I am trying to design a low-pass filter for a 24V, DC, 10A power line to suppress EMI, RF interference and noise in a bus. There is no ground(chassis). I am trying to simulate this filter that was tried in the bus and worked (the audio amplifier was noise-free). This filter is the filter on my previous question.

I couldn't calculate the cutoff frequency because of the common-mode chokes. Could you help me calculate?

I am trying to simulate it, however it doesn't seem right. Am I simulating it right?

Here are the schematics and graphs from LTSpice:

Schematic Graph1 Graph2

  • 1
    \$\begingroup\$ You have high resonance peaks which might be due to no series equivalent resistors in the chokes. Even though you are using 10A rated chokes, they will have some resistance, which will damp the peaks. \$\endgroup\$ – Martin Oct 5 '11 at 12:11
  • \$\begingroup\$ Thanks! This really damped the peaks. Now that I can see that this filter has -6dB at about 160 MHz. Still this doesn't seem right. \$\endgroup\$ – abdullah kahraman Oct 5 '11 at 12:41
  • \$\begingroup\$ The filter you link has common mode inductors, while your simulation will have simple inductors. \$\endgroup\$ – Connor Wolf Feb 23 '12 at 11:27
  • \$\begingroup\$ @FakeName I thought K L1 L2 1 is for simulating it as common mode inductor. How can I simulate common mode inductors? \$\endgroup\$ – abdullah kahraman Feb 26 '12 at 15:06
  • 1
    \$\begingroup\$ @abdullah kahraman - I didn't see that. I would have thought a common mode inductor would have a different schematic primitive. Anyways, apparently you know more about SPICE them I. \$\endgroup\$ – Connor Wolf Feb 27 '12 at 6:24

You have high resonance peaks which is due to no series equivalent resistors in the chokes. Even though you are using 10A rated chokes, they will have some resistance, which will damp the peaks considerably. In other words, there is considerable difference between ideal component parameters and what is achievable in a practical component that has a big effect on the simulation.

| improve this answer | |

It looks like you might have not selected the current symbols active load check box.
Try right clicking, selecting advanced, then checking "this is an active load" - it should then show "load" next to the symbol.
Also, I think the series capacitors junction needs to be connected to ground (the 2400p caps) and try setting the coupling factors to slightly less than 1 (e.g something like 0.98)

| improve this answer | |
  • \$\begingroup\$ Nope, it is not about the load. However, if I make the coupling factor 0.98, resonance peaks arise. \$\endgroup\$ – abdullah kahraman Oct 5 '11 at 15:13
  • \$\begingroup\$ Did you connect the capacitor junctions to ground? \$\endgroup\$ – Oli Glaser Oct 5 '11 at 17:35
  • \$\begingroup\$ There is no ground in the bus. This was connected without the chassis connection and worked. That's why I want to simulate without ground connection. \$\endgroup\$ – abdullah kahraman Oct 5 '11 at 19:32
  • \$\begingroup\$ "ground" may just mean the neutral line for the purpose of simulation. However, I noticed your graph is from Out to ground (single ended) Try a differential probing from Out to the bottom of I1 (click and drag from one wire to the other, you should see a red and black probe) \$\endgroup\$ – Oli Glaser Oct 5 '11 at 20:05
  • \$\begingroup\$ Here is the differential measurement with 10mohm inductor series resistances and 100mohm capacitor series resistances i.imgur.com/UgLdf.png \$\endgroup\$ – abdullah kahraman Oct 6 '11 at 7:26

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.