I'm building another (!) rather large WS2812B-based LED installation project and I'm now at a point where I'm suffering from what looks to be a power problem. The standard advice is to put a large capacitor (1000 µF, 6.3V or higher) across the power rails, as close as you can to the supply.

I bought a bunch of said capacitors (EEU-FC0J102) but omitted installing them when putting the whole thing together the first time because I didn't quite understand the datasheet of the Murata OKR T/10-W12 DC-DC converters I've installed to feed the strips. It has what I took to be dire warnings that "Excessive capacitance can make step load recovery sluggish or possibly introduce instability" and "Do not exceed the maximum rated output capacitance listed in the specifications" (which I take to mean those under "Maximum Capacitive Loading" earlier in the datasheet.)

SO... given all that I was going to try installing the capacitors I've got. Two questions:

  • Are those capacitors an appropriate voltage spec? I've had a warning that 6.3V caps for a 5V system is not enough of a derating and I might pop them. Is this a serious risk?
  • Are those capacitors an appropriate capacitance and "ESR" rating according to the Murata's datasheet—and if they're not, does it matter much? (This is a one-off project, I'm not going for UL listing or anything.)

With a bonus round: because of the physical constraints of my project (everything's stuffed into a narrow tube), I may need to mount these caps on short (~5cm) wire "tails". I get the impression this is a bad idea, but I may not have any choice. What are the ramifications of this?

  • 1
    \$\begingroup\$ I routinely fit 63V capacitors in an audio power amplifier with 52V rails, which is the same situation you're asking about. There's a 20% margin in both cases. The manufacturer did the same, I didn't invent this. \$\endgroup\$ – user207421 Apr 6 '15 at 22:13
  • \$\begingroup\$ The closer you are to the voltage rating, the more leakage there will be relative to a higher-rated cap. That's not necessarily bad, just design for it. \$\endgroup\$ – Nick T Aug 17 '15 at 0:38

6.3V rating is okay for 5V, provided you meet all the other datasheet limitations.

The ESR appears to conform to the Murata datasheet (> 15m\$\Omega\$), so you should be able to put one of those 1000uF capacitors on each converter (unless you start worrying about the tolerances, but it's probably okay).

The ripple current rating of 755mA (if I read the data correctly) could be an issue causing the capacitor to overheat and fail, but it's difficult to predict what that RMS current will actually be. You might want to either measure the current under worst-case conditions or check for any measurable heating of the cap under worst-case conditions. Worst case is probably maximum current load.


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.