Go Down

Topic: ULN2003 - Is there any other stronger tips than ULN2003? (Read 370 times) previous topic - next topic


Hi, I am currently working on a project of 500 LEDs powered by two pieces of ULN2003 and Nano.
I connect around 23-25 pieces of LEDs on each pins, which is about the maximum current of 500mA for each pins.
However, the brightness of the LEDs are not enough for my project, and I need these LEDs to be brighter.
I wonder if there is any other tips that can provide more than 500mA for a single pin such as 800 or 1000mA.
I cannot change Nano to other shields for this project. I might consider an external power output but I do not wish to burn out the shield.

I would deeply appreciate if someone could advice me on this issue.


The ULN2003 has an absolute max switch current of 500mA for ONE pin. Not all seven.
Absolute total max for seven is 2.5A, but you can't reach that because of dissipation limits (chip temp).

It seems you have connected 23-25 LEDs in parallel to one pin. That would explain your current requirements.
Can't you connect the LEDs in series, and use a supply with a higher voltage.
Post a circuit diagram.


No technical PMs.
The last thing you did is where you should start looking.


Feb 10, 2018, 12:19 am Last Edit: Feb 10, 2018, 12:21 am by Grumpy_Mike
That is very odd, yes I get a page not found from that link.
However when I go through google " Adafruit FET" I get the page and from what I can tell has exactly the same URL.
Try again https://blog.adafruit.com/2011/09/28/a-beginner%E2%80%99s-guide-to-the-mosfet/

Ah it is the apostrophe that is different, did the first on on the iPad second on my Mac.

Go figure?


Thank you for advices. I attached pictures of what I have been working on.
It is a very simple schematics - 13 lines of 23-25 LEDs in series and each line is connected to ULN2003 with 5V from a mobile charger plugged into a power outlet.
In precise, I'm using two Nanos to power the total of 500 LEDs; so around 250 LEDs with 2 ULN2003 on one Nano shield. There are also some RGB LEDs and these lights up very bright. These are connected directly to Nano.

I mounted those LEDs on a wooden panel and these are supposed to be seen through a paper on the front side. As you can see the picture, these red LEDs are not really visible unless in the dark.
I need them to be bright as the RGB LEDs without making the room dark.

@Mike  I wonder if MOSFET tips can be used with ULN2003. The input of ULN2003 is cathode so I think MOSFET cannot be used with the tip, or do you think if using MOSFET on each LED row can make 25pieces of LEDs brighter than ULN2003?

@Wawa  I think I connect the LEDs in series. I wonder if you know any good external power supply for this type of LED project. However, it has to be not powered by a battery but plugged into an outlet.


10 leds in series do not draw 10 times the current. With 10 leds in series you need a voltage source that can provide a voltage that is at least the sum of the forward voltages of the leds; and you still need.to take the voltage drop over a current limiting resistor into account.

A 5V supply might do 2 or 3 leds in series but that's it.
If you understand an example, use it.
If you don't understand an example, don't use it.

Electronics engineer by trade, software engineer by profession. Trying to get back into electronics after 15 years absence.


It looks to me like 28 groups of 23 leds connected in parallel. Except for the missing leds where they have been replaced by the rgb strip.

I'm sorry, but this is all wrong.

Leds need series resistors to limit the current flowing through them. You have none that I can see. When leds are connected in parallel, there should be a series resistor for each led, otherwise the led in the group with the lowest forward voltage will be the first to fail. Then, the others in the group will fall like dominoes. So if you want to do it like this, you would need to have soldered in 644 series resistors.

If you had used, say, 220R series resistors, each group of 28 leds would require around 320mA. That's around 9A in total. Without the series resistors, it would, until it started to fail, it would need even more current. Is your "mobile charger" power supply able to supply 9A+? If not, that might explain the dimness. Is the power supply getting hot? Are the ULN chips getting hot? I expect they are.

What I would suggest you re-wire each group of 23 leds like this:

Re-connect the leds in series (not in parallel as you have them now) into sub-groups of 5, each with a 100R series resistor. If there is a final group of less than 5, you will need to use a higher value series resistor with that group. For example, a sub-group of 3 will need a 300R series resistor.

This way, each group of 23 leds will consist of 4 or 5 sub-groups of up to 5 leds. Each sub-group will draw around 20mA, but will need a 12V supply. So your 12V supply will need to supply only around 3A. You can run the Nano's on 12V also because they have on-board regulators (use the Vin pin).

Unfortunately, The RGB strip still needs 5V. With around 50 RGB leds on the strip, this will require around another 3A. To avoid having a second power supply, you could use a DC-DC converter, rated for at least 3A output. This can be powered by the 12V supply but will need another 1.5A. So increase your 12V power supply to around 4.5A and that should do it.


@Mike  I wonder if MOSFET tips can be used with ULN2003. The input of ULN2003 is cathode so I think MOSFET cannot be used with the tip,
No you are wrong. The FETs pull down, that is  sink current and this is exactly what the ULN2003 does. You connect the cathode to the ULN2003 just like you connect the cathode to the drain of the FET.


It is a very simple schematics - 13 lines of 23-25 LEDs in series and each line is connected to ULN2003 with 5V from a mobile charger plugged into a power outlet.
What ever you have you don't have that. Please post a schematic.


Thank you for your advices. I attached more pics - LEDs &power source and a brief schematic because I think it is a bit confusing. Sorry, the schematic is not really neat but I hope it explains.

@PaulRB  I appreciated your detailed advice. As you and other people said, these LEDs are wired in parallel, not series. I am sorry about this confusion.
I think that it will be better if I make some sub-groups of LEDs with a resistor instead of 25LEDs in parallel altogether. Do you mean to connect the 5 sub-groups into one series or one parallel for a pin? It will be great to do this method with MOSFETs and 12V. However, my issue here is I do not have so much time to explore the best solution, and I have to finish this project asap. I am not familiar with using external power sources so I may have to stick to 5V and try how I can make them brighter with available materials and a time frame.
I once heard that when LEDs take up all the current from a pin such as 500mA from ULN2003, they do not need any resistor because there would not be additional current flowing into LEDs to burn out. For instance, A LED with 20mA x 25pieces = 500mA. It seems that while thinking about using the project for a long term, your way of wiring will be better treating LEDs.

@Mike  I attached my schematic. So I think that it might be possible to use a MOSFET and ULN2003? such as placing it between Nano digital output pin and an input pin of ULN2003.

@sterretje  I know it is a bit bold project to use too much numbers of LEDs, but consider it as a learning/experimental process...


You have no current limiting resistors and you have wired raw LEDs in parallel. That is not a good thing to do.


It's a common mistake for beginners to read a data sheet or tutorial and misinterpret "the maximum current of the pin is X mA" to mean "the chip will safely limit the current to X mA". But in fact it means you must take care not to exceed X mA or the chip will be damaged.

I am unsure what to advise you to do. You must finish the project asap with the components you have, but you need to start over with a better understanding of circuits and leds. I think perhaps you should hand over the project as-is, with all its faults, and in the write-up, describe the errors you made, why they have the effects they do, and how you would design the circuit differently if you were starting over. At least that way you can show that you have learned from your mistakes and now better understand the principles involved. The most important lesson being "don't buy any components or plug in your soldering iron until you have had someone more experienced review your design on paper".

Go Up