I have to add driving/dimming circuit to the design based on NHD-1.8-128160EF-SSXN-F display. The schematics suggested by the manufacturer is practically taken from MIC2289 datasheet (Figure 4-4).

Now, I do not have experience with backlight, but it seems to me that their choice of a driver is far from optimal. First, MIC2289 is a booster designed to drive up to 34V diodes in series, while the display has five LEDs in parallel. Second, the chip requires external components for soft-start, while many of the alternatives have that built-in.

While looking through LED drivers at digikey I realized that this particular application falls into weird "grey area" because of the way the LEDs are connected. The majority of chips are either step-ups designed to drive series of diodes from single lithium battery in mobile applications, or step-downs designed for higher voltage sources in lighting/automotive applications.

The display datasheet says backlight should be driven by constant current 100mA and "for the reference only" implies that this would require 2.8V to 3.4V. In my case there is 3.3V available from on-board DC-DC, which falls smack-dab in that range. So, any step-up seems to be wrong, for that potential 0.1V difference. Step-downs could be OK, but I don't know if they really necessary. There are some linear chips like BCR421U that can do the job just fine with only one external resistor, while dissipating 0.05 Watt in the worst case and supporting nice 25kHz PWM dimming.

QUESTION: Which driver topology would you recommend? Do you know of any drivers better suitable for this application and requiring minimum external components?


1 Answer 1


Since this question did not get any attention I am posting my own solution just in case.

The 5 LEDs in parallel can be treated as one, so I found nice single WLED driver chip MIC4801 that needs only one resistor to set current and supports up to 500kHz PWM control, which is on par with complicated DC injection into feedback line in the schematics suggested by the manufacturer.

Unfortunately, despite 0.13V dropout claimed by the datasheet, it was only able to supply 50 mA @ 2.8V to the LED when powered from 3.3V. However when powered from 5V it immediately settled at 100 mA @ 3.07V on the LED, which are spot-on display datasheet numbers.

I do have 5V available in the circuit, so this $0.80 solution was perfect for me, requiring only 2 components and no EMI emitting booster as opposed to 7 components of the suggested $4.50 solution.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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