1
\$\begingroup\$

According to the Battery University, charging a Li-ion cell to just 3.92v is the optimal voltage to avoid stress and lithium plating. Problem is, nearly all consumer chargers focus on going to 4.2v for longest battery life.

When charging, should you disconnect the (consumer) charger when the charging cell voltage reaches 3.92v or charge until the cell hits the end of its constant current stage at 4.2v and would subsequently settle down to around 3.92v?

Sources:

Charging Lithium Ion batteries

How to prolong Lithium based batteries

\$\endgroup\$
  • \$\begingroup\$ A LiIon cell charged to 4.2V WILL NOT "settle" to 3.9x volt in a few hours. LiIon maintain terminal voltage extremely well (unlike most other chemistries). || Battery University BU808 table 4 shows capacity versus Vchg max. Rule of thumb - capacity drops 10% per 70 mV reduction. | 3.92V ~~= 58% capacity and 2400-4000 cycles. | "Somewhere [tm] you'll find a superb long Mars Rover battery system design paper - they get maybe 8000 cycles - at AFAIR 3.65V max? ||| ... \$\endgroup\$ – Russell McMahon Apr 12 '18 at 12:04
  • \$\begingroup\$ ... Wow - Mars orbiter - 30000 cycles at 30% DOD full paper here || In your case - adapting a charger to stop at 3.9ish V will pay dividends for cycle life. You MAY be able to add a series Schottky diode with a parallel resistor so Vbat ~= 0.4V below Vsense BUT this is liable to be hard due to charger design and a added "stop at xxxV is probably easier. || Your attitude to Ali Chen will not do you much good. He put substantial effort into trying to help you. "Thanks" as you ... \$\endgroup\$ – Russell McMahon Apr 12 '18 at 12:10
  • \$\begingroup\$ ... gave it will drive people away wholesale. A brief polite comment may be in order but 'getting on his case' will do YOU no good. To get the rep he has you have to be helpful and useful to a large number of people over substantial time. | Welcome to SE EE - it's a valuable resource but a bit of tolerance of the crew will be of great benefit to you. There are some great and helpful minds here. Try to survive long enough to benefit from them :-). \$\endgroup\$ – Russell McMahon Apr 12 '18 at 12:13
1
\$\begingroup\$

I understand your question. I did something similar in one of my projects some years ago. I wanted to prioritize cell life over runtime, so I just charged the cell to 4.1V and then disconnect the charger. The constant voltage charging phase is what makes you able to have the maximum runtime, so in this case my opinion is that you can go without it.

I actually don't know where the 3.92V come from, but it is not uncommon to use Li-ion cells in a narrower voltage range in order to improve battery life. Some EV manufacturers, for example, start with a narrower voltage range which guarantees the declared vehicle range. Then, as battery ages and loses capacity, the voltage range is progressively extended to compensate and to maintain the original range.

\$\endgroup\$
  • \$\begingroup\$ I see. Also, I hadn't considered extending the voltage range as age compensation, may be useful in my case. Thanks for your help! \$\endgroup\$ – I Used To Be Funny Mar 6 '18 at 23:33
1
\$\begingroup\$

The question is a bit messy.

First, "optimal" might means different things. Charging Li-Ion to 3.92V will likely yield 70-80% capacity. Is it optimal? It depends on battery use/discharge pattern. If the device is meant to be in "hot reserve" like emergency flashlight, 80 % is optimal. If the device is meant to start working at full power immediately (like LED lights or quadcopter), then you wasted 20-30% capacity for no reason.

Second, "consumer chargers" do not "focus" for longest battery life when charging to 4.2V. The magic "4.2V" is because of mainstream battery specifications. And this level is determined by battery manufacturer, which is based on standard expectations of having 500-1000 charge-discharge cycles before rated capacity drops below 70% level. Due to improvement in technology and materials, there are Li-Ion cells that can be charged with 4.35 V while maintaining the same or better cycle life.

Third, if you charge a cell to 4.2 V, it might take a while before it self-discharges to 3.92 V, which makes no sense.

In general, there is certain (complex relationship) trade-off between charger float voltage, termination current, rate of charge, rate of discharge, shell time in fully-charged state, shell time in discharged state, and SOH - State of Health of battery. Siting in a fully charged state reduces battery cycle life, and sitting in fully discharged state has similar effect. So there is no universal recipe how manage "lifespan" of a Li-Ion battery.

EDIT: Based on studies of some cells from some manufacturer of 7-12 years ago (manganese-based cells for Nissan Leaf), it was found that charging level of about 3.92 V provides the best balance between two major deterioration mechanisms - build-up of SEI (Solid Electrolyte Interface, which accelerates at high charge levels), and EO (Electrolyte Oxidation). That's where the magic 3.92 V number came from. However, Li-ION cells are under continuous improvement, with different anode-cathode-separator materials and slight modifications of electrolyte chemistry, and the parametric space for exploitation of cells is vast, this magic number might be not universal. More, using 3.9V leads to about 60-65% usable capacity.

For practical purposes, if someone wish to employ 3.92 V charging schema, simple disconnect at 3.92 would lead to additional loss of capacity, since the CV part will be missing. To get proper charge to 3.92V, chargers must be re-programmed for new set of parameters. Most of modern ICs do have the ability to be controlled over I2C interface, or with external hardware biases.

\$\endgroup\$
  • \$\begingroup\$ Capacity is not an issue in this case, I would like to preseve the cells for as long as possible. Some of your information contradicts Battery Universities findings does it not? Ive added links to their articles above. It explicitly says "Chargers for consumer products go for maximum capacity and cannot be adjusted" and cells charged to 4.2v get an average of just 300-500 cycles before losing capacity. Still, all of this still doesn't clear up whether they recommend disconnecting at 3.92v (because the cell would naturally drop voltage after just a couple of hours) \$\endgroup\$ – I Used To Be Funny Mar 5 '18 at 4:27
  • \$\begingroup\$ You need to concentrate more carefully on context. Maximum capacity is at 4.2v... I have not seen many/if any consumer li-ion chargers that specifically stop at anything less than around 4.2v, have you? What I am looking for is clearly for a charger to stop around the end of the CC phase (re-read my original post). It has already been stated that capacity is not an issue, please stop wasting space in the forum with repetitive unnecessary information. Your petty comments have been unhelpful and irrelevant to my original question. \$\endgroup\$ – I Used To Be Funny Mar 5 '18 at 5:38
  • \$\begingroup\$ @IUsedToBeFunny, you should worry less about forum space, and concentrate on clearing your confusions. I am just trying to help. I repeat: most normal charger's IC have configurable float voltage, either by resistor feedback (e.g. LT1512), or via I2C interface (e.g. BQ24296 and hundreds of others). If your experience is limited to cheap knock-offs from Aliexpress, it is your problem. \$\endgroup\$ – Ale..chenski Mar 5 '18 at 6:58
  • \$\begingroup\$ It was a basic question, a simple "Disconnect at 3.92v" or "Have the cell settle at 3.92v" would have sufficed. You mention my 'confusions' but you keep rambling on while not attempting to answer the question. Also, why did you delete your first comment?.. You're attitude is an example of why these forums can be so inhospitable to newcomers. \$\endgroup\$ – I Used To Be Funny Mar 5 '18 at 12:20
  • \$\begingroup\$ @IUsedToBeFunny, it is not funny anymore. You still didn't get it, both of your preconceived variants are wrong. But you can do whatever you wish, disconnect, settle, whatever. \$\endgroup\$ – Ale..chenski Mar 5 '18 at 15:00

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.