# Supercapacitors in series

Can I charge series of 4 Maxwell BCAP3000 supercapacitors with constant current method using an adjustable Power Supply, and exceed the voltage rated for single cell?

What I mean is to charge the series of 4 ultracaps, that should be able to charge up to 10.8V, and have voltage higher than 2.7V on the power supply. I'm sorry if this question is somewhat silly, but this is my first take on supercaps connected in series. I've done single cell tests but never exceeded rated voltage on it. Therefore I have this small concern to solve.

They are rated for 2.7V. I have the manufaturers' balancing transistor boards attached to the cells, so single cell should be protected, but I am unsure if they are working since no LED is shining so far.

EDIT: Answers and comments provided solved my problem so this topic can be closed. Thanks for all of the concern and replies.

• What is the charged voltage? Commented Jul 5, 2018 at 18:13
• I charge them with constant current of 10 Amps, and power supply adjusts voltage by itself. So it goes from lower to higher voltage as capacitors are charged.
– QST
Commented Jul 5, 2018 at 18:16
• You have not done it yet. You are asking whether it is okay to do.. I re-read your question now. I think it is okay. But I don't know about super caps balancing. Let the experts speak Commented Jul 5, 2018 at 18:19
• By the way,why are you charging them with constant current? If you have adjustable power supply? Commented Jul 5, 2018 at 18:23
• I want to make a characteristic of charging/discharging of a supercap, and from it calculate the ESR, capacitance and other parameters. I've read that constant current is apropriate but constant voltage can be used as well.
– QST
Commented Jul 5, 2018 at 18:27

Can I charge series of 4 Maxwell BCAP3000 supercapacitors with constant current method using an adjustable Power Supply, and exceed the voltage rated for single cell?

Any capacitor put in series will increase the voltage rating of the capacitor. Keep in mind that supercapacitors are different from normal capacitors because of their very low ESR (Equivalent Series Resistance). The ESR could create problems if the current is being sourced fast enough to heat the capacitors (if the ESR rating varies by say 50% then the highest ESR capacitor will source most of the heat, in this case you will need to balance the capacitors.

You also need balancing if the capacitor values are different:

Since the individual ultracapacitor cell voltage is relatively limited compared to the majority of application requirements, it is necessary to series connect the ultracapacitors to achieve the voltage required. Because each ultracapacitor will have a slight tolerance in capacitance and resistance it is necessary to balance, or prevent, individual ultracapacitors from exceeding its rated voltage. Consider a string of 3 ultracapacitors with the following performance:
C1 = 100F and 0.011 ohms
C2 = 110 F and 0.012 ohms
C3 = 95 F and 0.010 ohms

If each ultracapacitor is initially at 0 volts and the string of ultracapacitors is charged to 7.5 volts at a constant current then C3 will reach 2.5 volts before C2 or C1. dt = IC•dV Thus if the string is not at 7.5 volts C3 will continue to charge above its rated voltage of 2.5 volts. In order to address this issue, balancing is required to maintain the ultracapacitors within its rated voltage. Balancing can be achieved through two different methods, active balancing or passive balancing.

Source: www.maxwell.com/images/documents/PG_boostcap_product_guide.pdf

• Thank you for the solution, I've already made charging/discherging cycles in my project. Apparently the last capacitor in the string was charging the fastest and balancing boards kicked in earlier than in front ones. I've also used the resistor to initially charge capacitors, so the power source woldn't see a shortage. When capacitors charged to about 4V in series then I could switch off the resistor and they were charging. Apparently their ESR had to raise so the power source let me charge them without a resistor. I'd consider my problem as solved by now. Thank you for all replies.
– QST
Commented Jul 6, 2018 at 20:43
• meta.stackexchange.com/questions/126180/… Commented Jul 6, 2018 at 21:28
• Pick the best answer and vote if you feel that it was a good answer, you don't have to vote for mine, but its a simple way to reward the volunteers on this site Commented Jul 6, 2018 at 21:31
• I've voted for both answears as they face my problem but in parts. Your answer is relevant to the charging process and Jack's answer faces the problem from the heat dissipation approach. Both of them must be faced, however heat dissipation wasnt much of a problem in my case as those capacitors can take a current higher by order of magnitude and be alright. Unfortunately I can't accept 2 answers, yet both are very helpful. And due to my low reputation on the site upvotes aren't publicly shown yet.
– QST
Commented Jul 6, 2018 at 21:38

There is no indication when using the SAB devices (likely the Maxwell balancing board you are using). It's based on balancing the leakage currents of the series connected supercaps and will not protect you from poor balancing at high charge current. Since your cap values can vary by more than 20% (datasheet) you will have unbalanced terminal voltages, so must provide this protection.

Yes you can easily end up stressing a capacitor due to overvoltage.
If you are going to charge the capacitor string at say 10A (your PS in CC mode), then you must allow for the possibility that any one of the caps may enter overvoltage (whatever you select as the terminal voltage ..in this case 2.7V maximum). At that point in time you need to stop charging that capacitor so must shunt the 10A around the cap, which will stop it charging.
At 2.7V and 10A you obviously need to dissipate 27W in the shunt while the other caps continue to charge.

Given you have 4 caps in series you will need a shunt balancer across each cap to protect it.
How you create your protection mechanism depends on what maximum charging current you want.

For example if you set the maximum current at say 1A, then you only have to dissipate 2.7W in each shunt which would be much easier to design. Worst case you will have 3 of the 4 shunts active, so a total of about 8W. At 10A you may need to dissipate 80W worst case.

There is some added complexity due to the ESR, but in the case of the caps you are using this is very low so can be almost ignored (max delta V would be 290mV @ 1A).

• Thank you for the answer, it shines a brigther light on the problem, I think I'll be measuring voltage on every cell, then if one will overvoltage I'll shunt it. Initially I'm charging those caps with a resistor, so the power supply won't see a shortage, which I know it isn't the most effective, but it's most basic way. I'll also think to decrease the initial charging amp to 5A.
– QST
Commented Jul 6, 2018 at 6:05