# Achieve the same intensity with many LEDs

I want to ask question about intensity control of parallel LEDs.
How can I achieve the same intensity for all LEDs?
Here follows some details.

What I want to achieve is to light up 10 (or more) high-power LEDs (EDC780D-1100).
$$(V_f = 2.1 V,\ \ I_f = 800 mA)$$

To achieve the same intensity, I am now thinking to use STCS1 constant current LED driver.
If I set the value of sensing resistor (R1 and R2) to 0.130 Ω , I will get a theoretical current of ~0.769 mA (100 mV of feedback voltage divided by 0.130 Ω).

The problem is that I can not make all LEDs in series because my input voltage is limited to 12 V, so I have to use two STCS1, as shown in the attached image.

However, these two groups of LEDs (D1-D5 and D6-D10) do not necessarily have the same intensity, because of the variations of STCS1 (U1 and U2) and resistors (R1 and R2).
According to the data sheet, the variations of feedback voltage (VFB) of STCS1 is 100 ±10 mV, but I want the current tolerance to be at least ±5% or less, because I want to do scientific biological experiments with these LEDs.

Increasing the voltage and make all LEDs in series will solve the problem, but in the future I want to increase the number of LEDs (50 or more), so this is not a realistic solution.

I thought the best way would be to use potentiometers for R1 and R2 to fine tune each current, but I could not find such potentiometers with low resistance.
Could someone tell me how I can achieve exactly the same intensity (the same current) for all LEDs?
Or, if there are other better ways to achieve the same intensity with many LEDs, could you give me an example of the circuit?

Any advice is welcome.

• Comments are not for extended discussion; this conversation has been moved to chat. Dec 12, 2022 at 6:34

## 1 Answer

You can make a precision current sink using an op-amp, a transistor, a voltage reference and a current-sense resistor. You can get select the desired precision by using a precision reference, a precision current sense resistor, and an op-amp with low offset voltage. Those three variances will account for most of your error. Don't forget to include temperature variation in your analysis.

It may be beneficial to use a low voltage reference to avoid excessive power dissipation in Rs. Also, the supply rail should be tailored to the forward voltage of the diodes. If you have unnecessarily high voltage, then the transistor will have to dissipate extra power.

The basic idea is shown in this diagram from a TI document:

Document is copyright by TI, but I claim fair use for this brief excerpt. Document was obtained here: https://www.ti.com/lit/an/snoaa46/snoaa46.pdf