[Nixie]

How should I model a transformer with more than two windings in SPICE? For two windings, the standard way is to have two inductors mutually coupled, such as:
K1 L1 L2 1
Where 1 is a coupling coefficient of one. Now, if there are more than two windings, should the coupling coefficient be still the same (say, K1 L1 L2 L3 L4 1), or does it need to be split up somehow?

 

[chesebert]

oh man. brings back to those school days.

sorry I don't really know the answer as my concentration was in CPU architecture and digital circuits..all these opamp and 24 stage RCL circuits are making me dizzy...lol

I am sure if you search around you will find a already modeled spice code which you can then just plug and play.

 

[Fresno Bob]

....

 

[rickcr42]

general info :

Linear Technology has an exeelant free spice program that has a support group at yahoo Groups where you can obtain many many more files than are offered by LT with full support.I use this pspice program and it is a good one

 

[Fresno Bob]

.....

 

[Nixie]

rickcr42, I'm using that exact program, and the diagram in the first post was drawn in it, but I made it B&W to save on bandwidth (while it makes no difference for JPEGs, it does for PNGs, which is a better format for these kinds of images, as it doesn't blur the sharp lines and text for the same level of compression).

From the help file of LTSpice, as well as the web, a (near-ideal) transformer can be simulated in SPICE using two inductors and a mutual coupling constant (usually set to 1 as lossy coupling makes the simulation very slow). My question is, to extend this to multiple secondaries, is the coupling constant still 1 pairwise between all inductors, as the extended command line I quoted would implement it, or is it different (in which case it needs to be specified for each pair)?

Thanks for pointing out the yahoo group. I'll also post my question there.

 

[rickcr42]

I understand what you are attempting and it is not the most efficient way to go at it.if you sign up for the Yahoo group you will have access to batch files and spice files which will model what you need far better and make the end results much more accurate.good luck man.

rick

 

[chesebert]

yeah. once you get a nice simulator (cadence) with spice II built in you just have to plug in the modules and just point and click your way to your result..lol

 

[SnoopyRocks]

Rather than explain it myself, I'll direct you to some reading material. If one were to google "spice transformer model," the anwer would present itself. Look at the fourth link, especially page 4. The document explains how to model transformer nonidealities as well.

 

[Nixie]

I don't need such a complex model that slows down the simulation. I just wanted to use the simple coupled inductors model, and I was trying to make sure that I was specifying the coupling right. No matter, the yahoo group that rick suggested answered my question.