1
\$\begingroup\$

I'm planning a hobby project with the following characteristics

  • 3.3 V circuit, battery operated (by regulating standard AAA batteries)
  • Most of the time in "sleep" mode, so I need a very small quiescent current
  • When operating it will consume 100-350 mA
  • Linear regulator to avoid noise and ripples

I've chosen the MCP1825 linear LDO regulator because it has enough power (500 mA), very low dropout voltage (less than 210 mV) and reasonable quiescent current (less than 120 µA).

My questions are:

  • What if I don't put any reverse current protection? Will the PMOS pass element in MCP1825 provide some protection?
  • Could I avoid adding reverse current protection if I'm careful with batteries' polarity?
  • If I have to put some sort of reverse current protection, how would you advice designing a simple one with very low (or non-existing) dropout voltage and quiescent current?
\$\endgroup\$
6
\$\begingroup\$

You only need a reverse polarity protection for your device if there is a possibility that the power input can get applied in the wrong direction. You would have to ask yourself how likely it would be that a reverse polarity condition could happen and then decide from that whether you need protection for that scenario.

Most simple reverse polarity protection schemes are going to involve the use of a diode in the input power path. You can select a Schottky diode to minimize the forward voltage drop so that you can get more usable voltage range from your battery supply. There are diodes available that have pretty low forward voltage. For example a generic SB30 power diode would come in at a Vf of ~400mV @ 100mA.

If you do not mind getting a bit more complicated you can also use a discrete PMOS FET to provide reverse polarity protection like shown here. One recommended P-FET for this circuit could be the Vishay Si2323 which would show only .068 ohm resistance when the battery voltage was at 1.8V or more. FET selection needs to be done carefully to find one that can be fully on at the minimum battery voltage. At the 1.8V this FET can easily support the 500mA rating you called out for your LDO regulator.

enter image description here

The MCP1825 is not going to provide reverse polarity protection for your device. If you look at the functional block diagrams in the data sheet you will see that there is a diode inside across the internal PMOS FET that would forward bias in a reverse polarity situation and allow a negative bias to be applied to any down wind circuitry.

enter image description here

\$\endgroup\$
  • \$\begingroup\$ Nice answer. (+1) I only wanted to add that the other way I've done reverse polarity protection (really over voltage protection.) is with a series poly fuse and a diode in shunt. (Diode in parallel with power supply.) The diode conducts with reverse voltage, and then the poly fuse trips limiting the power dissipation in the diode. For over-voltage protection replace the diode with a zener. \$\endgroup\$ – George Herold Jan 7 '15 at 14:38
  • \$\begingroup\$ There are two issues with the diode & poly fuse implementation. First off it exposes the down wind circuits to a negative voltage level that may not be tolerable for some circuitry. Secondly the poly fuse can take some time to open up. This exposes the circuits to the reverse diode level voltage for whatever 10s or 100s of milliseconds it takes for the poly fuse to act. I normally just use the series SB30 diode. Many of the modern battery technologies have a fairly flat discharge characteristic and the amount of extra time of battery life one gets with a lower drop solution is not that great. \$\endgroup\$ – Michael Karas Jan 7 '15 at 15:00
  • \$\begingroup\$ You can't really go wrong with a Schottky diode in series; the forward voltage drop is insignificant for 4.5V worth of AAA battery, and it's a single run-of-the-mill component. Michael's second solution is very elegant (I haven't come across that before) but for your situation I'd say it's overkill. If the power supply were higher, say 12V+, then you begin to approach the reverse breakdown voltage of a (cheapo) Schottky. This is where I'd use the much more robust PMOS FET option, especially as then the choice of FET would be easier due to the higher gate voltage. \$\endgroup\$ – CharlieHanson Jan 8 '15 at 15:28
  • \$\begingroup\$ @chaaarlie2 - The drop of the Schottky diode can pose more than an insignificant amount. Let's say that three AAA batteries are 1.2V NiMH batteries used at a nominal 100mA load when an MCU is active. A low cost diode such as a BAT54 will have nominal Vf of 0.5V at 100mA. At 0.5/3.6 that computes to 13.8% of the energy supplied by the battery getting lost as heat in the diode. Another thing is that a BAT54 has a -30V reverse voltage rating so it is a better choice for a 12V battery solution. The reason being that a P-FET like the Si2323 has a gate voltage range of only +/- 8V. (continued) \$\endgroup\$ – Michael Karas Jan 8 '15 at 16:15
  • \$\begingroup\$ (continued from above) This means overall that the P-FET is a winner for low voltage applications. For higher voltage applications such as 12V or 24V applications a diode would be needed and the Vf of the diode is a much lower percentage of the battery voltage. The SB30 type diode that I have recommended can even be used there because it has a max reverse rating of 40V and it has the advantage of an additional 0.1V Vf savings over the BAT54 at 100mA. There is no doubt that the P-FET solution costs a little bit more in part price. It really is east to implement. Si2323 is SOT-23 package. \$\endgroup\$ – Michael Karas Jan 8 '15 at 16:25
1
\$\begingroup\$

It is possible to construct a battery holder that will not connect if the cells are inserted the wrong way. If you are in control of making the enclosure, this may be an option.

\$\endgroup\$
  • 1
    \$\begingroup\$ The usual way is to have a couple of plastic lugs each side of the positive terminal. This prevents the flat negative terminal from ever touching it if the battery is inserted in reverse. \$\endgroup\$ – Simon B Nov 27 '15 at 21:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.