Genome used to have a calculator available on his website, but I don't know if it still exists.
Genome's program estimated some of my inductors' a.c. resistances as
3-5x their d.c. values--ouch! That epiphany guided me to fewer layers and experimentally braiding my own 29 AWG x 12-strand Litzendraht for a small 300kHz supply's inductor (but in the end the problem was core loss).Cheers, James Arthur
Yes, 3 to 5 is not unusual; if the coil has enough layers and is long compared to its diameter, the Rac/Rdc ratio can exceed 20. Scary stuff, and a big motivator for the use of litz wire.
yes this is true as in spice you cant specify a frequency dependant resistor, but I usually chose a value that is close for the frequency.
obviously for smps this includes a dc current as well, here an aditional parallell resistor would in fact be a much better model, I had assumed the op was ony using a parellel r.
you could use the series resistance to model dc resistance and work out a parallel resistor to give the same overal disipation at a specific frequency calculated from the frequency dependant series resistance from data sheet if it has it.
most inductor data sheets only give series resistance v freq. some do however give complete eqv circuits including series restance and also parallell rc networks wich more acuratly model the losses.
Colin =^.^=
thinking about it a bit more ... ive often thought how cool it would be to be able to have a model wich allowed you to just enter physical details of the windings/core and the nature of the core material.
might need some considerable maths to be done to convert the physical parameters to usable eqv spice circuit, but shldnt be imposible, even to take into acount core saturation.
shld be quite easy to make a model wich is a lot better than just rule of thumb.
you could even transform it into laplace functions, wich tbh I dont know wel enough to use often ... actually just googling for "spice laplace inductor model" gives a whole load of hits for doing this very thing ...
I dont know why spice doesnt have this sort of model. its not like its fantasticaly more complicated than any of the semiconductor component models for example where you can enter chanel length etc ...
Colin =^.^=
AMS Simulator - Magnetic Parts Editor (PSpice)
Charlie
Have something to add? Share your thoughts — no account required.
Ask the community — no account required