I'm designing a product that will be powered by a single 18650 cell.

Specifically, I am basing my numbers off this cell in particular:

Samsung INR18650-25R

It has a typical full voltage of 4.2V and cut-off of 2.5V.

I've looked through TI, ABLIC, Diode Inc. and a few others for battery protection ICs for single cells and noticed that the OVP parameter for offerings that have a UVLO of 2.5V are typically higher than 4.2V by anywhere from 0.075V to 0.125V. Is this so the chip does not cut out too early when the battery reaches a full voltage of 4.2V?

If so, would an IC with a OVP of 4.325V be safe to use with a cell that has a full voltage of 4.2V?

Sorry if this question has a simple answer, this is my first time designing a full product, and want to understand fully before moving forward with my design.

I am also using the BQ24075 with the BQ27441 for charging and fuel gauge. This is from the reference design of the Battery Babysitter breakout from Sparkfun:

Battery Babysitter from Sparkfun

  • \$\begingroup\$ I'd be contacting someone in Samsung field service, myself. They can probably help you work out a successful approach for their battery. They may also be able to suggest alternative batteries that may be better for some IC you are considering, too. You should contact them and give it a go. See if they help. If not, let us know what happened. Otherwise, they may be your better bet. \$\endgroup\$
    – jonk
    Nov 16, 2018 at 6:44
  • \$\begingroup\$ Do you plan to give the user of your product an access to the battery cell (removing it and installing it), or the cell will be embedded and charged/managed solely inside your device (without user ever removing the cell)? \$\endgroup\$ Nov 17, 2018 at 20:44
  • \$\begingroup\$ Samsung will not even give you the time of day if you call asking about their batteries. But you may get help from a distributor or reseller. \$\endgroup\$
    – user57037
    Nov 18, 2018 at 1:04

3 Answers 3


The purpose of a protection circuit in a single-cell pack is to provide backup protection in the case that the charge or discharge circuitry should have some kind of fault. This is why the protection circuit overcharge threshold is a bit higher than the recommended maximum charge voltage. This is normal. The idea is that we don't want the protection circuit to interfere with normal charge and discharge. We want it to operate only if the charge and discharge parameters are exceeded.

One thing that jumps out at me is that the cell you have chosen is a power tool type cell. In other words, it is specifically designed to deliver high power, and can be safely discharged at relatively high rates (e.g., 20A). Does your application require such high currents? If so, careful design is required for the power path. If not, you may consider looking at other cells (so-called laptop cells).

When you select a protection circuit, it is essential that you take the normal charge and discharge rates into account. Most single-cell protection IC's I have seen are not designed to support such high currents. But if you are using the BQ24075, then it seems that you are not intending to support high discharge rates. As a side note, the BQ24075 is NOT a charge protection IC. Just google "Seiko charge protection IC" to find examples of what a protection IC is.

I am not sure if it is still the case, but for a long time, Seiko basically owned the market for single-cell protection IC's. They have thousands of parts and you can fine tune over and under voltage cutoff points over a wide range. Someone commented that protection IC's are normally used by battery pack vendors, and this is true. But I guess there is no reason why you could not add the protection circuit yourself. But there is more to it than just the protection IC. It is a big topic.

Cell cycle life can be greatly extended if you keep your max charge voltage to 4.1V or so, in which case, you could possibly use a protection circuit with a 4.2V or 4.25V cutoff. Just make sure the protection IC cutoff is a bit higher than the charger voltage so that they play nice together. If you only charge to 4.1V, you will not get full rated amp-hours from the cell. So it is a trade off. I think you will get around 90% at 4.1V.

  • 1
    \$\begingroup\$ This is a great answer. You bring up a good point with the battery. My application is not going to be drawing anywhere close to 20A, closer to 300mA while transmitting over Bluetooth. As for the BQ24075, I know this does not offer protection features which is what I was alluding to. I've seen plenty of options for protection IC's, my main concern was that most were showing a OVP of higher than 4.2V, but you've reassured me that this is because these are meant for backup protection, so it should be slightly higher than the battery voltage to avoid interfering with the charge cycle, thanks! \$\endgroup\$
    – seggles
    Nov 17, 2018 at 18:06
  • \$\begingroup\$ "there is no reason why you could not add the protection circuit yourself" Wrong advice. The reason for cell protection is to avoid mishandling of cells when they are removed from equipment and charged separately. Since the OP device does include the charger (and hence the cell has no need to be removed/re-inserted), there is no reason to bloat hardware by adding this dedicated circuitry (and two MOSFETS), all protection can and should be implemented using the charger (overvoltage-overcharge) and gauge monitor (over-discharge). \$\endgroup\$ Nov 17, 2018 at 21:33
  • \$\begingroup\$ @Ale..chenski, it is absolutely industry standard practice for battery packs to have a protection circuit built in regardless of whether the pack is user accessible or user replaceable. It is not only for mis-handling by end user but also as a secondary protection in case the battery charger circuit fails. Normally it is incorporated inside the pack by the pack vendor. I admit that this is a slightly different scenario, but I don't think it is bloat, and there is still a case to be made in favor of it. \$\endgroup\$
    – user57037
    Nov 18, 2018 at 1:01

The BQ24075 does not offer any feature to prevent the cell from over-discharging though.

After reviewing the datasheet for the BQ24075, it does mention a typical UVLO of 3.3V. Seems awfully high when the cell can go down to 2.5V

You may be confusing the protection on the INPUT voltage pins with battery voltages.

You also seem confused on the cutoff. A battery charger is going to try and charge a dead battery so there is no low battery voltage cutoff.

The only battery voltage a charger must know is the max charge voltage.

You may be thinking of low battery indication. Some chargers have a LBO that can be used to shut down or reduce the load. The BQ24075 is a basic no frills charger targeted at small devices where real estate is at a premium. Features take pins and consequently, real estate. If you want a more full featured charger look at the MCP73871

BTW: Just because the cutoff is 2.5V does not mean you should allow the battery to discharge that low. I would not discharge below 3V. Discharging below 2.8V can damage the battery.

  • \$\begingroup\$ +1 from me. For UVLO, and for 3.0V cut-off. Also, since OP uses I2C gauge, there is no reason not to use I2C and other much more advanced parts from BQ24xxx/BQ25xxx families, like BQ24297, with every voltage monitoring over I2C, and various means to connect-disconnect battery power path. \$\endgroup\$ Nov 17, 2018 at 23:57

"Battery protection IC" are used by Li-Ion cell manufacturers, and are built into the cell itself. This is for protection and safety compliance for Li-Ion cells to be used in consumer equipment as loose batteries.

Since you are making a device with unprotected cell, I assume that the cell is embedded (soldered/welded) inside your device. (If not, you can't let your customers to have loose unprotected cells). Therefore, the BQ24075 and gauge monitor BQ27441 is your protection, you just need to set it up properly to meet the cell charging specifications. You don 't need any additional "protection" IC.

As for the particular question, "would an IC with a OVP of 4.325V be safe to use with a cell that has a full voltage of 4.2V?", hell no, the cell must be charged at nominal voltage listed in its specification. You could overvoltage this parameter, but this will lead to vastly reduced cycle life of the cell.

  • \$\begingroup\$ The BQ24075 does not offer any feature to prevent the cell from over-discharging though. You'll notice on Sparkfun, they specifically tell you, the product is used only for cells that have built-in protection circuitry. It could be argued that the BQ24075 would prevent an over voltage situation since it's charging voltage is 4.2V, so the potential of the battery would not go over this, however, there is nothing to stop the battery from discharging past 2.5V. I could potentially shut down the device through software once the fuel gauge shows 2.5V, but that seems unreliable to rely on software. \$\endgroup\$
    – seggles
    Nov 16, 2018 at 13:21
  • \$\begingroup\$ After reviewing the datasheet for the BQ24075, it does mention a typical UVLO of 3.3V. Seems awfully high when the cell can go down to 2.5V. \$\endgroup\$
    – seggles
    Nov 16, 2018 at 13:46
  • \$\begingroup\$ The BQ24075 is not a protection circuit. It is a charge controller and power-management IC. It is industry standard practice to have a protection circuit (or BMS in multi-cell packs) integrated with the pack. The BQ24075 should NOT be integrated with the battery pack (it will add too much heat). It should be on the system PCB. \$\endgroup\$
    – user57037
    Nov 17, 2018 at 17:36
  • \$\begingroup\$ @mkeith, I see nothing in the OP question that the intent is to make "an integrated battery pack". \$\endgroup\$ Nov 17, 2018 at 21:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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