Understanding constant current/constant voltage charging of LiIon battery

I'm trying to understand exactly how CC/CV charging of a LiIon battery works and I tried to replicate it with a cheap bench power supply (RD Tech DPS3003) and a Samsung INR 18650-25R. The bench power supply is capable of both CC and CV. Here is what I tried:

• Discharge the cell to 3.1V
• Set the bench power supply CV to 4.2V according to the battery data sheet
• Also set the PSU CC to 1.250A or 0.5C, according to the data sheet
• Connect the PSU's positive lead to the battery's positive terminal; same for negative

What I thought would happen:

• The PSU will supply as much voltage as necessary to make the set CC of 1.250A flow into the cell
• As soon as the cell voltage reaches the set CV of 4.2V, the PSU switches to CV mode and the current will decrease steadily

This is what I saw instead:

• When I measured the voltage at the battery terminals while charging, the voltage went up only a little, to around 3.3xV and slowly rising as the battery charges
• The PSU display shows around the same voltage and only a current of around 530mA
• The PSU displays that it's in CV mode, rather than CC mode.

Clearly my understanding was wrong and I'd love to understand better. Why did I not see the set CC of 1.250A? Were my settings wrong, e.g. do I have to allow for a higher voltage? But then how will the PSU limit the cell voltage to 4.2V and switch to CV mode later? Or could it be that my cheap bench power supply does not behave as it should?

Additional information: I also did a quick check with my Liitokala Li500 battery charger. I set it to charge the same battery (by then at around 3.3V cell voltage) with 1A and measured the voltage at the contacts. It was around 3.5V. Unfortunately I could not easily measure the current. I'd be very glad for any insights on how my understanding was wrong.

2 Answers

I might say compute delta V /ESR of load to choose 1A assuming 50 mohm ESR. Then plot Vbat vs I and compute A-sec and ESR vs t in a spreadsheet or compute Coulombs or VI*t or Ec stored energy and then efficiency. The Li Ion cell is about 10 kfarads.

So 1A*50mohm = 50mV above Vbat and keep adjusting or add series R to improve PS stability and manually adjust Vout CV mode to achieve this, or use a 1kW tungsten NTC bulb in series and adjustments will be easier for tiny SMPS that is sensitive to load ESR. This SMPS is clearly not designed for battery charging unless tweaked for stability.

Also compute ESL or inductance of your cable loop and keep short so that resonant freq does not interfere with SMPS.

SRF=1/{2pi√(LC)}. Also beware cells have memory and thus two RC equivalent series networks in parallel or more. (See my answers related this site ... search )

• Wow, it will take me some time to fully understand all of it fully, but I understand now that there is a lot more to the LiIon chargind process than my initial understanding. Thank you very much for the valuable input, I will use it as a guide to research and learn to understand the process better. – derfraenk Oct 3 '18 at 18:42
• Battery university is your best bet... safety and VI profiles optimum lifespan vs mAh reduction – Sunnyskyguy EE75 Oct 3 '18 at 18:43
• Also search me here – Sunnyskyguy EE75 Oct 3 '18 at 18:43

Your understanding is correct. You want to limit the current until the voltage reaches its final value, then limit the voltage. I did exactly this with an HP bench supply (obviously many years ago) so I suspect a problem with your power supply.

• Thank you very much for your quick reply. I'm glad my understanding is not wrong, it has been bugging me. Then I will contact the manufacturer of the power supply. – derfraenk Oct 3 '18 at 14:08
• WIth a LiIon cell with an ESR perhaps lower than your PS output cap, it becomes unstable and cannot achieve full current . 0.2V rise at 0.53A=0.4Ohm while your cell may be 10% of that, so increase voltage slowly to achieve 1A or add series R+L=0.1Ohm at smps f. It is not an HP or Lambda or Murata PS!!! Nor an ideal lab supply. I might say compute delta V /ESR of load to choose 1A assuming 50 mohm ESR. Then plot Vbat vs I and compute A-sec and ESR vs t – Sunnyskyguy EE75 Oct 3 '18 at 15:35