LOW BATTERY CUT-OFF, REVERSE POLARITY PROTECTION

[mick43]

I posted two earlier questions about this:
thread956-188710
thread240-174281

This seems like a long post, but often when I read through an Engineering Tips Forum thread, your first responses ask for more information about what you are trying to do, why you want to do it, and what you’ve tried already.

So, here’s the background:
Practically, I want to learn how to monitor and shut off 4-AA batteries in an energy efficient way, so that batteries have long run time on one charge, almost without SMD components (SOT223 is ok), this is the goal.

Goal is to use 4-AA nimh batteries to operate a portable device as many hours as practically possible, a device that has an input voltage range of 4.8-5.8V (130 ma). Because the device input and 4-AA battery output match, I guess no regulator is needed.

Question: IS THIS A SAFE GUESS? I don’t understand “surges” that might occur when power is connected or disconnected; I’ve assumed no voltage regulator is needed. That saves some voltage drop already. Is this a safe assumption? PCB components in portable device have 6.0V max rating

Low Battery shut off circuit:
To automatically disconnect batteries from portable device, to protect nimh batteries from over discharge, a simple circuit was found:

http://www.edn.com/archives/1998/091198/19di.htm

showing how to use a voltage detector with a MOSFET switch to disconnect the power. Thanks felixc (Electrical), for confirming that this 1998 circuit remains a valid design, with 1 microamp drain. This circuit used a TO-220 MOSFET, which was a bit big for the space available.

Keith Kress found a smaller MOSFET FDT457N, in a SOT223 package, it does the job when coupled with Dallas Semiconductor DS1810, 5V EconoReset with push pull output, which turns the MOSFET off when voltage drops to 4.7 volts. Thanks Automatic2 for telling me to search for “voltage supervisor.”

Thanks to all of you for taking time to steer me this far.

Reverse Polarity Protection
I searched earlier Engineering Tips Forum posts on this subject and they led to an article explaining how to use FET for reverse polarity protection:

http://www.elecdesign.com/Articles/ArticleID/9945/9945.html

I have tested this reverse polarity circuit with three different N-channel MOSFETS, the TO-92 BS-170-N-channel MOSFET (6+ ohm R-on) mentioned in the above article, and with the n-channel MOSFET FDT457N in a SOT223 package (suggested by Keith as a very low R-ON (0.06 ohm), and with n-channel IRLZ14 in TO220 package, also with low R-ON (0.12 ohm)

Question: Does the low R-on mean it drains less battery when used in reverse polarity circuit?

Question: Can one correctly measure the output of these MOSFETs with a VOM? Or, maybe one can’t? When I measure the output voltage of the above-mentioned MOSFETs when connected to the portable device, the BS170 and the FDT457N both show a voltage drop of about 0.5V from the 5.0 input, yet the device continues to operate. My middle of the road quality VOM LCD shows 4.5V. But, I know from past experience portable device will not operate at less than 4.7V (from benchtop variable voltage power supply). What am I missing here? IRLZ14 shows a drop of only 0.03V, but it's R-on is 0.12 ohms, higher than FTD457N.

Last Question: What difference does it make it you use n-channel MOSFET or p-channel MOSFET to protect from reverse polarity? The article (above)shows using a p-channel mosfet, or, an n-channel MOSFET to protect circuit from reverse polarity. Is there any advantage one has over the other?

Mick43

 

[bogeyman]

Brand new alkaline batteries can have a terminal voltage of as much as 1.65v for the first part of their life this could mean that with new cells you have a 6.6v battery, a bit over your 6v limit.

Low Ron means low voltage drop across the FET in your aplication, if the gate is biased sufficiently, you therefore need an fet with a lowish gate saturation voltage.
with ron =120 mohm at 130 mA, vds should be 156 mV ( if the bias is sufficient).

P and N types both work in the same way with of course oposite polarties, P types tend to have poorer Rds on figures and cost more than N types.

 

[loicalset]

Just something to add (correct me if I am wrong):

P type FETS require a 'negative' voltage in order to activate, which may be a problem depending on his circuit design. These would typically be used in a high-side switch (disconnect V+). N-type FETS require a 'positive' voltage and are typically used in a low side switch to disconnect ground.

 

[mick43]

Thanks Bogeyman for pointing out the over-voltage problem (6.6V) if four Alkaline cells are accidentally used, instead of nimh. The device I'm powering operates from 4.8 5.8 volts.

Is it ok to use ldos below their drop out?

Fixed 5V ldo regulators drop out at 5.2 or higher. But I want to cut-off the batteries only when they fall below 4.8V. The device I'm powering stops operating when voltage falls below 4.8, but it continues to draw about 30ma and can drain the batteries.

So, to protect the batteries, I want to cut off power when voltage falls to 4.4-4.8V, about 1.1-1.2V per cell.

Is it ok to run the ldo this way?

mick43

 

[bogeyman]

The LDO will not regulate with vin less than (vout + vreg) of course but it won't blow up either.

 

[zeitghost]

I seem to recall that LDOs don't drop out as such, just stop regulating as the input voltage drops...

The tail current goes up spectacularly though, if I remember correctly, as the thing tries to turn on the series pass as hard as it can.