"Bob Engelhardt" wrote in message
...
On 3/23/2020 12:47 AM, Cydrome Leader wrote:
... I have a supply of large electrolytic caps and some mega
SCRs. Don't mind trying something out if the magnetizing currents
and
times aren't something too insane.


I once tried to re-magnetize an indicator base & don't remember any
details, just that it involved really large current and didn't work
very well.

I do remember what I thought was a neat trick for switching. The
switch was 2 copper leafs, separated by an air gap. To turn it on
you hit it with a hammer! Very fast and very low resistance as the
oxide is displaced and you have a large area bare copper-copper
contact.

An SCR switch blocks the LC oscillation that would immediately
demagnetize the magnet. You still need a flyback diode to suppress the
reverse voltage spike.

https://forum.allaboutcircuits.com/t...athode.111625/
"As a side note, please be advised that proper (mechanical) 'clamping'
is essential to device operation and integrity and not merely for
heat-sink efficiency! It seems the construction 'expects' a precise
compressive force (should be found in the Specs) --- Indeed many such
units will test (electrically) open while 'uncompressed'..."

The actual SCR is a thin silicon wafer slice between the anode and
cathode contact disks. The specified clamping pressure ensures that
enough of the area of the wafer disk will be in electrical contact to
distribute the current and the heat it produces across the wafer
rather than allowing it to concentrate in a few spots and burn out the
SCR. Driving the gate harder than the minimum helps rapidly spread the
initial area(s) of conduction and even out the instantaneous heating.

The clamps GE sent us for our test stations consisted of two U
channels connected by bolts. The dimples accept alignment pins that
center the anode and cathode contact disks while you are assembling
the awkward sandwich of SCR, contacts, insulation and clamps. We used
a simple sheetmetal gauge that aligned two points when the channels
had bent far enough to produce the specified clamping pressure.

http://appliedps.com/clamp-heatsink-assemblies-kits

We didn't need heatsinks to test the SCRs with one or a few
full-current pulses. The heat generation can be estimated as 1 Watt
per Amp continuously, or 1 Joule per Coulomb for a capacitive
discharge pulse.