TPS92641EVM: IADJ to IOUT ratio not constant

Robin, please help in this case.

Hi Martin,

For low IADJ, the TPS92640 dims well but may have more variation between circuits. So I suggest you use PWM diming to achieve small output current.

Besides, for analog dimming, it's most accurate when device is in CCM mode. BTW, would you mind introducing your application simply? Thanks!

Robin

Hi Yuting Shao.

PWM is not wanted because of SVM requirements, we want to achieve low analog dimming. This is why this device was recommended to me.

Regardless of CCM or DCM, the average inductor current across Rsns should be proportional to the IADJ setpoint.

Our application is residential lighting. We want to dim down to about 50 ... 100 mA current.
Technical Data:
Vin 48...50 V
I,out 750 - 100...50 mA
VLED appr. 36 V at 750 mA, 32 V at 100 mA

Hi Martin,

The inductor current ripple is related with the inductor value, switching frequency, the voltage difference between input and output, the on-time setting and so on. Pls keep the half of inductor current ripple is lower than average current to maintain CCM mode.

http://proddmsvm.itg.ti.com/viewdoc/slva419b.pdf

I think you can refer this AN and try with the third method with TPS92640 to achieve deeper analog dimming.

Robin

Yuting Shao said:

The inductor current ripple is related with the inductor value, switching frequency, the voltage difference between input and output, the on-time setting and so on. Pls keep the half of inductor current ripple is lower than average current to maintain CCM mod

Of course you are right, that CCM is dependent on those factors. However the set point for the LED current is given by VADJ / 10 / Rsns. But in my application it seems that the 10 factor is not constant and  not 10.

Thank you for the application note. This shows interesting ways to analog dim the LED current.
What are the advantages over using IADJ of the TPS do influence the current?

Martin