RE: Terry's DRSSTC - Phase Detection/Controller

["Tesla list" <tesla@xxxxxxxxxx>]

Original poster: "Adam Horden" <adamhorden@xxxxxxxxxxxx>

Hi Terry,

Looking very nice and professional

I recently obtained a half length 2u rack enclosure for my OLTC
controller.

http://www.acidtech.net/ah-oltc-rebuild.html Some photos are here.

It shows the photos before I rebuilt the coil and the current rebuild
going on over Christmas. I am trying to get every last inch of arc from
this coil. At the moment it did 12.5 inches but I am hoping for more.

Terry your CT looks nice looked at the data. Shame its no good for my
intended use I need around 800A pulse for my small OLTC. I will wind a
rogowski coil.

Adam

-----Original Message-----
From: Tesla list [mailto:tesla@xxxxxxxxxx]
Sent: 24 December 2004 05:10
To: tesla@xxxxxxxxxx
Subject: Terry's DRSSTC - Phase Detection/Controller

Original poster: Terry Fritz <teslalist@xxxxxxxxxxxxxxxxxxxxxxx>

Hi,

Here is the latest idea I am working on to detect the phase of the
primary
loop current to drive the IGBTS:

http://drsstc.com/Design/CurrentPhaseDetector.gif

It is pretty similar to the one in Dan's book but this one does hundreds
of
amps for the primary side current control.

CT1 is a center taped 100:1 Current Transformer.  The center leg is
taken
as ground.  D1 and D2 clamp the negative going signal and D3 and D4
clamp
the positive against a 4.7 volt Zener power rail.  The CT+ and CT-
outputs
get slammed against these rails pretty hard so the 4.7 volt Zener system
is
isolated from the +5V power supply.  The +5 volts just charges the rail
during power off.  The output swings between -1 and +6 volts but that is
ok
in this case.  I like the center tapped CT since the forward logic only
has
to play with positive signals now which saves a lot of trouble.

The main controller is still being worked on but it looks a lot like
this

http://drsstc.com/Design/Controller.gif

The signals from the CTs got to two OR gates.  One is null and the other

(U1b) is used to inject the initial firing pulse.  The rising edge of
the
PULSE signal fires a ~3.3uS pulse from U5a which turns the H-bridge on
and
gets things going.  U2c "can" shut the thing down if some odd current is

detected during the initial pulse but it is really just there because it

looks good ;-))  U2a and U2b prevent both sides of the H-bridge from
being
on at the same time no mater what.  U3a and U3b are the main AND gates
that
take all the signal and produce the logic drive signals.  They also add
the
FAULT shutdown at this point.  U4 is a QUAD balanced line driver chip
that
drives the LEDs on the TLP250s  It puts out about 0.5 to 3.5 volts under

load thus the seemingly small resistors.  That helps since the smaller
resistors can drive the LED junction capacitance better.  The capacitors

(C2-C5) give the LEDs a nice hard kick to help them turn on and off
really
fast.  Since all the ICs are really fast, the propagation time through
the
whole mess is a little less than 40nS!  The circuit does have plenty of
hooks to play with timing and dead time if needed.  It will be on a
bread
board at first so I can fiddle as needed with it once it is actually in
circuit.

I have started working and the cabinet:

http://hot-streamer.com/temp/PC230007.JPG

Just mounted the power supply, capacitor drain/charge resistor, bridge
rectifier, little power resistor, feet, and the rack handles.  I'll add
the
line filter and connector blocks to the back next and then do the front
panel last after I have had practice ;-)  Most of the ordinary stuff
will
mount on the sides out of the way, leaving plenty of room for the cool
stuff.

So that's where I am at.  Most of the stuff at www.drsstc.com should be
up
to date now too.

Cheers,

	Terry