Homebrew PCBs

A few other comments:

First: 200uF @ 2KVusing Watt-secs or Joules = 1/2 (CV^2) gives 200 
Watt-secs.That seems like a lot of heat in a 10 mill dia 62 mill long 
hole.More information on the source of this quote of a "typical" amount 
of capacitance/voltage might be in order.Exploding wires in general is 
an exotic topic and unless the quote was for plating holes it could be 
for another purpose requiring this extraordinary amount of energy.For 
instance exploding wires under water between a backing block and a 
microscope cover glass could make an extremely fast strobe flash (think 
pictures of bullets exploding apples).The technique depended on the 
microscope glass literally pulverizing and scattering the light, 
effectively acting like a shutter.One place I found the old column was:
http://jesseenterprises.net/amsci/1957/11/1957-11-body.html
In this experiment the authors were using about 30uF @ 25KV, or about 
9000 Watt-secs.

Second:The capacitor selection.As already mentioned electrolytics need 
voltage balancing due to unequal leakage.This can be a source of 
considerable power drain while you're trying to charge them up.But in 
addition, electrolytics generally have too much inductance due to the 
construction technique of winding the electrodes like a tube of toilet 
paper.This inductance interferes with the fast discharge rates 
required.Capacitors used for pulse discharge applications are 
constructed differently to minimize the inductive component.Still 
another problem is that there are electromechanical problems due to the 
stress of the high voltage on the electrode structures in the capacitor 
so capacitors such as caps used for power factor correction on the 
utility lines (the caps one might likely see on Ebay or find in a good 
junk yard, won't have great lifetimes in an application like this and 
may fail prematurely.So have a spare or two on hand.)

Third:There is an alternate technique one could contemplate.Years ago 
AT&T (?) experimented with repairing PCB edge connectors by placing gold 
foil on the finger to be repaired and setting off a small amount of 
explosive on the surface, driving the foil into mechanical bond with 
finger copper/laminate.Is gun bullet reloading mercury fulminate or some 
similar available?You would need a fast detonation rate material such as 
the primer.Load a pre-cut length of fine copper, brass or bronze tubing 
with the primer (one could easily buy rivets of the proper size.)Slip 
the rivet into the PCB hole, place board w/ rivet on a heavy steel plate 
and use heavy hammer like a small sledge to tap the exposed top of the 
rivet.The resulting explosion might expand the rivet in the hole bonding 
it to the layers.The key to explosive forming like this is that the 
metal acts much as if it has liquified so it fills the surface.This 
might be easier than trying to thread small wires and explode them in 
the PCB.No experience here, but it makes an interesting thought 
experiment.If you do try this -- be really careful and use really, 
really small amounts of primer and don't accumulate it in large amounts 
in any one place.From painful experience as a child, a small amount of 
gunpowder in a small firecracker put a large gash in my hand.As the 
watch captain said on NYPD Blue, "Be careful out there."

The explosive technique was the original method (don't know if it still 
done this way) that Texas Instruments (?) used to form the multi-layer 
sheets that US dimes used to change from pure silver.The metal sheets 
were joined with explosives on the surface, then rolled to final 
thickness prior to punching.Sectioning showed a interface that looked 
like liquids intermingling.

Enough rambling.Enjoy.
Charles R. Patton


On 1/2/2013 8:37 AM, Norm wrote:
>
> On 1/2/2013 8:03 AM, Jeff Heiss wrote:
> >
> > Can you suggest a place where 800uF, 500v caps can be found?
> >
> At the risk of offending some on this list, be aware:
>
> 800 uf at 500 v can be a lethal combination! 200 uf at 2000 v is
> absolutely LETHAL - and you only get one mistake!
>
> How do you intend switching this power? How will you generate the
> charging voltage (the charging voltage is also well into the lethal
> range!)? Safe enclosures, relay isolation, interlocks and grounding
> rods are required. It's not like transistor supply voltages. If you
> are not accustomed to this kind of voltage, please don't attempt it.
>
> Sorry to rant, but I speak from 65 years of industrial experience, much
> of that with 1Kv+ operations .
>
> Norm
> W6NIM
>



[Non-text portions of this message have been removed]