2013-05-20 03:11 AM
Hello,
I'm currently designing my own schematic based on the STM32F4DISCOVERY board.The two BAT60 power supply protection diodes (D1 & D3) are not relevant for me because they generate a dropout voltage: +5V is in fact +4.7V and +3.3V is +3V.I want a full +5V to supply my analog functions (based on opamps) and a full +3.3V to allow the STM32F407 to generate/convert 3.3V signals with the DAC/ADC. (my only board power supply is the USB VBUS)Are the BAT60 really critical concerning the USB power supply protection ? I need to remove them...Thanks !2013-05-20 06:07 AM
Without the diodes, it is up to your design to ensure that the USB does not see more than 5V and the F4 does not see more than 3.6 V.
Cheers, Hal2013-05-20 08:55 AM
Thanks !
Is there a cheap solution to protect these power supplies from highter voltages, without decreasing the voltage ? For example with zener diodes...2013-05-20 04:50 PM
Why not use a 3.6V regulator instead of a 3.3V regulator to get the desired ADC range of 3.3V ? Convert internal channel 17 to get the actual reference voltage.
I am puzzled about using 5V for the analog functions. Are you mounting the board on a motherboard with analog amps ? Cheers, Hal2013-05-22 08:05 AM
The 3.6V regulator can be a good solution.
The +5V of the USB is enough for my application (amplifing the analog DAC signals outputs). Yes, I already mounted the discovery board on a motherboard with analog opamps, but now I want to create a single board with the uC and analog part, all powered by the USB 5V, because it's a part of my specifications...2013-05-27 11:42 PM
You can implement polarity protection with a MOSFET rather than a diode.
To protect from the positive rail, use a P-channel MOSFET. Connect the positive input to the drain, the load to the source, and the gate to ground. This is 'backwards' from the usual configuration. The device initially conducts because of the body diode. With higher voltages the gate turns on allowing resistive flow, just as with 'forward' operation. An N-channel MOSFET could do the same job on the negative power side. N-channel devices are more efficient (either lower cost or better specs for the same cost), but sometimes a design that disconnects the negative/ground isn't suitable. Extra low dropout voltage regulators are available. They are based on MOSFETs rather than bipolar transistors, thus operating with less than the 300mV drop of a saturated bipolar transistor.