[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [TCML] Cap Discharge Questions



----- "Phil Tuck" <phil@xxxxxxxxxxx> wrote:

> Hello Dave,
> Thank you for the reply. Yes, I was referring to the individual caps.
> My old MMC of 216 caps had its bleed resistors twisted with the cap's
> lead wires, so when I rebuilt the MMC I had to cut the connections
> away, leaving too short a lead on the old cap for a new bleed
> resistor. The result is I now have 12 strings of 18 in series with no
> bleed resistors. I was going to use a resistive stick to discharge,
> and then keeping the stick in place, apply a link across the MMC to
> leave on for when not in use.

> So I presume (actually  I am hoping) your reply of  "No DC path from
> the 'inner' caps, round the loop, to equalize." Applies only if
> whatever was shorting the MMC (HV winding or stick etc.) is removed.

   If i'm right (and i think i am....), without individual
   bleeders, charge will stay stuck 'in' the series string...

> Let's say I have just three hypothetical, fully discharged, caps in
> series, like this: 

> connection 'a' ---  [CAP 1] ---  connection 'b'  ---  [CAP 2] --- 
> connection 'c' ---  [CAP 3] ---  connection 'd'    
 
>  I then apply 1000v just across CAP 2 using connection points 'b' &
> 'c'. Once the charge is removed I then connect, and leave connected, a
> resistance across all three caps at points 'a' & 'd' Will not the
> presence of the resistor and the outer caps provide a path for the
> middle cap to discharge?

    I'd expect not.  Simple solution: try it and see.
    10 to 100VDC, and a hi z voltmeter should do
    nicely.

> (I have never fully understood the mechanism that allows a cap to
> regain some of its charge.)

    As i understand it:
    1) thats quasi irrelavant to this discussion, mostly.
    2) Assuming a single cap, for now:  The charge in the
    'plates' is 'imaged' by charges in the dielectric.
    Being a dielectric, the imaged charges get 'stuck'
    (cf: electrophorus).  The surface portion 
    discharges, the remainder, over time, will 'diffuse'
    to surfaces, appearing as remnant charge.  (Some Years
    back I was involved in high precision 'flying cap'
    analog scanners/multiplexors & we needed to 
    account for this.)

    Back to the series string of just capacitors:
    I picture it this way:
      Think of the individual electrons (lots) on
      opposite plates of the center cap in the 
      string above.  (OK: one plate has 'holes'.)
      connect 'a' to 'd'.  (zero ohms or resistor...)
      How do the excess electrons on one plate of
      [CAP2] find there way thru the dielectric
      of [cap1] and [cap3]?

    For AC(ish) voltages/currents (normal Tesla operation):
     not an issue: AC couples thru the caps.
    for any residual DC (the concern here?) SOME of the
    charge will go in the transient when the connection
    is made, however the (reduced) charge will remain
    trapped on [CAP2], and even lesser on [CAP1] and
    [CAP3].  The amount of charge/voltage on each will
    be affected by the tolerance on the cap values.

    repeatedly making the discharge connection will
    remove 'some more, smaller' amount of charge
    each time.

    As I say: if a 100vish DC supply and a hi z meter be
    handy, try and 'c'.  (Presumably discharge thru an
    _inductor_ would go into oscillation: THAT would 
    drain the charge(s) completely.)

    best
     dwp     

_______________________________________________
Tesla mailing list
Tesla@xxxxxxxxxx
http://www.pupman.com/mailman/listinfo/tesla