Circuit design help to add reverse polarity protection?

EDC Gear for Flashlight Carriers USB Power Devices
1

Hi all,

I have a baby monitor system that’s about a year old. The battery is junk so I replaced it with two unprotected 26650s in a holder piggybacked on the device. I just realized that if my wife or I accidentally insert a battery, the device will fry as it was designed with a non-removable cell in mind.
Circuit draws 150mA on standby, ~350mA with the screen on and low brightness, ~550mA high brightness.

Problems:
I have some electrical and soldering experience, but I am not an Electrical Engineer.
The monitor goes into LVP at !!!! so a standard diode would probably be too much voltage drop.
LVP with these cells is about 45% capacity remaining, so I’m looking for ways to “fix” this.
I do not know if this system can run on 5v, as it won’t power on with just USB power.

Possible solutions for RPP:
Schottky diode?
P-Mosfet circuit (I have no idea how this would work/what to get)

Possible solutions to LVP capacity issue:
Boost to 5V, buck back down to 4.0 - seems wasteful, but would work.

2

Yes a PFET, drain to batt, gate to batt- , source to load.

Edit: using a NFET on the minus side will work too and will be cheaper. Drain to batt- gate to batt+ , source to load-

3

I appreciate the info! That design seems simple to implement with the proper parts.

Do you by chance have a NFET part number I should look for? I know very little about transistors or what ratings mean for them.

4

I guess you’d want a through hole one for soldering wires directly on it. For a 0.1V drop at 0.55A would be 0.1/0.55= 0.18Ω, but since you’re short on voltage you might want even lower drop, says <0.1Ω

Here is a Mouser search
Nch, through hole, Id>0.7A, <0.1Ω, Vgs th <1.5V

Vgs th <1.5V is for filtering the one that needs a high Vgs to turn on (your battery voltage, so 3.6 to 4.2V) there are still some that need higher Vgs to properly turn on in there so you have to check the drain to source on resistance vs gate voltage plot in the datasheet.
This 100mR one for example

Or this 8mR one

There isn’t that much choice actually, low Vgs through hole mosfet aren’t very common apparently.

5

What was the OEM voltage and battery that you removed? Were they rechargeable? Does the monitor have a built-in charger? Maybe some details about the monitor would help, like the make, model and user’s manual with spec sheet.

Are the 2 26650 cells in series or parallel? How will they be charged?

You would need 12A-hrs to operate for one day if it is always ON. Seems a dc adapter that plugs into the mains would be the way to go for 24/7 operation.

6

Thank you, I’ll research my options.

We run the monitor in an auto screen off mode which “idles” at 150mA, if noise is detected it pops back on for a few minutes @ 350mA.

OEM was a single flat cell, 3.6v nominal, 2900mAH, rechargable with attached protection board.
Monitor does have a built in charger, but it was consistently overcharging the old cell to 4.24v.
Manual here: https://d2211byn0pk9fi.cloudfront.net/spree/accessories/attachments/75495/T8300_EN_manual_20190327_V01.pdf?1562639429

My mod was two 26650s in parallel, charged externally, but the battery holder they reside in can be easily reconfigured to 2S. I have a few different test boards coming from Aliexpress to test with. Charging, charging/boost, buck step down and a few other goodies. One is a 3A USB-C lipo charging board, which I might attempt to implement and eliminate removing the cells all together.

Currently the plan is to work on an reverse polarity protection circuit, but I do want to test a few things in regards to battery capacity yield and efficiency. The current battery “yield” is roughly 55% on two Vapcell k62s 6200mAH in parallel. I figure if I use a high efficiency buck converter to step it up to 5V, then buck it back down to around 4v, I’m still coming out ahead even if both are only 90% efficient as long as the standby current isn’t significant.