Learning Forum

[Serdar KALAY]

I wanted to design a coil in Maxwell 2D (cylindrical about Z). Specifications of my coil in reality are like this: It has 80 turns, 1 layer, its material is copper and it consists of a litz wire which has 125 strands each of which has 0.0799mm diameter. Litz wire (125 strands all together with insulator around them) itself has 1.2mm radius.  After creating a rectangle and assigning a coil terminal to it, I specify the number of conductors as 80. I set the material type as copper, compositions as litz wire, strand number as 125, wire diameter as 0.0799mm. I arrange the length of the rectangle as 80(turns)*1.2mm= 96mm. What makes me confused is that what happens to the wires when we change the length of the rectangle? It does not give any error when I set the length shorter than 96mm or make it longer then 96mm, how does Ansys arrange the location of litz wires in the 80 conductors that I set, does 1 conductor contains only 1 litz wire?

Another question I have is that is it possible to create multiple layers in a single rectangle (cross section of a coil). As above if I have a coil that has 80 turns but has 3 layers instead of 1 what should I do. I thought to set number of conductors as 3*80 = 240. However, I am not sure if it suffices because if Ansys divides rectangular cross sections to 240 subsections and locate them next to each other as if they are the 240 turns of a single layer, I will not be able achieve my goal. I also tried creating three different rectangles next to each other to represent different layers, however; since my design contains 16 different coils and each coil has at least 4 layers it does not seem as an effective solution.             

Finally, is it possible to measure the self-resonance of a coil in Ansys Maxwell? Because of the potential difference between the turns and layers of a coil, coil also have a capacitance and for a specific frequency impedance of the coil must have peak value. I did a frequency sweep in the eddy current solver type for a coil with 20 turns and 1 layer. Even though I made a frequency sweep from 1Khz to 100Mhz impedance of the coil increased linearly. Normally it is expected to have a sharp peak at some point and decrease afterwards.  I did this in both 2D and 3D designs. Using cylindrical representations instead of helix could be the reason why I am not seeing a self-resonance effect or isn’t Ansys Maxwell capable of it?

 

[HDLI]

Hello Serdar,

       Maxwell will determine wire locations based on litz wire size and strand automateically. There is twisting length factor to adjust length in 2024 version.

       You could separate coils into 3 layers. 

       For self-resonance, it is better to use Q3D or HFSS to get RLC together. Otherwise, you need to use both electrostaics (Capacitane) and eddy current solver (R and L) to make co-simulation between Simplorer and Maxwell, to determine the resonant frequency.

HDLI