Toroidal inductor for crossover

Jump to Latest
#2
Those are wound on what appears to be Micrometals powdered iron, mix 26. (or a copycat). The permeability of this material changes fairly dramatically as a function of AC flux density, so the cutoff frequency of the filter can vary with the signal amplitude. A lower permeability Kool-mu or molyperm core would be more suitable, as the pemerability changes a lot less with AC flux density.
 
#4
Since it is rated at 3A I think it would fit for a rated 40w rms tweeter for example, at least regarding power handling.
But if, as stated, the frequency of the filter can vary too much with the signal amplitude, then it wouldn't be a good choice, right?
Any other issues? I'm saying that because it's hard to find off-the-shelf inductors here in Brazil for audio purposes, and I can't spend too much, on the other hand. But if there isn't any other reasonable option out there, I will pay someone to do it for me. They charge 15 dollars for each one, I need 4 of them.
Thank you
 
#5
This might work at least for tweeter use at least. Especially if other parts are hard to get.

Maybe start over and let us know what you have for drivers and what you plan to do.

I think I have that core type. I may see if I can make a 470 microhenry choke and see if inductance really varies over different power levels like .1w , 1w 10w. Would probably test at about 3 KHZ
 
#7
Hi ozasco,

The link below describes some of the aging issues in Micrometals powdered iron cores. I should warn you that there are companies in the far east manufacturing powdered iron cores that are significantly inferior to those made by Micrometals. Visually, there is no difference.

Micrometals Powder Core Solutions

Hope this helps.

Peter
 
Last edited:
#8
ozasco said:
But if, as stated, the frequency of the filter can vary too much with the signal amplitude, then it wouldn't be a good choice, right?
Click to expand...
Correct. This causes intermodulation distortion. A core can only support a certain strength magnetic field. The magnetic field is proportional to the current flowing through the windings and the number of turns. When the core saturates due to excessive magnetic field, the effective inductance falls to that of an air core inductor - the core stops being effective at increasing the inductance of the windings. Saturation does not happen like a light switch, but follows at a smooth 'knee' in a BH curve so even at <3A some distortion will occur.

Steel laminate cored speaker crossover inductors typically have much lower permeability cores than powdered metal or ferrite cores which means that while you don't get such an increase in inductance compared to an equivalent air core (it may only be 2-3x the inductance for the same amount of wire where a powered metal may be 10-50x+), the inductor will support much higher current before the core saturates. If an Isat rating was provided it would be about 50-100A for an inductor with a steel laminate core about 0.5x0.5x2.5"

Ultimately cored inductors are a cost and weight saving measure by reducing the gauge and length of wire required to produce a certain inductance and DC resistance (if the wire is shorter it can be smaller gauge for the same DCR). The disadvantage is core saturation and the distortion it causes. A balance must be struck between sound quality and size/weight/cost. Many of the crossover inductors i've salvaged from commercially made speakers do have very small cores and probably do make appreciable amounts of distortion if overdriven. I often remove the cores from them and just use them as air cores, because cost/weight/size is often less of a priority to a DIY'er.
 
Last edited:
#10
Powdered iron toroids are extremely difficult to fully saturate due to the distributed gap inherent in their construction, but the variable permeability of core material with both DC and AC flux will make the filter cutoff frequency a moving target at high levels of excitation. Gapped ferrites are far more stable, but they will go all at once when they saturate. Lower perm (60mu and below) molyperm and sendust (Kool-mu) cores are much more stable, and not all that hard to find on the surplus market these days. Of course, if you want something impossible to saturate, there's always air core inductors, but those have other issues.

As a parting thought, you can always go with Micrometals mix 0 if you insist on using a toroid. The stuff is impossible to saturate, as it's phenolic resin with mu = 1, and gives you the advantages of toroidal geometry.
 
#12
I have read the Thread with interest as I am thinking to use ferrite material in my crossover - not a toroid but a simple rod. THere are two types, one with initial permeability u2300 and saturation 480 mT. Another one initial permeability u1800 and saturation 450 mT. Perhaps someone could tel is the material suitable for crossover at frequency of 120 Hz. Length of rod is 38 mm but only 25mm will be available after attaching left and right side of bobbin; diameter of the rod is 19 mm .
 
#14
Oh? I've had trouble finding solid data on these. If you look at this type they appear to be designed for less wire, to reduce resistance going to the bass driver. With possibly fewer turns, the core appears to be doing a lot to bring up the inductance. This also suggests that as saturation comes in it will have a significant effect. I've seen the advertising copy, it suggests that they know that we know this but that they've tried to make it workable for a crossover...you pick your priorities, this is a low resistance inductor.
 
#15
diypass - yes I believe it but their coils will be prohibitively expensive, I need something DIY

AllenB - What do you mean by "designed for less wire"? In my case, at 120Hz, second order, I need 21mH - it will be significant number of turns which ever way I look at it. I am expecting 1.5 to 2 Ohms resistance with 1mm wire. I inquired at Dyne transformers, they can make it , 0.7 Ohms, almost $200; by 4 is $800. I respect their work and I love my speakers but I will accept 2 Ohms in DIY version and see what happens.
I can choose 13 or 20mm diameter rod; is it more important to have longer or larger diameter rod?
In your opinion, what would be ideal dimensions for bobbin, if there is something ideal?
 
Last edited:
#16
diypass said:
Jantzen c-coil in crossover application never will saturate.
Click to expand...
I'll reiterate the comment I made 3 years ago. Saturation is not like a light switch. Just because you use an inductor with Isat=50A and only subject it to say 10A or even 1A doesn't mean that it does not create distortion. "Exotic" core shapes don't beat physics. Belt and braces engineering of using larger cores and/or larger wires to achieve a given inductance and DCR will always beat small cores and small wires with 'exotic' materials or geometry. I'm lead to believe that size for size and material for material, toroidal cores saturate at a lower level than I and E-I cores. The advantage is that stray magnetic fields are lower on a toroidal, however this is easy to design around in a passive loudspeaker where an enclosure is fairly large and there is design freedom to physically place crossover components sufficiently far away from each other.

Whether or not the amount of distortion is considered too much depends on the application. In most home audio applications where power is low (<100W), cheap steel laminate cores (like from ERSE) are more than acceptable. If distortion is -80dB at a few tens of watts you probably don't care. Jantzen, while making a quality product do seem to charge a premium for marketing fluff and not actual objective improvement in performance for many of their offerings.

If you want absolutely zero distortion at any power level, you have to use an air core. I would suggest that if money is no object, buy air cores with significant wire cross section if you need to achieve low DC resistance. The geometry of the wire(e.g. circular vs foil/flat), or type of insulator(s) (e.g. enameled and/or waxed) makes no difference. If low DC resistance isn't required, fantastic, a small cheap air core will do and being physically smaller it'll do a better job at reducing stray fields.
 
Last edited:
#17
Tomom said:
What do you mean by "designed for less wire"?
Click to expand...
A core that has no air gap makes the inductance appear much higher than it would be if there were no core, so this is a way to turn a small inductance into a large one.

Designing wound devices is involved, and unless I have a project in front of me then I'm not going to be in the mindset. However, if you know what you need regarding current, self-resonance frequency, resistance and distortion then you may be able to design it in air, solenoid or shell core.

You would have to go through the designs to know which one is going to be cost effective, acceptable size, sensitive to its environment. This is why DIYers often either ask someone who does it regularly to do it for them, or take a suitable looking power transformer and start de-constructing it to make their own.
 
#18
AllenB said:
A core that has no air gap makes the inductance appear much higher than it would be if there were no core, so this is a way to turn a small inductance into a large one.
Click to expand...

Since powdered iron cores are internally gapped (unlike ferrites), their inductance cannot be increased as the internal gapping cannot be removed. Thus, all one needs to do is calculate the inductance using L= AL * N^2. The Micrometals -26 is an SMPS grade material that is designed for soft saturation (see datasheet) and is therefore not suitable for crossover design.

However, if the OP still wants to use this material, a more cost effective solution would be to salvage these yellow cores from scrap ATX power supplies (plentiful). If I remember correctly, the AL value for a 1" diameter yellow/white core is 96nH/turns^2.
 
Last edited:
#20
  • Like
Reactions: Audio>X
You must log in or register to reply here.
Top Bottom