Does ADPD103 can drive a LED ( 5V~7V@20mA ) by the following Schematic？ And the VCC is 8V.
The maximum voltage that can be applied at the LEDX pin is 3.6V. If the voltage of the pin is pulled above that voltage it could result in damage to the device. There are two things to watch for, the first is what the forward bias voltage of the LED will be when the driver is at 20mA. The cathode of the LED will need to be <3.6V when the LED is on. The second thing is what will the drop across the LED be when the LED driver is OFF. There will be some leakage through the LED driver and the ESD diodes that will cause some voltage drop across the LED, but lower than the turn-on voltage of the LED. Typically, this will be on the order of nanoamps. With green LEDs operating on a 5V VLED, for example, the voltage drop of the LED is large enough due to this leakage current that the 3.6V maximum voltage of the driver pin is not violated. You would need to experiment with your particular LED to see if this is the case.
For some reason I assumed that we were talking about the ADPD105 since it's the latest device in the family. If this is a new design, then you should be looking at the ADPD105 (or ADPD107 if you wanted SPI). The ADPD105 is backwards compatible. There is more detail included in the datasheet with respect to some of the registers. For example, register 0x42/44, you will see the detail on bits [5:4] for the TIA_VBIAS
Sorry for the confusion.