I've always gotten a kick out of doing things with "found" materials,
and adapting cheap stuff and simple (i.e., lazy) approaches to do the
job of fancier equipment (oscilloscopes aside, obviously -- I never
found a plausible substitute for one of those). Unfortunately, a
consequence is that I tend to cut corners where it would have been
better not to do so. I thought folks on the list might be amused by a
few tidbits illustrating what happens when a math guy fiddles with
electronics.
So, the main point here is that I finally put together an electrophorus
(it really _is_ insulated, but the insulator is clear packing tape,
which is kind of invisible in the photograph):
http://www.physicsinsights.org/images/electrophorus-1.jpg
Unfortunately I cut too many corners and it doesn't work as anticipated
-- the clear packing tape I substituted for a sheet of polyethylene in
the plans refuses to take a charge when rubbed with anything I own, so
if I just touch the pan and then lift the top part, nothing happens.
However, it still makes a dandy variable capacitor; it's around 300 pF
when together (determined from discharge curve through a 100K resistor),
presumably falling to zip when the pan is lifted (haven't actually
measured it in that configuration). So........
I tried charging it up to about 1500 volts, disconnecting the "power
supply" (quotes explained below), and /then/ lifting the pan by the
handle. Bingo! Got a spark roughly a quarter inch long. Not
spectacular, but at least it demonstrates the "varying potential"
effect: separate the plates while keeping the charge constant, and the
voltage goes up, just as it's supposed to do. (And then I got
distracted with tinsel experiments and didn't pursue the electrophorus
-- such is the lot of a dilettant ...)
The 1500 volt "power supply" is worth a word or two, also; it's home
brew, stepped up from 18 volts (originally conceived to run on
transistor radio batteries, but in practice driven from another DC
supply). I originally planned to cobble up something along the lines of
a single-coil flyback rig, until I realized the gutsiest transistor I
own can only withstand 140 volts; if anything significant flew back from
the coil the transistor would fly away. So, I fell back on Plan 2, and
tried using a custom-built monofilar-wound transformer with laminated
carbon steel core:
http://www.physicsinsights.org/images/wire_wrap_coil_1.jpg
The green stuff is about 2 dozen turns of hookup wire. The yellow and
black spools are exactly what they say on the label, with some leads
soldered to the "inside" ends which stuck out just far enough from a
small hole in the middle of each spool to make that possible. The two
(partly used) spools alone, wired back to back and placed one on top of
the other, without the penknife, are about 0.08 henries (measured via
discharge curve across a resistor), and almost exactly 100 ohms
resistance. Didn't measure the inductance with the penknife in place,
and in fact it only helps the transformer function when partially
inserted. Since I was just holding it in that position by hand, that got
old fast, so I dispensed with the penknife in the "final design", and
went with straight air core.
The actual HV "power supply" consisted of a square wave generator
running at about 50 kHz feeding the primary, and one rung of a
"ladder"-type voltage multiplier on the secondary using 1kV diodes
(1N4007's) and some 100 pF resistors:
http://www.physicsinsights.org/images/kv_supply_1.jpg
The basic 18v DC supply is also partly visible in the background of that
shot, hiding behind the computer monitor; it's also home brew but a far
more respectable design than the HV jobbie... The red and black twisted
pair visible in the middle ground is my home brew "100x scope probe",
and it doesn't work very well but at least it keeps me from accidentally
blowing the front end off the scope when looking at 1500 volts.
I'd have gone for more multiplier stages (and hence more volts) but I
couldn't find any more 1kV capacitors in the basement (and I need 2kV
caps, anyway -- each ladder stage can apparently hit about 1.5 kV). I
have the makings for a couple Leyden jars here (i.e., empty butter
containers and some aluminum foil) but the easy way to make the "inside"
plate uses a liquid electrolyte and I'm hesitant about containers filled
with conductive liquids in this very cramped home office.
(I'll have more to say about HV supplies in a later post -- I seem to
have some "over unity" diodes here and I've been going nuts trying to
figure out what I'm doing wrong...)
*******************************
Once I had the HV supply in hand, tinsel experiments seemed called for.
A kilovolt is apparently enough to get a piece of tinsel very
interested in another electrode. That's fun, of course, but it also
brought to mind the question as to whether a dielectric really does
block an E field. Could I try that experiment? Answer: No, not
quite.... the dielectrics I could bring to bear (primarily empty butter
dishes) turned out to have their own, permanent charges along for the
ride, and their permanent fields totally swamped my little kilovolt
field. The tinsel just went nuts over the butter dish... sigh.
*******************************
OK enough of that. At some point I decided I wanted some real sparks.
One of my plans is to home-brew a Tesla coil, but not today (no magnet
wire in the house, for one thing). However, in the course of packing
before we moved I did run across an old Allison high performance coil I
got for a long-gone car, long long ago, and never got around to
installing. I also happened to have an old spark generator I had built,
which was intended to drive an ignition coil, but it never worked very
well, possibly because the coil I originally used it with was NG to
start with (it was a discard from the tuneup of some car or other, years
gone by). The "sparker" used a square wave oscillator and voltage
multiplier (diode/capacitor ladder) to step 18 volts up to between 65
and 70 and then discharge it through the primary on the coil. Some
debugging with the new scope (storage scopes are magic, really!)
revealed a number of problems, starting with the fact that I had much
too much capacitance in the final stage and all it was doing was making
the power supply hot (or killing the batteries -- this actually started
life as a battery based gadget). A snapshot of the nearly final
version, with outboard power supply used in place of batteries (which
run down kind of fast):
http://www.physicsinsights.org/images/spark_generator_1.jpg
The blue tweak pot adjusts the spark frequency.
Anyhow, after a bit of fixing, it produced some reasonably satisfactory
results. Still no tesla coil, not by a lot; probably not even the
60,000 volts for 3 mS which are claimed in large letters on the coil's
side. But based on spark length it appears to be in the neighborhood of
35,000 volts (assuming 30,000 volts per centimeter) and based on a look
at the hash coming back on the primary side, it appears I'm getting a
solid millisecond of arc. Here are a couple short clips:
http://www.physicsinsights.org/images/mvi_0780.avi
http://www.physicsinsights.org/images/mvi_0781.avi
Whatever, it's all fun, even if I never really learn anything new from
any of it...
