Terry's DRSSTC - Phase Detection/Controller

["Tesla list" <tesla@xxxxxxxxxx>]

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