TPS61158: LCD backlight driver

Erfan Omidvar

Part Number: TPS61158

Tool/software:

Hello,

I designed a LCD backligh driver using TPS61158. but I faces a problem.

I am trying to balance fast voltage response for brightness control with minimal ripple for my LCD. When I reduce the time constant (using smaller capacitors), I get faster charging and discharging, but it increases the ripple( around 10V ripple) beyond my LCD's acceptable range. Increasing the capacitance reduces the ripple( Around 500mv), but makes the charging slower and causes the capacitor to remain active even when the PWM is off, affecting the brightness control. This creates a trade-off between faster response and controlling ripple within the desired voltage range.

Does anyone can help me to get the solution?

I also sees the figure provided in datasheet. it has a very slow time constant. So How should we control the brightness with such slow discharging time?

So, Main question: How can I have fast discharge without cost of higher ripple!?

6 months ago

0 Patrick Yao 6 months ago

TI__Mastermind 34760 points

Hi Erfan,

The waveform you attached is the switching node waveform, but the LED current is still a constant DC value which equals to 200mV * PWM duty / RFB. It is not PWM dimming. May I know what kind of symptom you have observed with your LCD?

Patrick

0 Erfan Omidvar 6 months ago in reply to Patrick Yao

Prodigy 80 points

What I see at the Vout, is eactly same as figure. When the PWM goes off, cap starting to loose its charge, but still high! so, this make difficult to control brightness using PWM.

For example, when I Change the duty cycle of PWM, the brighness of my display does not change effectivly. also some flikering at low duty cycle Lets say from 1 to 8. brightness-levels = <0 32 40 48 62 92 128 1000>; Changing from the duty, doesnt affect the brighness effectivly, since when the PWM is off, VOUt at COUT still high(because of discharging time constant).

I tried to decrease the COUT, to decrease time constant, so discharge faster. BUt, the problem isº the ripple , soike I have is huge now!

0 Erfan Omidvar 6 months ago in reply to Erfan Omidvar

Prodigy 80 points

Uper one, my PWM.

Lower one, V over cout.

0 Patrick Yao 6 months ago in reply to Erfan Omidvar

TI__Mastermind 34760 points

Hi Erfan,

So what is the input PWM frequency?

Patrick

0 Erfan Omidvar 6 months ago in reply to Patrick Yao

Prodigy 80 points

around 33kHz

0 Patrick Yao 6 months ago in reply to Erfan Omidvar

TI__Mastermind 34760 points

Hi Erfan,

Is it able for you to capture the LED current waveform? And would you mind if sending me your schematic and indicate the location where did you add the probe to capture the "V over cout"?

Patrick