I understand that the hysteresis etc starts at 1khz in the iron coils.
The wires need to be accurately wound in circles, even and perfectly.
If you use air-core you don't need ribbon ones unless you are in series with high pitch frequencies.
Some XO have a 0.05mh in series with tweeter to filter ultrasonics, such filter would beneficiate from a ribbon coil.
Other than this plain air cores are doing fine, I cannot hear any difference, the only difference should be attributed to the resistance and small deviations in geometry which there are no way to detect.
The wires need to be accurately wound in circles, even and perfectly.
If you use air-core you don't need ribbon ones unless you are in series with high pitch frequencies.
Some XO have a 0.05mh in series with tweeter to filter ultrasonics, such filter would beneficiate from a ribbon coil.
Other than this plain air cores are doing fine, I cannot hear any difference, the only difference should be attributed to the resistance and small deviations in geometry which there are no way to detect.
Coil saturation
See Crossover coil saturation | Audiokarma Home Audio Stereo Discussion Forums
... all about why I prefer air coils if at all possible.
Andre
See Crossover coil saturation | Audiokarma Home Audio Stereo Discussion Forums
... all about why I prefer air coils if at all possible.
Andre
In fact that's probably about the worst way electrically(!), but the easiest to manufacture. For low self-capacitance it's best to wind haphazardly from one end, or diagonally, and/or in bunches separated with spacers, or with the turns spaced apart in a single layer. Close winding boustrophedonically in perfect layers is the most reproducible method (its easier to sell a standard product with poorer specs than a variable product with better average specs), but has high self-capacitance and high proximity-effect losses limiting the frequency.
For cross-overs at audio frequency it doesn't matter, for RF coils you often need to consider self-resonant frequency and may need a modified winding scheme to raise the resonant frequency. For instance a lot of RF chokes are wound criss-cross diagonally in separate segments so they act as chokes, not capacitors at the desired frequency.
Using an air core means needing far more turns, which will make the self-capacitance (and copper losses) a bigger issue.
I am not aware of any ferro- or ferri-magnetic material with stable parameters (they vary significantly with temperature, sometimes with age, and have non-linear inductance that depends on the current).
You counter this by using gaps in the magnetic circuit, so that the air-gap becomes the dominant factor (provides most of the reluctance), and air has constant magnetic properties (essentially the same as a vaccum). This linearises and stabilizes an inductor by a large factor. So any inductor not wound on a continuous core will be much more linear - for instance a cylindrical core has a huge gap between the two ends and is commonly used in cross-overs (but sprays magnetic field around, of course). Iron dust cores are special - they are iron particles in an insulating matrix, a kind of distributed gap, and are well behaved and stable - used for RF tuned-transformer cores for instance.
Typically a core without gap (and not made of iron-dust matrix) is used for a choke or transformer, inductors with a defined value are normally gapped.
The linearising effect of the gap can't prevent saturation, but can delay distortions until closer to the limit.
It all works like negative feedback - you throw away magnetic permebility and gain stable linear behaviour in return. Upto a point of course.
For instance you might use a gap to turn a core with a relative permeability of roughly 2000 to something more like 20 but much more stable. You have to use more turns, but not as many as with an air-core, so you still have significant benefit in size and copper losses.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.