This question is almost answered in "Charging li-ion cell using constant-voltage only" but I wanted to know if this is true for an extra long period of charging.

I bought a broken electric bike from my nephew and the battery pack is fully functioning. I wanted to build a simple UPS for my server using the li-ion pack(LiCoMn).

Since I don't care about slow charging and want something relatively simple, I wanted to charge the batteries via a buck regulator connected to a laptop power brick.

The charging voltage for the pack is 21V(4.2V per cell and 5 in series) and the buck converter outputs 20.5V +- 0.05V.

Theoretically since the charger cannot charge the battery past the charging voltage, there is no danger of overcharging. However, as an UPS, this device will be plugged in for long period of time and I was wondering if there is any danger to this practice.


If you try to apply a constant voltage across a heavily discharged cell it will draw an enormous current, this will heat-up the cell and will stress its structure.

This is true even if the voltage is the usual target voltage of the CV phase of a conventional charger, i.e. 4.2V±.05V per cell.

See this document from TI: Battery Charging (application note - TI literature number: SNVA557).


enter image description here

Although the text reports the charging method in the initial phase as "Current-limited CV", from an electrical POV it is just a CC charging phase.

Moreover, you have another problem: LiIon cells don't lend themselves to be charged in series safely. The voltage you apply to the pack won't be shared equally by the cells, and this could result in overstressing some cells in pack (the cells which end up having the biggest share of the voltage).

Usually you need chargers that perform cell balancing during charging, i.e. they monitor and charge the cells individually.

To further reinforce those points, see this document from MIT about handling procedures for LiIon batteries.


  • Batteries must only be charged with a charger or charging method designed to safely charge cells or battery packs at the specified parameters. Be absolutely sure that the charger settings are correct for the battery pack being charged – both voltage and current settings.
  • Never leave a battery pack unobserved during charging. Always stay in or around the charging location so that you can periodically check for any signs of battery or charger distress. Occasionally check on output levels and balancing effectiveness.
  • For series packs (2S and above) always balance charge with a charger capable of monitoring the condition of individual cells to prevent individual cells being overcharged. This charger and the battery should be put on a heat-resistant, nonflammable and nonconductive surface. Fire-safe containers designed for Li-ion batteries are available. Never place them on a car seat, carpet or similar surface.

See also this article from DigiKey: A Designer's Guide to Lithium (Li-ion) Battery Charging.

Excerpts (emphasis mine):

Charging time (for a given current) is ultimately determined by the battery’s capacity. For example, a 3300 mAhr smartphone battery will take approximately twice as long to charge as a 1600 mAhr battery, when both are charged using a current of 500 mA. To take account of this, engineers define charging rates in terms of “C”, where 1 C equals the maximum current the battery can supply for one hour. For example, in the case of a 2000 mAhr battery, C = 2 A. The same methodology applies to charging. Applying a charge current of 1 A to a 2000 mAhr battery equates to a rate of 0.5 C.

It would seem to follow, then, that increasing the charging current will decrease the recharge time. This is true, but only to a certain degree. Firstly, ions have a finite mobility, so increasing the charging current past a certain threshold doesn’t shift them any quicker. Instead, the energy is actually dissipated as heat, raising the battery’s internal temperature and risking permanent damage. Secondly, unrestricted charging at a high current eventually causes so many ions to embed into the negative electrode that the electrode disintegrates and the battery is ruined.

Recent developments have significantly improved the ion mobility of the latest Li-ion cells, allowing the use of a higher charging current without dangerously raising the internal temperature. But even in the most modern products there is still a risk in overcharging because it is a direct result of the physical make-up of the cell. Consequently, Li-ion battery makers prescribe a strict charging regimen to protect their products from damage.

Carefully does it

Li-ion battery charging follows a profile designed to ensure safety and long life without compromising performance (Figure 2). If a Li-ion battery is deeply discharged (for example, to below 3 V) a small “pre-conditioning” charge of around 10% of the full-charge current is applied. This prevents the cell from overheating until such a time that it is able to accept the full current of the constant-current phase. In reality, this phase is rarely needed because most modern mobile devices are designed to shut down while there’s still some charge left because deep discharge, like overcharging, can damage the cell.

In all this, see the focus of not overcharging the cells. Without a current limiter a CV supply cannot guarantee that the maximum current won't be exceeded, since the charging current depends on the actual voltage applied to the cell and the internal state of charge. Even applying a voltage well below the nominal cell voltage (3.7V), say 3V, to a deeply discharged cell (say, 2V open-circuit voltage) could cause too much current to flow.

Unless you can guarantee that each cell in the pack is in a well defined state of charge before being connected to the CV charger (and this state is the same for each cell in the pack), you put the pack and yourself at risk.

  • \$\begingroup\$ Yes, I think I need some kind of balancing board to charge these batteries. I would also need a way to set the maximum current, a resistor or even a MOSFET. Also, I think I need a over-discharge protection as well. A BMS may solve most of these problems. \$\endgroup\$ Apr 24 '18 at 5:04

"The charging voltage for the pack is 21V(4.2V per cell and 5 in series) and the buck converter outputs 20.5V +- 0.05V.

Theoretically since the charger cannot charge the battery past the charging voltage, there is no danger of overcharging. However, as an UPS, this device will be plugged in for long period of time and I was wondering if there is any danger to this practice."

For those who, like me, find this now that it's been a few years...

Theoretically there is great danger to allowing that voltage to remain connected to the battery pack. Lithium Ion cells don't charge by voltage, but by current... and at that applied voltage - some current will always flow into the battery pack.

Charge voltages of 4.1/cell are MUCH safer on the cells, and will extend their usable life significantly.

All batteries have internal "leakage current" above a certain voltage threshold... meaning that when you are applying any higher voltage than the specified "operating voltage" (typically 3.7 to 3.85v/cell), you are into "charge conditions".. and causing current flow.

Once all available lithium ions have been moved back to the negative Anode material, any further application of charge current (even if it's only microamps) will cause lithium ions to continually move deeper into the graphite anode structure. This causes the lithium ions to become "tightly grouped together", which once again becomes metal (aka: "lithium metal plating")... and this metal lithium is no longer useful - which permanently degrades the capacity of the cell. It also becomes a point of attraction for additional ions, which causes the metal clump to grow - always towards the separator membrane, and possibly becoming a dendrite that can cause high internal current leakage... heating... pressure... and then explosions.

If you were privy to the exact chemical construct of the cell(s), and knew what the precise "operating voltage" was (typically 3.65 to 3.85V), you could allow THIS voltage to remain connected indefinitely... however the ability to charge at that voltage would be reduced significantly... on the order of maybe 30% of the original capacity... but it would last many thousands of cycles.

  • \$\begingroup\$ Interest only: Mars rovers manage about 8000 cycles at afair 3.8 V and somewhere under 50% of full capacity. \$\endgroup\$
    – Russell McMahon
    Jul 26 at 5:09

If you regulate CV to the recommended float voltage (non-charging voltage) but not 4.2V per cell. This may reduce rated capacity by 10% or so but then discharging in <1h may reduce the 10h capacity much more.

Conclusion —

Yes you can prematurely wear it out.


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.