The circuit I have attached is basically LED arrays with some transistor and N channel Field-Effect Transistor. It requires 17v, 400mA to work. I will only glow either white led or RGB. They won't glow at same time, due to less current. . The N channel Field-Effect Transistor (XORB48) has

  • IDS: 5.8A
  • VGS: ±12V
  • VDS: 30V
  • CISS: 623
  • VGS(th) = 1.05V

It has low gate charger and operation gate voltages as low as 2.5V.

Each white led is of 0.2W (2835) and RGB is of 1W (5050) each.

I will control and dim the LEDs using Arduino PWM, for that individual GPIOs are connected, mentioned in circuit.

I want to know what kind of supply is needed for this? Constant current? Constant voltage? Does it require High power factor? How about ST's Viper22a or power integration's LNK IC?

Schematic :-

enter image description here


closed as off-topic by Olin Lathrop, ThreePhaseEel, Daniel Grillo, Voltage Spike, Dave Tweed Nov 29 '16 at 17:56

  • This question does not appear to be about electronics design within the scope defined in the help center.
If this question can be reworded to fit the rules in the help center, please edit the question.

  • \$\begingroup\$ Hope you know those parallel connected LED strings (white ones) is rather bad, low reliability, design. \$\endgroup\$ – carloc Nov 27 '16 at 10:41
  • 1
    \$\begingroup\$ @carloc But due to less current, the white ones had to be place in parallel. How can it be improved? Please suggest. And also why this parallel design not reliable? \$\endgroup\$ – Embedded Geek Nov 27 '16 at 10:45
  • \$\begingroup\$ If you drive constant current two (or more parallel) string they are going to share 50%/50% only if they are identical. In real life this is never true and one of the two will get some more and hence get warmer then the other one. Unfortunately LED voltage drop temperature coefficient is negative so the warmer one will decrease its voltage drop and tend to get even more current and so warmer and again more current and so on. This can lead to different brightness or even one string failure. The best practice to avoid this would be to have an individual current source for each string. end part 1 \$\endgroup\$ – carloc Nov 27 '16 at 11:00
  • 4
    \$\begingroup\$ Closing and -1 because sending someone you don't know private email to ask them to answer a question here is totally inappropriate. SE doesn't have private messaging for a reason. Deliberately circumventing that is not cool. \$\endgroup\$ – Olin Lathrop Nov 27 '16 at 14:26
  • 1
    \$\begingroup\$ @OlinLathrop Extremely Sorry for that. \$\endgroup\$ – Embedded Geek Nov 27 '16 at 16:00

I want to know what kind of supply is needed for this? Constant current?

Well, your MOSFET has a 5R6 source resistor that will self regulate the current at about 125 mA so clearly a constant current supply isn't needed but, if as you say you designed this then why are you asking?

You might alo add a resistor in series with the GPIO line so that the BJTs that form part of the current regulation don't overstress the GPIO line if they conduct down to ground when current limiting.

You might also want to check what the gate threshold voltage is for those MOSFETs to make sure they can be activated with the limited voltage drive from your GPIO lines.

  • \$\begingroup\$ I agree with Andy's suggestions. Also, to answer your questions, the 17V DC supply stated in your schematic should work fine. (You need > 15V to get the blue string to light up.) Going above 17V will just make the MOSFETs run warmer. So it's not super-critical. Just find any DC power supply (constant voltage) that can deliver around 17V at 1A. \$\endgroup\$ – Rich S Nov 26 '16 at 19:44
  • 1
    \$\begingroup\$ @Andy I just wanted to confirm the supply before sending it for commercial use. The gate threshold voltage (Vgs) is min 0.65V and max 1.45V. I think that is sufficient to drive from Arduino or any controller with 3.3V GPIOs. and A 10K pull down is needed for the GPIO line?? Also let me know what particular kind of supply is needed? \$\endgroup\$ – Embedded Geek Nov 27 '16 at 2:24
  • 1
    \$\begingroup\$ @RichS I don't require current more than 400mA and it's a safe figure. Going for 1A would just make BOM more costly. \$\endgroup\$ – Embedded Geek Nov 27 '16 at 2:26
  • \$\begingroup\$ @AnujMattóõ 1A doesn't make it that much more costly , but the deisgn challenge is greater. heat sink for 17W and efficacy are certainly a challenge and this design has no heatsink and RGB LEDs probably only using 6W and not 11 W and 20xWhite LEDs using <1.4W(<0.12A*12V) and not 11W. I would use 0.5M~1MHz to permit ceramic & plastic caps with much lower ESR, greater efficiency and less ripple. Which means not this chip. \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Nov 28 '16 at 1:13

The circuit is quite simple and you don't need constant current. Constant voltage is enough. Just get a regulated DC supply. ST's Viper has some good reference design which can be helpful to you.

  • \$\begingroup\$ I assume If you have seen the 4S White load , if so you are completely wrong to apply CV 17 V with 4.25V per White LED \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Nov 27 '16 at 23:55

Not the answer you're looking for? Browse other questions tagged or ask your own question.