Hello,
I am experimenting with a 300VA 117VAC/234VAC dual toroid power transformer. I have a bench setup with a 0 -130VAC 2000VA variac connected to the input of the transformer for slow ramp up. On the output side of the transformer I have a full bridge (diode) rectifier, 400uF capacitor, and a load resistor.
I am searching for a way to simulate the setup for varying input/output configurations such as the primary coils connected in series or parallel and similarly for the secondary being in series and parallel and combinations thereof for various load resistors and output capacitors. I would eventually like to collect data using a scope (voltage/current probe setup) etc... and compare/verify the collected test data against a model.
The bench setup works as expected for various input output configurations (2Vs or 0.5Vs outputs) and I seem to be getting a handle, at least in hardware, on how things work. However, I constructed a LTSpice model, without prior LTSpice knowledge, and I cant really seem to get the model to behave as expected. I used a fairly inexpensive LCR meter with 100Hz setting to measure coil inductance but with very little intuition what values should be expected albeit the values I am getting from measurement seem unreasonably large. Here is a screenshot of my LTSpice schematic and the inductance values are from measurement. I physically wired the primary coils in series and used the LCR meter across the outer leads and then the secondary in parallel and once again measured across the two leads and then just backed out the values by dividing by two considering L1+L2=Leq and 1/L1+1/L2=1/Leq etc...
I cant figure out where I am going wrong. Am I using the simulation tool incorrectly? Is it the wrong choice of simulation tool for this level of power application? Or have a taken poor test measurements?
Any guidance on this would be appreciated but I suspect at the very least I should be able to get the simulation to output 2Vs or 0.5Vs. I thought about lumping the coils as a single coil on each side but at this point I would like to understand what is wrong with my approach thus far.
Schematic:
Sample Sim Output
Hi zirogravity ,
Hopefully, someone in the community can troubleshoot your circuit.
One thing to try: give the primary side a ground connection. LTspice needs a reference point to do it's calculations; doesn't like floating circuits.
mike
Hi zirogravity ,
If you need to see isolation, connect the node to GND through a high value resistor.
mike
Hi MStokowski
Thanks. I did try placing different GNDs on either side of the transformer. The output changed centers but did not result in "expected behavior".
Regarding the isolation suggestion. This is very helpful suggestion. Would I place the high value resistor on the supply side to ground or elsewhere? My original line of thinking was just the coil resistance would suffice but after your suggestion I realized the coil resistance would be too low for isolation so thank you.
Wess
Hi zirogravity ,
Thanks. I did try placing different GNDs on either side of the transformer. The output changed centers but did not result in "expected behavior".
Yeah, sorry I can't help with expected behavior.
Note that your K statement shorthand couples every inductor to every other inductor; is that your intent? I.e.:
L1 to L2
L1 to L3
L1 to L4
L2 to L3
L2 to L4
L3 to L4
If you want to be clear about coupling, use separate K statements for each coupling.
Regarding the isolation suggestion. This is very helpful suggestion. Would I place the high value resistor on the supply side to ground or elsewhere? My original line of thinking was just the coil resistance would suffice but after your suggestion I realized the coil resistance would be too low for isolation so thank you.
You could add a COM net name (Press N) to the primary side and simply add COM-connected-via-resistor-to-GND anywhere on the schematic. This problem has several videos describing it on youtube. Check out Fesz electronics, for instance. COM is just a name, has no intrinsic value or meaning.
mike
Hi MStokowski
Such great inputs especially regarding the mutual coupling statement. Thank you.
I guess I need to reconcile the physical world from modeling. The way I have been looking at it is that any single coil on the primary side or combination there of regardless of how they are wired (using L1 alone, L1+L2 in series, L2 alone, or L1+L2 in parallel) should "excite" or "couple" with the secondary side regardless of how the secondary coils are wired. The equivalent of two inductors on the primary side should couple with the equivalent of two inductors in the secondary side.
I will explore the coupling statement and the common net name approach.
Thank you for your time and inputs.
Wess
MStokowski - Thanks for all your help and inputs. You are correct regarding where to place "COM" or grounds in the schematic. Also, Fesz Electronics videos were great. My final schematic looks something like this and generally agrees with the bench hardware setup.