1
\$\begingroup\$

Our circuit has three supercaps in series across the Battery+ and GND rails. The average current draw is very small -- around 250uA -- with peaks of 20 mA.

Short form

Do we need load balancing across the supercaps?

Details

Our battery powered system needs to last for > 2 years, so minimizing current draw is crucial.

We've included the supercaps to soften the peak load on the batteries when the WiFi radio kicks in (about 40 mA), but most of the time, the system is in deep hibernation and draws about 50 uA.

The three supercaps are 3F, 2.7V devices, in series across a (nominal) 7.2V supply rail. In the units I've measured, the voltage drop across each capacitor is close to 7.2V / 3 = 2.4V, so -- at least for now -- the voltages are balanced.

But I'm concerned about variations due to ageing. Is it important to provide a load balancing mechanism, or will the capacitors continue to self-balance (as it appears they are doing now)?

If load balancing is important, then what technique do you recommend (keeping in mind that minimal current draw is crucial)?

Putting a 2.5V zener with a sharp knee across each supercap would offer protection in case things got really unbalanced. Similarly, a 1M resistor across each supercap would offer some balancing and would only increase the current draw by a couple of microamps, so that's also acceptable.

\$\endgroup\$

1 Answer 1

0
\$\begingroup\$

For discharge you only need to prevent under voltage which means reverse polarity on caps.

For charge obviously overcharge must be prevented.

-The very best tolerances you can buy from D-K is +/-30% while typical is -20% +80% and worst is -0/+100%.

  • The mismatched rise in voltage depends on C
    • dV=dt*Ic/C

It is well known that balancing extends the life of rechargeables because the weakest cell degrades faster. Although the weakest capacitance discharges fastest, that does not necessary mean it ages fastest.

Balancers are essential due to unbalanced capacitance, the same as Lithium batteries.

How you balance depends on power dissipation and speed or time dt.

\$\endgroup\$
2
  • \$\begingroup\$ Hmm. If you have three fully discharged supercaps in series (roughly equal capacitance), and you apply 7.2 v across them, in practice, is there anything that will prevent one capacitor from getting the full 7.2v ("bang!") while the other two get zero volts? \$\endgroup\$ Dec 19, 2019 at 1:33
  • 1
    \$\begingroup\$ final dV always depends on C mismatch which is expected. But where the charge current is near the self-discharge rate the leakage resistance rises with overvoltage , so with very low currents ,it self regulates. I don't know how elastic leakage rates are , due to chemical reactions. I think a string of LED's rated to handle the charge current would be good protection. \$\endgroup\$ Dec 19, 2019 at 1:43

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.