200nH inductor from PCBA traces

-Lasse

femm 4.2 by Dr. David Meeker with a VERY helpful 24/7 support group

free and ACCURATE, because you include he frequency, has thermal which might help too.

I use it all the time to characterize transmission lines, cables, PCB performance, eddy current losses, etc etc.

PS: remember your field from this PCB coil will be everywhere, even 'through' GND layers.

I've designed polyimide circuitry to go in a vacuum. The circuitry was super high speed, super low noise DAC system for accurate e-beam positioning. Believe me, when I say, fields go through EVERYTHING!. so, use femm to determine how much your field penetrates and gets into the rest of your circuitry. [also if you're after high speed, using PSpice or LTspice you can do a lot of modeling of your cabling and such to use the information from femm and INCLUDE radiation, not just transmission delay, and sometimes radiation has some weird effects on your 'terminal' voltages too.]

good luck.

On Saturday, 9 August 2014 22:48:14 UTC+10, Mook Johnson wrote:>

inductors cant be used due to extreme high temperatures >> 125C. I'd like to just spin one in the PCBA traces as an "air" (polyimide)..core.

put caps ( 4 X 1210 ceramics) to make a Pi filter (1C-l-3C) on the output stage. I only need ~0.2uH to meet the ripple spec. Right now, ripple i s 200mV p-p and I'm looking for < 75mVp-p (@ ~250kHz). slugging it with m ore low ESR caps is getting expensive from a board real-estate point of vi ew. This thing needs to be tiny.

corner of the core to increase the perm but am concerned there may be so me transformer action that is hard to predict and may cause increased ripp le instead of decreased ripple.

width, # layers, diameter, etc for an inductor like this.

Frederick W. Grover's book "Inductance Calculations" from 1946 was recently reprinted as ISBN-23 978-0-486-47440-3, ISBN-10 0-486-47440-2. I've got a copy.

Page 60 gives the formula for an octagon of round wire (of radius r in cm) as

L = 0.016 S[ln(s/r) + 0.21198 + u/4] uH

where S is the length of the side of the octagon in cm and u is the permeab ility (probably 1 in your case).

may be more immediately useful.

Then you're screwed, because barium titanate (high-K ceramic caps) have a Curie temp around there as well.

Powdered iron and C0G are still on the table.

I really doubt that you'll get enough inductance in a small enough area. If you can't do it with an off-the-shelf air core inductor, you sure aren't going to do it on the board.

Tim

-- Seven Transistor Labs Electrical Engineering Consultation Website:

Their 2222 size can do 220 nH in about a tenth of a square inch of PCB, probably less than a spiral inductor would need.

FastHenry from MIT (open source) can calculate the inductance and mutual inductance of arbitrary structures. You still have to create the file that describes the geometry of the structure. You would want to write a script to automate that. I used it together with FastCap to make models for on-chip inductors (and verified models against various expensive tools, and eventually against real chips).

There is a viewer for helping to debug the model geometry on this website:

You will probably do a lot better with an air-core wire-wound inductor than with a PCB trace.

Chris

Welcome back, Win!

Cheers

Hi, Win!

board.

PCB,

Happy to discover you posting again. Be welcome.

?-)