AIR-Core Vs Ferrite-Core Inductor in Class-D amplifiers

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Recently i look around several class-d amps which used air-core as well as ferrite core inductors at their outputs to suppress switching frequency component.

AS I am working on class-D amp project , Kindly Help me which filter should i use for suppressing switching frequency components and best performance in terms of sonic quality.:)
 
Have a look at this one:

http://www.lcaudio.dk/com/zpn10.jpg

(Hi Lars ! ;) )

The result is less core-saturation but also a lower AL value than the original core-set. But you still need less turns than a similarily sized air-core coil would need. Another advantage is that there is less EMC problems than using air-core coils.
I once thought about using a toroid-shaped "air" core. But that one would need much more windings than a cylindrical coil.
Someone else (on this forum, but I don't remember who it was)proposed two coils beneath each other connected in a way that they are cancelling out the stray field (in fact a cylindrical coil that is bent at half it's length). This could be a good compromise either.

Regards

Charles
 
Commentable Thoughts

SIR I DONT have access to ferrite cores right now, can u PLZ tell me that what are the disadvantages air core inductors and how can i calculate the number of turns for a given inductance and area of a core, can u plz tell me some formula to calculate it.
 
I guess using as big coil as possible makes current changes small?

Are you talking about the OUTPUT-FILTER inductor ?

If so, the peak current in the inductor at idle is:

carrier_peak_voltage / (2 * carrier_frequency * inductance)


Please keep in mind that this current is reactive and is only causing some conduction losses in the output stages, while too large a coil would lead to a low filter cutoff frequency.


Regards

Charles
 
Hi Charles

Someone else (on this forum, but I don't remember who it was)proposed two coils beneath each other connected in a way that they are cancelling out the stray field (in fact a cylindrical coil that is bent at half it's length). This could be a good compromise either.

Hi Charles, thanks for remembering this idea.;) Sometimes I rather stumble onto different things.

Amp_man, what I do is just try differing numbers of turns until the performance is optimized. But a ball park figure may be about 20uH for 30 turns on a 1" form for an air core choke. Using that as a reference, the inductance is proprtional to the square of the number of turns.
 
I'm planning to take feedback after output filter, and the topology will be self-oscillating (+ maybe some periods of doing nothing in case of small signal, this can be changed in the programmable logic chip). Still that leaves the question open about how much does the speaker and speaker cable capacitance affect self-oscillation? I read somewhere that cable capacitance can be around 1 nF or less. Assuming i use 100 uH and 100 nF in LC filter, it would seem that cable is not a problem, adding only 1nF to the C, but i'm not sure about speakers.
 
There are HUGE emc-problems associated with air core inductors in power switching. So much so that there is no point in trying, if the amplifier has to comply with IEC standards or anything similar.

DIYers can choose to ignore it.

Gapped cores also have some of these properties. To minimize stray magnetic fields in these cores, keep the gap inside the coil, like if you use a double E-core, only gap the center leg.

There are many advantages to using cores, and there are good materials out there to. If you are going for a toroid design, then MPP cores should be high on the list. These are expensive though. The cool-mu cores are a less expensive, and an ok alternative.

Gapped 3F3 or 3C90 ferrite cores are good if you use E-I og E-E combinations.
 
There are HUGE emc-problems associated with air core inductors in power switching. So much so that there is no point in trying, if the amplifier has to comply with IEC standards or anything similar.

Since it has been done it must be feasible, althoug not easily I have to admit.
Since we are DIYers and don't want to sell our stuff we don't care that much about complying with any standards.
As long as we don't disturb our neighbour's TV's and hearing-aids etc and we don't take any planes down with our equipment, it is O.K for us. We don't have any chances to test for compliance anyway.

Gapping only the center leg can be quite a task. For us DIYers it is much easier making the gap by inserting some insulating material. The EMC problems we had with a class-d amplifier were definitely not caused by the air gap.

Before I try to get some toroid made out of near unobtainium I get a standard RM, EFT or whatever core and gap it by myself.

Regards

Charles
 
There are HUGE emc-problems associated with air core inductors in power switching. So much so that there is no point in trying, if the amplifier has to comply with IEC standards or anything similar.

DIYers can choose to ignore it.

Gapped cores also have some of these properties. To minimize stray magnetic fields in these cores, keep the gap inside the coil, like if you use a double E-core, only gap the center leg.

There are many advantages to using cores, and there are good materials out there to. If you are going for a toroid design, then MPP cores should be high on the list. These are expensive though. The cool-mu cores are a less expensive, and an ok alternative.

Gapped 3F3 or 3C90 ferrite cores are good if you use E-I og E-E combinations.

What about air core inductors in foil version like this
Tobias Jensen Production A/S - Air Core Inductor 14 AWG, copper foil/paper
or this
Mundorf - Inner Excellence ?
Usually are to find inductors in this kind for the output filter:
Coilcraft Critical Products | Power Inductors | Surface Mount Inductors, High Temperature | Military Grade | Tin Lead - Coilcraft CPS
but the TacT Millenium uses air core inductors
http://www.hifihenkes.de/produkte/archiv/tact/tact_m1_inside_pic.htm
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.