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

365kJ Capacitor Bank

  1. Apr 29, 2005 #1
    I am pursuing a degree in Electrical Engineering with a minor in Physics, and I am almost done with my 4th out of 5 years ar Northeastern University (go SOX!) in Boston. I have a solid grasp of most theories related to the field.

    However, exploiting my knowledge upon a physical device for the first time is different from using a pencil and a piece of paper. I am sure anyone will agree that there is no shame in double-checking yourself and questioning others with more experience.

    I plan to begin experimenting with very high currents, and to do this I would like to build a capacitor bank. By luck I already have two Mallory 1800MF 450V capable electrolytic capacitors. Obviously, the energy I could store in these guys seems to be: (1/2)*(1.8)*(2)*(450)^2 which leads to 364.5kJ!

    This seems like an awful lot of power, and also seems fansastically dangerous!

    I plan on charging the bank with a rectified transformer (probably a microwave transformer since they are free from the junk yard) under a voltage divider for 450V. I will add safety devices like a complete 1/4inch plastic housing, bleeder resistors, and a voltage monitor, plus a huge red light that warns me not to go near the thing.

    A few questions (since I have never made something like this before)

    1. Is 364.5kJ correct?
    2. What are the not-so-obvious dangers?
    3. What can I use to discharge it safely?
    4. Anything else that I should know/research before attempting this?
    5. Once it's working, what devices can I use it for?
    6. How would a Cockroft Walton voltage mult. config. affect energy and such?

    If you would like to check out my website, it contains a lot of projects I have completed in the past:


    Thank you for your time!
  2. jcsd
  3. Apr 29, 2005 #2


    User Avatar
    Science Advisor

    Are you certain that its not microfarads instead of milifarads?
  4. Apr 29, 2005 #3
    Yeah, they definately say 1800MF (and not 1800UF)
  5. Apr 29, 2005 #4


    User Avatar
    Science Advisor
    Gold Member

    That doesn't mean that it isn't microfarads. I have seen caps printed that way. Physically they would need to be HUGE in order to have a capacity that large at a high voltage.
  6. Apr 29, 2005 #5
    Nate i have to ask , how large are they physically??
    dimentions in inches?
  7. Apr 29, 2005 #6
    They are probably about 8 inches tall with a diameter of about 4 inches (which would make a lot more sense for a smaller capacitance)

    So industry standards aren't as concrete as I thought?... some companies use an "M" and some use a "U" for microfarads?
  8. Apr 29, 2005 #7


    User Avatar
    Science Advisor
    Gold Member

    If you still aren't sure why not just charge it up and measure the charge/discharge time? You don't have to charge it up to full voltage.
  9. Apr 29, 2005 #8
    Because i'm on vacation and I don't have them with me haha.

    But you answered my question about the markings. Now im almost 100% sure theyre microfarads. Thanks

    I don't want to spend time doing the analysis...

    What are the differences in charging time, energy, etc of Cockroft-Walton voltage doubling configurations compared to other methods?
  10. May 1, 2005 #9
    Yeah, the largest capacitor I've ever seen sold is 5000 Farads, and it was quite expensive (since it was meant for industrial use).
  11. May 1, 2005 #10
    You also made a mistake with the formula 1/2CV^2..
    unless i am mistaken your new Joule calculation is .5(0.0018)(450)^2 = 182.25 J
  12. May 1, 2005 #11
    you have some nice projects on your website sepulker; I like the night vision project in particular.
  13. May 1, 2005 #12
    He has two caps (In parallel I presume) not one.
  14. May 4, 2005 #13
    Right, I was assuming two.

    I have another question:
    I want to increase the maximum voltage I can put into the bank. Assuming all capacitors are the same model... is it healthier to:

    1. Stack capacitors in series
    2. Use a voltage doubler like the Cockroft-Walton

    It would be interesting to see the pros and cons
  15. May 4, 2005 #14
    in my opinion , for raw power , stack them , besides the current output of the Cockroft-Walton and its variations is low, on the order of 100mA , although they do talk about one that looks like it could produce more current , it has sort of a complex triggering scheme because it doesnt use P channel FETS because at the time it was written they were more expensive than they are now.. http://murray.newcastle.edu.au/users/students/1999/c9221792/doubler.htm [Broken] your doubler circuit
    how much voltage are you looking to make?
    Power fets are only rated for around one hundred volts or so , so stacking them would be your only choice..
    Last edited by a moderator: May 2, 2017
  16. May 4, 2005 #15
    did you know that there are high voltage capacitors in microwave ovens , on the order of 2100V ..
    @ 1.1uF that comes to 2.4255 Joules..
  17. May 5, 2005 #16
    On Ebay, I found Nippon Chemi-Con capacitors rated at 400V 3600UF. The guy has 60 available. So let's see... each one can hold 288J each. If I buy 10 and split the stack, thats 9000UF at 800V (i think thats right?) at around 2.9kJ. I can stack it more for a better approach to a spark gap... actually...

    What would be better for switching the current? A spark gap? Thyristor SCR, Thyratron? (Maybe I don't have to stack them)

    If I assume my load is a coil and it has a impedance low enough to allow a fast pulse, which should be enough energy to crush a soda can, right?

    This would be perfect for my demonstration.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook