I am using a TPS2097 to control four loads all of which are less than 50mA
The control inputs to the switch are driven with 3.3V logic from an Atmel microprocessor.
3.3V is connected to IN3 of the device (pin 6)
4.5V is connected to IN4 of the device (pin 7)
The control input on EN3 id high (pin 8) which is disables OUT3.
When the control input on IN4 (pin 9) goes low enabling OUT4, the output OUT4 (pin 10) only goes to about 2,4V and does not turn on the load. There is about 200mV of low frequency stuff on it and it looks to be floating. This behavior is intermittent. The load is a 50mA fan.
I put a 2K resistor on the output to ground, assuming that the high side FET in the chip did not have a adequate reference which did not fix it. This resistor is not now in the circuit.
I put a 10K pull up on OCB (pin 12) to see if the device thinks that there is a fault. It stays high at all times.
The block diagram shows a block called Vcc Select and UVLO from which the charge pump is derived. This block is shared between channels 3 and 4. Are the voltage inputs on IN3 and IN4 allowed to be different? since I ave 3.3V on one and 4.5V on the other? Or is the device smart enough to pick the correct charge pump source for the channel it is driving?
The portion of the circuit in question is marked in red in the attached file.
This problem has shown up on all of the 20 boards that have been made so far.
To TI
SR # 1-1895053512 from Consultant regarding TPS2097
Thank You
Tim Woolaver
