3kJ coin shrinking

[davekni]

My 2002 coin-shrinker is built into a corner of my garage semi-permanently.  Wired into my house through 1meg array of power resistors, avoiding exposure to lethal current when charging and triggering.  Cap is an ancient oil/paper pulse unit, 14uF at 20kV, which I run at ~20.5kV for ~3kJ energy.  Over 1500 coins shrank in its 19 year life, with a few repairs along the way.

Optimization is different depending on capacitance.  At 14uF, two layer coils are better than the more common 1-layer coils.  Magnet wire is terrible, immediately arcing between layers.  Even with wraps of tape between layers isn't enough.  Enamel cracks as wire stretches, and arc paths develop.  Thin stretchy insulation over stranded wire works best for me (radiation-crosslinked PVC).  Broken strands overlap, so conduct better as wire stretches.  Coil shrinks axially and expands radially as it explodes.  My optimum ended up as 2x9 (2 layers of 9 turns each) of 18AWG.  Started with 4x6. but 2x9 passes through optimum shape (wire close to quarter rim) as it explodes.

A few pictures for comparison.  I haven't made a video.

Original wood containment box, lasted about 15 shots before 2x6 boards split.



After 3 rebuilds, changed to 1/4" thick aluminum cylinder.









Coil voltage at 5kV/div.  Pulse lasts about 17us.  This short pulse time maximizes shrinking, but does leave the outer rim thicker than the center.  Skin effect prevents coin current from penetrating to the coin center.  See above image.







Edit:  Thought I should add one more image showing the aluminum cylinder (160mm OD, 6mm wall) 20 years and ~1500 shots later:



Not certain of the alloy, but this is hard aluminum, used in a solid-ink printer that required minimum deflection during ~6000N load from an adjacent rubber-coated transfix roller.  Forming is all from impact of broken strand bits from 18AWG wire.  Shows clearly that coil explodes radially.  (And contracts axially as mentioned above.)

[klugesmith]

Applause!

I agree that coil needs to be matched to capacitor to control the frequency (pulse time).   When there is time to shrink coins with my 52 uF capacitor, I think I will need fewer turns than Dave and slightly more turns than Steve. 

Volts per turn could be a parameter where shrinker designs converge, given their naturally similar coil sizes.

[davekni]

Another 40 shrunk quarters and my aluminum containment cylinder failed:



Put some scraps of the 18 AWG stranded wire that caused the damage into a plastic bag, included in the image.

Replaced with other section of same original printer drum.  Likely won't experience another 1500 shots in the rest of my shrinker's life, so hopefully this is the last replacement.

[FPS]

Thankyou for that photo it remarkable the energy the wire scraps are hitting it with! If you have the same gauge wire with a coil in the centre, is the coil going to explode before the straight wire?

Let's just say you have 1m of solid core wire, so ------------////------------ with a few turns in the centre. Can you tell the wire to fail at the point of the turns?

[davekni]

If you have the same gauge wire with a coil in the centre, is the coil going to explode before the straight wire?

Yes, my coils are this 18AWG stranded wire.  Coils explode.  Lead portion typically gets torn off, but does not explode into pieces.  Coil has much more concentrated magnetic field, so much higher force on wire.  Wire strands (tin plated copper) break out of insulation.  Most insulation scraps are many times longer than wire strand scraps.

Let's just say you have 1m of solid core wire, so ------------////------------ with a few turns in the centre. Can you tell the wire to fail at the point of the turns?

Yes, wire fails where it is wrapped into a coil.  Each coil turn experiences magnetic field from all the other turns added together.  Lead wire section experiences field mostly from other adjacent lead.

BTW, solid wire does not work as well as stranded.  Strands break in different places and slide past each other as coil explodes.  That maintains conductive path longer than solid wire.  Solid wire breaks in many places too, with each break requiring an arc across the gap to keep current flowing.

[MRMILSTAR]

The velocity and energy of those flying shards is indeed amazing. I built my pulsed power generator with a closed blast cage made with 3/8 inch steel sheet. The lid was even bolted down during operation. Even so, one of my early coin shrink operations at 7K joules injured me. I use 12 gauge wire for the coils.  Even though there was no visible crack in the seam between the lid and the wall, a copper shard hit the seam with such velocity that it extruded itself through the seam. It then struck me in the forehead from 10 feet away. The shard embedded itself in my forehead causing a lot of bleeding. I didn't even know that it was embedded in my forehead until about a week later because the wound wouldn't heal.

I added additional shielding to the seams to prevent that from happening again. I have made many shots since then with no additional problems. I heard that those shards can travel at up to 5000 ft/sec which is faster than a bullet. I was very careful in designing this device but even so I overlooked a flaw. I have seen other coin shrinkers use containment structures which have visible seems. I think they may be running on luck.

[davekni]

Even so, one of my early coin shrink operations at 7K joules injured me. I use 12 gauge wire for the coils.  Even though there was no visible crack in the seam between the lid and the wall, a copper shard hit the seam with such velocity that it extruded itself through the seam.

Wow, extruded copper is surprising.  Was the coil positioned such that the seam was in the radial plane of the coil center (or close)?  As you can see from my aluminum shaping, the coil explodes radially.  Shrinks axially.  My lid is just a piece of 2x6 lumber set on top and held on by gravity.  Not even tight, as leads pass in above aluminum cylinder and hold back side of 2x6 up 10mm or so.  However, my shrinker is in the corner of my garage, but charged and triggered from inside my house.  No person is ever close to it when charged and fired.

[MRMILSTAR]

The seam is radially located from the work coil so it is in the direct line of fire. Here is a picture of the blast cage before improvements were made. Even though the seam appears large in the picture due to shadows, in reality you can't even slip a thin piece of paper into it. Also included is a picture of the improved blast box lid showing the added shrapnel shielding around the edges. I also included a picture of the improved blast shield showing shrapnel scarring so you can see where the shards are hitting.

[davekni]

The seam is radially located from the work coil so it is in the direct line of fire. Here is a picture of the blast cage before improvements were made. Even though the seam appears large in the picture due to shadows, in reality you can't even slip a thin piece of paper into it.

Makes sense now.  Still amazing that scraps of wire managed to extrude through the thin slot.