Common mode choke puzzle

I needed an inductor for a signal filter, very low current. I have some small common-mode chokes. I measured the winding inductances as 1.7mH each, thinking I'd get 4 times this with the two windings in series, but that only measures 3.2mH, a bit less than double (or zero if the windings are connected opposing).

This happened with a few different value CM chokes, all similar to KEMET SC-02-200.

formatting link

I know that CM chokes aren't designed for this, but can they really be saturating at whatever small level my meter uses, or is there some other explanation?

Cheers

--
Clive
Reply to
Clive Arthur
Loading thread data ...

Quite the opposite, you're probably down in the initial permeability region. Try driving the doubled winding with double the voltage, if you can.

k is typically around 0.98 so the series-aiding should be close to 4x, assuming equal volts/turn that is.

If you're down in the initial permeability range, note also that it will continue to rise, until, uhh, somewhere in the 10 to 100mT range maybe, and then saturation (as such) will be kicking in, and average permeability drops.

Obviously -- don't count on this signal filter to do anything very precise, especially if amplitude is varying, and if low distortion is critical!

Tim

--
Seven Transistor Labs, LLC 
Electrical Engineering Consultation and Contract Design 
 Click to see the full signature
Reply to
Tim Williams

Thanks Tim. How I hate magnetics. The more I find out the more there is to know.

What would be a good core type for 2mH and 10mH low signal level inductors for Pi filters in the range 1kHz-200kHz?

Cheers

--
Clive
Reply to
Clive Arthur

That's gigantic for low current use.

Winding capacitance often fools inductance meters; what instrument are you using, at what frequency?

CM chokes can be complex. I sometimes just use two separate ferrite beads for the equivalent CM filter; I can understand that better.

--
John Larkin   Highland Technology, Inc   trk 

jlarkin att highlandtechnology dott com 
 Click to see the full signature
Reply to
John Larkin

Good for what?

Those values are commonly stocked and have okay Q at those frequencies, why not buy off the shelf?

Tim

--
Seven Transistor Labs, LLC 
Electrical Engineering Consultation and Contract Design 
 Click to see the full signature
Reply to
Tim Williams

Well, if doubling the number of turns on my CM choke doesn't quadruple the inductance for the reason you gave, then I guess that type of core material isn't good for a filter - the inductance would seem to vary with excitation amplitude. That suggests that there are other types which aren't good either, and thus some types which are better.

As you can tell, this isn't my area of expertise. Yes there are plenty to choose from, they seem mostly (if not all) to be made for SMPS's - will this type they work well in filters too?

Cheers

--
Clive
Reply to
Clive Arthur

What are you filtering, and to what end?

Tim

--
Seven Transistor Labs, LLC 
Electrical Engineering Consultation and Contract Design 
 Click to see the full signature
Reply to
Tim Williams

Low level signals in the range 10kHz to 100kHz to reject lower frequency (~1kHz) higher level interference (can't be removed at source) so the high gain amplifier doesn't clip.

The simulation low pass with 2 L's and 3 C's (actually 6 C's, it's balanced) works very well, but the simulated inductors are perfect barring the series R.

Cheers

--
Clive
Reply to
Clive Arthur

You might consider an active filter; inductors get big at low frequencies.

There is a form of Sallen-Key filter where the first stage is a passive RC, which absorbs the first level of noise without overloading or rectifying.

I suppose one could make a filter with an initial RCRC too.

--
John Larkin   Highland Technology, Inc   trk 

jlarkin att highlandtechnology dott com 
 Click to see the full signature
Reply to
John Larkin

Yes, I had considered that, but it's an existing low-noise amplifier and I'd rather not mess with it too much, and the simulation of my LC filter design is just right. I have some 10mH inductors which are about 8mm diameter 10mm high and I have plenty of room.

Cheers

--
Clive
Reply to
Clive Arthur

Clive, I probably know less about magnetics that you, but if you get an inductor (OTS) with more turns then it may avoid this low current (low B-field) region.

As an aside can, can you use an active filter? Maybe one of those Multi-pole ones from Linear. LTC1562... well as long as price is not a big issue.

George H.

Reply to
George Herold

An active filter would help where hum pickup is a problem with inductors.

Reply to
Steve Wilson

If you want a well-defined amplitude response, use a gapped core - any non-linearity in the permeability of the core material matters a whole lot less.

If you are simulating an inductor in LTSpice you can use the John Chan model to simulate the effect of core non-linearity.

Farnell stocks a gapped R8 core.

formatting link

It offers 0.4uH per root turn which means you'd need a 158 turn winding to get 10mH +/-3%. For 2mH you'd need 71 turns.

The gap has dropped the inductance by a factor of 400/3300 - what you see is 7/8th air-gap and 1/8th ferrite - so weird initial permeabilities become less important.

The ungapped version of the core offers 3.3uH +30%, -20%. per root turn, not a way to get a well-defined inductance.

--
Bill Sloman, Sydney
Reply to
bill.sloman

Thank you Bill, a good explanation. I have some gapped pot cores somewhere, IIRC Al = 400nH...

Cheers

--
Clive
Reply to
Clive Arthur

I didn't, so I ordered some. RS components didn't seem to realise that the two halves of the pot core are different, so a slight delay while they sent another set. Anyway, it works well and the measured response is almost indistinguishable from the simulation.

Cheers

--
Clive
Reply to
Clive Arthur

And you should know what to do about getting something smaller if you need it.

RM cores go down to RM4

formatting link

finding somebody who stocks them isn't always easy, but Mouser seems to have a decent range which does include gapped cores.

There are even smaller parts around, but getting them with the sort of gap you want is even trickier.

Element 14 (Newark to you) isn't as comprehensive. Specialist suppliers can do better - when I was living in the Netherlands I found a ferrite specialist in Germany who stocked a wide variety of parts and would ship small quantities.

The UK was never that well served.

--
Bill Sloman, Sydney
Reply to
bill.sloman

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.