I have a battery pack of 2 18650 cells in series, so it is possible for me to charge it without connecting anything to the central 3.7v potential point as many answers mention? The battery in question is https://uk.rs-online.com/web/p/lithium-rechargeable-batteries/1449410/, which is sealed and has only 2 wires coming out. This means that I cannot access the central part. I understand that a 7.4v charging circuit is most appropriate, but is it possible for me to implement 2 TP4056 boards in series somehow? I wouldn't be able to do a setup like


simulate this circuit – Schematic created using CircuitLab

given that the grounds would be shorted, and thus so would the second cell. So, does this mean I have to resort to a specialised 7.4v circuit, or could I manage with a 3.7V circuit (that I have 3 spares of with me currently, which is why i want to use it).

Furthermore, the datasheet on Rapid Electronics for that battery pack lists that it has over-charge and over-discharge protection - does this mean I can hook it up to a 8.4V (safe current limit of, say 500mA once near full) bench power supply? This would be even more convenient to charge, but I do not want to damage the battery if this is dangerous.

Many thanks.


page three of the data sheet has the internal circuitry of that battery pack.


enter image description here

the chip https://pdf1.alldatasheet.com/datasheet-pdf/view/736555/SII/S8252.html doesn't seem to have the capability to balance the charge on the cells - VC is an input only. I don't know if that will cause problems or not.

but either way it will be safe to use a voltage and current limited source to charge the battery pack.

  • \$\begingroup\$ So I can use a bench power supply at 8.4V (limited to something safe like 300mA) to charge it? Also, would a circuit like mouser.co.uk/new/dfrobot/dfrobot-dfr0564-usb-charger be useful for this type of battery (with internal circuitry)? \$\endgroup\$
    – QuickishFM
    Jan 26 '19 at 21:59
  • 1
    \$\begingroup\$ yes to both those suggestions \$\endgroup\$
    – Jasen
    Jan 26 '19 at 22:02
  • \$\begingroup\$ Great, bench power supply it is until I can find a cheap source of the circuit. Thanks for your help \$\endgroup\$
    – QuickishFM
    Jan 26 '19 at 22:05
  • \$\begingroup\$ Please note: The bench supply will charge the cells, but it is not a recommended way to charge them. Actual chargers have other features besides CC/CV control. They detect when end of charge has been reached and when that happens, they completely turn off all charge current. Furthermore, they have backup safety timers. If end of charge is not reached, they will turn off all charge current. After transitioning to CV mode, the charger monitors charge current. End of charge is defined to be the point where the charge current drops to some low level (maybe 25mA... varies with cell capacity). \$\endgroup\$
    – mkeith
    Jan 27 '19 at 3:50
  • \$\begingroup\$ I have on many occasions charged lithium ion cells with CC/CV supplies. But only when I am sure I will not forget and leave them connected to the supply for an extended period. It is definitely not safe to leave lithium ion batteries connected to 4.2V supply for an extended period of time. (I mean 4.2V per cell). \$\endgroup\$
    – mkeith
    Jan 27 '19 at 3:52

Jasen's answer includes a good solution.

Consider the requirements.

If Vbat = CC to 4.1 then CV with cut-off at 10% CC and say mAh capacity difference between cells reaches 10%, one cell will reach 4.1V before the other so as the charger target is now 8.2V the weaker cell is now getting overcharged while the stronger one is still catching up.

Consider Vbat must has an ideal Vmax of 4.1V and a limiter of 4.13 per cell then the tolerance is 30mV of overcharge. Depending on use tradeoff of capacity and life time, the overcharge voltage accelerates ageing exponentially thus making the weaker cell even worse yet increases their capacity only up 10% from 4.0V

Hence a Balancer reduces the rate of aging or cell mismatch because the pack fails with the weakest cell.

So let's examine the S8251 Seiko specs inherited by ABLIC

Overcharge detection voltage n

(n = 1, 2):3.55 V to 4.6 V (5 mV steps) ±20 mV (+25°C), ±25 mV (-10°C to +60°C)

Overcharge release voltage n (n = 1, 2):3.15 V to 4.6 V*1 ±30 mV

Overdischarge detection voltage n (n = 1, 2):2.0 V to 3.0 V (10 mV steps) ±50 mV

Overdischarge release voltage n (n = 1, 2):2.0 V to 3.4 V*2 ±100 mV

Discharge overcurrent detection voltage:0.05 V to 0.4 V (10 mV steps) ±10 mV

Load short-circuiting detection voltage:0.500 V to 0.900 V (50 mV steps) ±100 mV

Charge overcurrent detection voltage:-0.4 V to -0.05 V (25 mV steps) ±20 mV

Charge overcurrent detection function “available” / “unavailable” is selectable.

  • \$\begingroup\$ Since my battery pack is in series straight from the manufacturer, is it likely to cause this much damage still? It's very likely that they are batteries from the same batch, and will be used/charged with the same current each time, so the level of wear is likely to be identical - is it safe to still use a bench supply? I know that a dedicated circuit is imperative for a long lasting product, but this is for a short time for a robot so I am not too interested in the longevity of the cells at the moment. \$\endgroup\$
    – QuickishFM
    Jan 26 '19 at 22:33
  • \$\begingroup\$ @QuickishFM OK as long as you know the quality of a pack is unknown yet related to the 0.x% mismatch in capacity and ESR which accelerates with CC and CV rate relative to OEM recommended values. Such CV charge level and cutoff threshold is a tradeoff between that extra 10% of capacity and accelerated rate of wearout. For example using 4.0V CV max and 40% SoC cutoff means you get more than 4x to 10x as many charge cycles but using only half of the capacity. Your mileage may vary with quality ref Battery University \$\endgroup\$ Jan 26 '19 at 22:37
  • \$\begingroup\$ It would be a good idea to measure the pack capacity now while the pack is new and record it. Occasionally re-measure it. If it declines dramatically, stop using it. That is an indication that something has gone wrong. Also, if you have the ability, measure series resistance at some particular state of charge. Re-check that from time to time. If the series resistance goes up, that could be a sign of a problem. Series resistance naturally varies somewhat with state of charge. So be consistent. \$\endgroup\$
    – mkeith
    Jan 27 '19 at 3:59

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.