# 10 3W RGB LEDs in series control current

I'm having the following schematics connected to a 5V 25A DC power supply. All LEDs are 3W RGB, additional values are. R1: 6.8 ohms 1/2W R2: 4,7 ohms 1/2W R3: 4,7 ohms 1/2W

I'm using an Arduino Mini and TIP120 to control each channel (R, G, and B). The 10 leds are connected in parallel, and they actually emit light (which is quite important).

During testing, and connecting a single led, this led was drawing about 0.4A for the Red channel. Now I have connected the 10 LEDs in parallel, the Red channel is only drawing 0.8A. I tried placing a resistor between pin 6 and the base of Q1, but the current drawn stays the same. The resistors tried vary between 10 ohms - 10 Kohms.

Could someone help me out? How can I direct more current through my Red channel, to make the LEDs shine brighter?

EDIT1: I've tried connecting all 10 LEDs directly to my 5V power supply. The light up all more brightfull and draw 2A in total. So I assume the problems ly in wiring up my transistors?

• Is the issue present only for the red LED? have you checked the actual voltage on the Base pin of the TIP120? have you checked the voltage drop on each TIP120? can you try with two LEDs instead of only 1 or 10? have you checked the actual voltage on the 5V rail? – FarO Jun 3 '15 at 15:30
• I don't think your Arduino has the output oooomph to kill a TIP120 (or itself), but connecting a transistor in this way without a base resistor is bad practice. In theory either your Arduino outputs or your transistors (or both) could be damaged. – Wouter van Ooijen Jun 3 '15 at 15:38
• You say one red LED connected directly draws 0.4A, and 10 LEDs in parallel draw 2A. Doesn't that strike you as odd? Can your PSU realy supply 25A, or is there something wrong with your measuremenst? – Wouter van Ooijen Jun 3 '15 at 15:40
• Keep in mind that if you connect the LEDs directly they get the full 5V, if you use the transistor, they will get (5-Vf) volts: the transistor is not a switch that either blocks or lets the currnt go: it brings always a voltage loss. The same for MOSFETs. – FarO Jun 3 '15 at 16:19

You are doing something wrong. I suspect your either your wiring (you may be using much too small wires) or your power supply is not behaving correctly.

When you connect all 10 LEDs to the 5 volt supply, you claim a current draw of 2 amps. That is .2 amps per LED. In that case, a 4.7 ohm resistor will drop just about 1 volt, leaving your LEDs dropping 4 volts. And that is not possible. Without a picture of your setup, I suspect that you are not using thick enough wire to connect to your power supplies, and the wire resistance is causing you problems. You may even be on the edge of melting your insulation. When you do your high current tests, do you smell something funny?

In addition, if you ever do get the current right, you will kill your transistors. From the data sheet, the VCE(SAT) can be as high as 2 volts at 3 amps. Let's use this as a working number rather than 2 amps, since you also stated that a single LED will draw .4 amps, so with all of them working you should draw 4 amps, not 2. With a 2 volts drop on the TIP120, and a current of 3 amps, the transistor will dissipate 6 watts, and this is more than a bare TO220 package can handle. You must provide a good heat sink for your transistors, or change transistors to something with a lower voltage drop. As Asmyldof suggested, other transistors should be used. Personally, I'd recommend a MOSFET.

Apart from the good comments to your question, I would like to add the following:

TIP120 Datasheet

See page 2, "Collector-Emitter Satuarion Voltage" written as Vce(sat) is very important to know about. It is the voltage that the transistor will have across its C and E pins when you use it as with the numbers given. I don't think 12mA is very far off the mark of what your Arduino is driving into the base under the circumstances, might be 15mA, certainly not 20mA or more. With a collector current of 2A that's not going to be much less than the 2V for 3A as listed.

So, what happens? Your LEDs can't pull the maximum current, because they don't see 5V. So the transistor and the LEDs start "trading" on their voltage-current curves until they end up at what is appropriate.

Regardless of the fact that your supply should be stable and reliable and that you should always use at least some base resistance not to break anything, or at least to be able to predict behaviour. This set-up just can't work, because a darlington device has a relatively large Vce-sat, even though it has a nice high current amplification.

You're better off using either: 1 PNP type driving a single NPN type; A MOSFET; or a Power Transistor with a high amplification without the extra base transistor stage, but off the top of my head I can't think of any.