SS Variac Update / Improved AC Chopper

["Tesla list" <tesla@xxxxxxxxxx>]

Original poster: Sparktron01@xxxxxxxxxxx 

Steve Conner, Terry Fritz, ALL

I did a "crash-and-burn" test on prototype last night, ran for 1.5 hours while
the 5kV, 60mA NST was arcing into a Jacobs Ladder!  The second isolation
transformer is being blatantly overloaded, and you could smell the hot fish
paper.  The controller heatsink was about 140F, with a 4" muffin fan it would
be fine, it is switching 10A at 32Vpk FWRECT dc Vce.  The IGBT has a Vfwd 
of approximately 2.0V average (20W IGBT contribution) and bridge has a Vfwd
of 1.4V (14W contribution) measured- total on controller heatsink of ~ 34W.
This is equal to 1.2kVA  (heat generation wise) with straight 120VAC, so I'm
confident it will work!

I also increased PWM carrier frequency to ~12kHz, now LP choke is cooler
and low duty cycle output has much less loaded waveform distortion.  Next
modification on plate will be to add a current monitoring feedback loop and
comparator (LF412) to disable of 555 using reset line (async, pin4).
This technique allows very simple pulse-by-pulse real-time current limiting 
at setpoint...   :^)   ).

-------------------------------------------------------------------

Here is a much improved the PWM AC Chopper circuit that appears on surface 
to be the "real deal", particularly at high power.  Please reference
Figure 1 of below URL, save to your PC!  I have already and will refer to this
reference frequently!

http://www.pp.bme.hu/ee/2001_3/pdf/ee2001_3_01.pdf

You would use 600V, IGBT's; SOT227B packages would be ideal, 350V
Bidirectional TVSS, 1500W rated.

Devices S1, S2, S3, S4 would be used, S4 would be tied to middle B phase.
C (bottom) phase devices S5-8 would go away.  Gates and emitters would be
paralleled with switch pairs S1,S2 and S3,S4 respectively.  Low pass filter
(L,C) would be wired downstream (load side) between IGBT's and motor
 (inductive load, transformer,etc.).  This would turn this circuit (a 3 phase
induction motor soft starter) into a single phase, manual (or automatic with
PIC, or PC DAQ card) voltage controller/regulator.

Description of operation is "BRUTALLY" simple...:^)   S1, S2 are commutated
as a normal PWM control (low power = 10% DC, high power = 90%).  S3, S4
are commutated by INVERTING signal to S1, S2 (when S1/2 is on, S3/4 are
off and vice versa).  This allows 4 quadrant freewheeling recovery currents to
flow to load, irregardless of position on AC sine wave.

Control is equally simple, use (2) isolated DC power supplies to run TLP250
optocouplers to drive IGBT switch pairs, and a TL494 (or 555) could be used as
a single ended PWM controller.  There is a DC controller on web available
from AUS that has V adj, I adj and drive for a FET already built for ~ $25 US;
using a TL494, schematic is posted as well.  Feedback current through a hall
device and rectify for your IDC FB, likewise for voltage FB and system is 
pretty
much in the bag. Would have to add inverting drive to operate OC and S3/4
switches.

The circuit illustrated and described above would truly make a bullet-proof
PWM AC controller, that could be expanded to 100kVA+ with minimal cost or risk.
I am working on a Express PCB schematic that will detail system more fully,
and will post it and pictures of my present development "POUNDER" system
through Terry this week, pictures of my prototype "beater" I'll forward
to Terry tonight.

Best Regards
Dave Sharpe, TCBOR/HEAS
Chesterfield, VA USA