RE: SSTC battle continues!

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "Steve Conner" <steve.conner-at-optosci-dot-com> 

 >He's talking about battles with the driver

Yes, it can be a battle just getting a SSTC to work and stay working for
more than 10 seconds. All those problems multiply as the coil gets bigger
and more powerful.

But knowledge of practical EE matters like layout, shielding and grounding
can help you a lot. These are far more important than they were in spark-gap
coiling- it takes a lot less interference to destabilise a self-resonant
controller than to burn out an NST :(

In practice, it's probably a good idea to build your SSTC controller, and
plan the wires that connect it to power supplies, H-bridge, ground etc, with
the same attention to shielding and grounding as a computer CPU box or a
high-sensitivity radio receiver. Radio techniques are quite appropriate, as
the idea in a driver is to let the power supply and feedback signal into the
box, and let the gate drive out, while keeping all other conducted/radiated
interference out.

Possibly the worst offender is ground loops that include your SSTC driver
circuit board. These can pick up magnetic field from the primary/secondary
coils as well as the large dI/dt caused by an arc to ground.

Another bad source of interference is the dV/dt caused by an arc to ground-
as the toroid capacitance is suddenly discharged to ground, it pulls the
local RF ground up to a remarkably high voltage. This effect is discussed in
textbooks on commercial lightning protection. It can cause electrostatic
coupling of hash into your control circuits, and even cause the RF "ground"
to arc over to your control circuits which are hooked to a different (AC
line) ground.


Steve C.

best SSTC spark so far: 185cm

Death toll: 9 small IGBTs, 2 fast recovery diodes, 4 MOSFETs, 6 UCC3732x
driver chips, a pile of 555/74HCxx/LM339/2N3904/etc, a heap of fuses, and
one 240V 40W filament bulb burnt out by ground strike current.