3
\$\begingroup\$

It is common to see capacitors in parallel, with smaller and smaller capacitance, in order to reduce the parasitic ESR (and the equivalent series inductance). This allows obtaining relatively big capacitances of very low ESR. But I've not seen in the literature that a similar, dual trick, is used for inductors: to connect in series inductors of smaller and smaller inductance (and hence, generally speaking, of smaller and smaller parasitic effects), with suitable shielding. I had a look at various inductor models, and if my understanding is correct, this would reduce enormously the parallel capacitance (probably the most annoying parasitic element), as well as the parasitic parallel resistance (magnetic losses). Of course, this would also somewhat increase the ESR of the inductance, but this is not a problem in general.

My question is: is this idea correct, and where is it dealt in the literature?

\$\endgroup\$
  • \$\begingroup\$ If one needs an ideal inductance then one usually uses a gyrator instead. \$\endgroup\$ – Ignacio Vazquez-Abrams Sep 28 '16 at 11:05
  • \$\begingroup\$ Maybe I'm stretching definitions a bit too much but.. a Litz wire inductor is sort using "different" inductors do minimize parasitics right? Something that I think is called "spiderweb" coil is also a topology made to reduce parasitics (by having strands of conductor less parallel than if directly wound). I think there are also some audio inductors that use foils instead of wires. \$\endgroup\$ – Wesley Lee Sep 28 '16 at 11:05
  • \$\begingroup\$ Ignacio, a gyrator behaves like an inductor only at a definite frequency no ? \$\endgroup\$ – MikeTeX Sep 28 '16 at 11:36
  • \$\begingroup\$ Wesley, If my idea is correct (and it is probably not), this may reduce the parasitic effects much more than a special winding topology. \$\endgroup\$ – MikeTeX Sep 28 '16 at 11:46
  • \$\begingroup\$ This depends on what you're trying to do. For some applications like power conversion the ESR is extremely important, and the parallel capacitance less so. \$\endgroup\$ – John D Sep 28 '16 at 15:46
3
\$\begingroup\$

But I've not seen in the literature that a similar, dual trick, is used for inductors: to connect in series inductors of smaller and smaller inductance (and hence, generally speaking, of smaller and smaller parasitic effects), with suitable shielding.

This is a very common practice when designing bias tees. A bias tee requires an inductor that can block a wide range of frequencies, for example, 10 kHz to 20 GHz in one application I've worked on. So a low-value, but also high series resonant frequency (SRF), inductor is connected directly to the input trace to block high frequencies, while one or more higher-value (and lower SRF) inductors are connected in series to block lower frequencies.

Here's a high-end bias tee taken apart: enter image description here

You can see a fairly complicated network of radial coils, toroidal coils, and ferrite beads used to obtain wideband blocking.

\$\endgroup\$
  • \$\begingroup\$ Exactly the answer I dreamt. Thank you so much the photon. \$\endgroup\$ – MikeTeX Sep 28 '16 at 16:26

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.