? If so, you need a very large capacitor on the one driving the LED strip.
How many LEDs are there on this strip and what strip is it? Link please.
5V regulator is for C2073 transistor. TLC5940 is not powerful enough to open 16 C2073 transistors so between TLC and C2073 I put BC556B transistor to open 5V for C2073 transistor. I dont want to use other power supply (5V) so I use L7805 to make 5V from 12V.
No, its one 12V power supply.
In strip is 60 x SMD3528, 4.8W, 12V. One TLC channel = 1m LED strip
I would be more worried about it being able to fully turn ON with 5V signals. Unless OP dropped the regulator and has 12V to the gate. Then still more changes have to be made to that circuit...
No, it needs to be a pull up because the TLC5940 is only a current sink and pulling it up to 12V is the right way to drive an external N-channel FET.
But the point the OP is missing is that the strip is not switching off fully, maybe he hasn’t noticed yet or tested it in a dark environment but the strip will not fully turn-off if wired like this.
This is because when the TLC5940 is sinking current and thus an LED wired in the conventional way will be on, in that circuit it will be off. There is an inversion between signal and light level.
This would not normally be a problem just that the numbers would be inverted driving it. But this chip works in slightly an odd way in that when the LEDs are set to be on at full brightness it produces a PWM signal that is not fully on but one that has a duty cycle such that is one clock pulse short of fully on. Fine in normal operation as you won’t be able to see such a small brightness change. However, when you invert that signal like in the circuit here it has one clock cycle on and the rest of them off. Meaning the LEDs can never be fully off.
Intriguing why the OP doesn’t spot this, but it is real and so the circuit used is crap. No one should believe it works. The only thing I can think of is that the circuit is so piss poor it will not respond quickly enough to drive the LEDs on for such short a time.
So I would conclude that it is the IRF640 that is the poor component and the IRF610 that is the good one.
The transistor fully turning off is not the problem. The transistor not turning on “fast enough” is masking the problem you have due to faulty design of the circuit.
The transistor fully turning off is not the problem. The transistor not turning on “fast enough” is masking the problem you have due to faulty design of the circuit.
so how should I build everything to do it right? can you draw I scheme?
I run fade up and down code and it looks "fast enough" for me, no lag or blinking. But if circuit is bad I want to know how to do it right.
