Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

My 27kJ capacitor bank seems to be only 27J. Need your help.

  1. Jun 27, 2015 #1
    So, I've been doing a lot of spot welding and research/development of this manufacturing process I came up with a while ago, both of which involve high-energy capacitor discharges, and my previous 2.3kJ capacitor bank just wasn't cutting it anymore. So, I invested in 100 "Green-Cap" 2.7v 100F supercapacitors made by a company called Samwha. I spent the last two days soldering 74 of them in series to create the equivalent of a 200v 1.351 F capacitor, which according to my math, should be able to store about 27kJ of energy.

    Which is why I was shocked when I tested it for the first time only minutes ago. Oddly enough, using only my 200W power supply, it charged to 200 v one or two seconds when it theoretically should have taken 135 seconds. Then, when I discharged it, the spark was so tiny, it didn't even make a noise. I've played with 50J camera capacitors that had more of a bang than this thing.

    I must have checked 8 or 9 times, but, no, the polarity is correct, and the circuit is not shorted anywhere.

    What could cause this? What massive, fundamental error have I must have committed for my math to be 3-4 magnitudes off? Or did the semi-anonymous chinese dude who sold me these give me a box of lemons? Or could Samwha be guilty of false advertising? What is it? What could it be? I'm completely stumped here.
  2. jcsd
  3. Jun 27, 2015 #2
    Supercapacitors are slow.
  4. Jun 27, 2015 #3
    As PietKuip hinted, this is the wrong part for the job.

    Supercapacitors have a number of http://eprints.soton.ac.uk/272829/1/Supercap%20Journal.pdf [Broken]. Not only do they have an ESR, but they have a series inductance and parallel resistance. I suspect your problem is with the inductance.

    Supercapacitors store their energy in an electrolyte. It can take hours to charge or discharge fully.

    The name "supercapacitor" isn't that great. They are more like lame batteries than capacitors. They have niche uses where they are clearly the best part for the job, but outside those areas, they tend to be problematic.

    I hope your vendor is willing to take them back, perhaps trading them for the real thing.

    It always pays to read the datasheet before buying the part. I can't count the number of times I've had to change parts from what I thought I needed.
    Last edited by a moderator: May 7, 2017
  5. Jun 27, 2015 #4

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

  6. Jun 27, 2015 #5


    User Avatar
    Gold Member

    If you short a supercapacitor, then open it, you will see the voltage climb back up. The energy is there, but it acts like lots of small capacitors in parallel with resistors in in between. SO the last capacitors have extremely long time constants (high esr).

  7. Jun 28, 2015 #6
    I read that spot welding is done with low voltages, not 200V. Also 1 to 2 seconds of charging time for your caps just can't be. With a 200V, 1A power supply it should take 270s to fully charge them. How did you measure the voltage on those caps? And did you remember to add balancing resistors to avoid damaging them?
    Anyway, super capacitors can certainly be used for welding metal. I found this article
    And also there are quite a few videos on youtube demonstrating the power of those caps.

  8. Jun 28, 2015 #7


    User Avatar
    Gold Member

    I'm pretty out of date on this one. The series resistance of those youtube ultracaps is much less than I was aware of.

    The spec for the caps the OP is using is 10 mOhm. So 74 in series would only be 0.74 ohms. ( http://www.produktinfo.conrad.com/d...74999/451437-da-01-en-GREEN_CAP_100F_2_7V.pdf )

    Checking the numbers, given 1A, 200V, 1.3F where 1A = 1.35F *200V/T so T= 270 seconds.

    You should experiment with 1 capacitor. Charge one to 2.5V and see how it discharges. Also, you need to read this and deal with balancing the charging for series capacitors.
    http://www.tecategroup.com/ultracapacitors-supercapacitors/ultracapacitor-FAQ.php (see section on load balancing)

    If you don't load balance you will blow the capacitors when charging (they explain why).
  9. Jun 28, 2015 #8
    You're math checks out.

    You may want to be careful that the leakage current through one cap doesn't build up on the next in series by using voltage balancing resistors. [Edit: OK, this is load balancing that meBigGuy already covered.] Also, any capacitor with less capacitance than average will have a proportionally greater voltage across it.

    You may be pushing the envelop operating so close to the rated voltage. Have you checked each cap for short?
  10. Jun 28, 2015 #9


    User Avatar

    Staff: Mentor

    Thread closed temporarily for Moderation...
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook