voltage regulator bypass

Discussion in 'The Projects Forum' started by max2, Jul 24, 2006.

  1. max2

    Thread Starter New Member

    Jul 24, 2006

    I was going to design a general microcontroller board using an ATmega controller and some of the usual interfaces such as RS232, RS485, LCD and keypad. I have an ordinary 7805 voltage regulator onboard, and would like to have the possibility to feed the circuit with an external stable 5V supply from a switched power supply and bypass the 7805. I would like to be able to have an overvoltage protection built in on this 5v supply using a MAX4866 or similar. These devices switch a MOSFET on/off depending on the voltage trip level that they are prefabricated with. My plan was to use either of these devices in order to switch off the power supply if something else than 5v was fed into the circuit.
    But the problem is the current limiting resistor to the LCD backlight. It's only at 3.5 ohms according to the datasheet and would make the rest of the circuit see only this low impedance. Right?
    So the ON resistance of the MOSFET would contribute to a voltage drop from 5V to around 4.7V with a ON resistance of 0.2V max. Calculating with lower onresistance doesn't help much, so I guess some other solution must be used in stead. I can't figure out how to solve this I'm afraid, but there's probably a better way of doing this.
    Does anyone have any suggestions for changes or another solution?
  2. mrmeval

    AAC Fanatic!

    Jun 30, 2006
  3. hgmjr

    Retired Moderator

    Jan 28, 2005
    Greetings max2,

    mrmeval is right, the 3-terminal regulator with an SCR crowbar circuit is the easier and cheaper solution. However, If you are still interested in using an SMPS and you haven't already procured the +5V switching power supply maybe you can kill two birds with one stone and buy a switcher that has the over-voltage protection feature built in.

    As for the 3-terminal regulator, my experience has been that they can become significant heat sources since the voltage dropped across the regulator times the current being delivered to the load is power that gets dumped as heat in the regulator. This heat can be dissipated by using a reasonably efficient heatsink if the heat is not too large.

    If you decide to use a 3-terminal heatsinks you may want to consider using several of them. You can divide your circuit up into smaller sub-circuits and then power each circuit with its own 3-terminal regulator. You will need to consider some sort of fail safe scheme that senses when one regulator fails and disables the remaining regulators.

  4. max2

    Thread Starter New Member

    Jul 24, 2006
    I think I have decided to go with the crowbar solution where a thyristor short circuits the output of the voltage regulator. Unfortunately all fuses I have seen demands that the current be several times the rated current before it breaks quickly and the max short current output from the voltage regulator is 2.2A. I guess the board takes 700-800 mA under worst case and the fuse must handle that at the same time and not accidentally blow under "normal" transients before the regulator has seen any load. But I have never used a fuse before in a circuit like this so I don't exactly know how they act under real conditions.

    Thanks for the help