I want to use PWM to adjust the input of a 60 Hz high current output transformer (instead of using a powerstat), and I would like to run a simulation on it to see the waveform of the current applied to the load under various conditions. The transformer I want to model consists of a 240 VAC primary toroid, with a single turn of bus bar through the hole. This produces about 1 volt per turn, at about 2 kVA, or 2000 amperes continuous. I will need to produce up to at least 20,000 amperes for short durations (0.1 seconds).
I made a circuit in LTSpice that seems to be fairly close to the experimentally observed characteristics of the actual transformer (or similar units). I used a primary inductor of 1600 mH, and a secondary of 25 uH, for a ratio of 252. Under open circuit secondary conditions, it draws
707 mA, which seems about right. I added 10 mOhms primary resistance and 2 uOhms secondary resistance, and I also added a non-coupled inductance of 200 nH to the output. I get a current of 13.6 kA into a 10 uOhm resistive load at 64 degrees phase shift, which is about what is usually expected under these conditions.However, I am really just guessing the inductance and resistance of this model. I would like to be able to model this more accurately, based on actual measurements I could make on the prototype. I can measure the primary resistance by using a DC current source, and I can measure the inductance by using AC voltage below saturation. The secondary inductance is determined by the turns ratio, but there will also be a component of external inductance that will depend on the geometry of the bus connections. I might be able to determine that by experiment. I don't need the model to provide accurate results for saturation or hysteresis, but it would be a plus if it could.
Following is my model for this transformer. If you have any ideas on just what measurements I should make, and how to add the characteristics to this subcircuit, please let me know. I have designed transformers like this for several circuit breaker test sets, and they work well, but my PWM design will need a fairly realistic model to observe and control switching transients. I already blew out a large IBGT in a small prototype, and my recent LTSpice model showed transients of over 1 kV before I added a snubber.
Thanks,
Paul
===================================================================
Version 4 SHEET 1 880 680 WIRE -96 144 -256 144 WIRE 64 144 -32 144 WIRE 240 144 144 144 WIRE -192 224 -256 224 WIRE -96 224 -192 224 WIRE 96 224 -32 224 WIRE 240 224 96 224 WIRE -192 256 -192 224 WIRE 96 256 96 224 FLAG -192 256 0 FLAG 96 256 0 SYMBOL ind2 -112 240 M180 WINDOW 0 8 120 Left 0 WINDOW 3 -31 -6 Left 0 SYMATTR InstName L1 SYMATTR Value 1600m SYMATTR Type ind SYMATTR SpiceLine Ipk=100 Rser=10m SYMBOL ind2 -16 240 R180 WINDOW 0 6 121 Left 0 WINDOW 3 -30 -4 Left 0 SYMATTR InstName L2 SYMATTR Value 25µ SYMATTR Type ind SYMATTR SpiceLine Ipk=10000 Rser=2u SYMBOL voltage -256 128 R0 WINDOW 3 -55 -44 Left 0 WINDOW 123 0 0 Left 0 WINDOW 39 -55 -16 Left 0 SYMATTR InstName V1 SYMATTR Value SINE(0 350 60 0 0 0 10) SYMATTR SpiceLine Rser=1m SYMBOL res 224 128 R0 SYMATTR InstName R1 SYMATTR Value 10µ SYMBOL ind 48 160 R270 WINDOW 0 32 56 VTop 0 WINDOW 3 5 56 VBottom 0 SYMATTR InstName L3 SYMATTR Value 200n SYMATTR SpiceLine Ipk=10000 Rser=1u TEXT -128 280 Left 0 !K1 L1 L2 1 TEXT -312 506 Left 0 !.tran 100m TEXT -208 320 Left 0 ;240 V to 1 V at 2000 Amps