One of the recurring questions that seems to be of interest to speaker builders is the matter of copper foil coils- does it make a difference, and is it worth the cost?
Well, first off, I'll be the first one to say, "Well, try it, listen to them in your own projects....see what you think..."
But some folks have been noticing there aren't any foil coils in the projects I've "published" to date, and wonder if it's something I'd consider, and when.
Now, a lot of you probably realize that my day job is technical, having something to do with power semiconductors. All true. More to the point, my background is SMPS design (prior to joining Siemens Semiconductors, now known as Infineon). In Switch Mode Power Supplies, we use coils (inductors), and sometimes at fairly high frequencies. How high? Well, some of my past projects, even 10 years ago, were operating in the 500 kHz region Quite a bit higher than CD Audio, in fact, that's almost the bottom of the AM Radio band; and those are square wave converters, with rise and fall times measured in a few nanoseconds. So, for the sake of arguement, let's say I have some familiarity with HF power conversion.
OK, the arguement for foil inductors is to reduce skin effect. Hmmm, good idea. I worry about that in power conversion, and in fact, usually use custom litz wire conductors or planar foil conductors in high frequency applications for inductors and transformers- generally, above 100 kHz. In fact, I use a simple MathCAD calculation to compare the resistance of the wire with the total impedance at a given frequency due to the combination of resistance and skin effect. That's not hard to estimate for circular conductors at all, like conventional magnet wire. It's a good idea to do, because often, in a given frequency range, it may be more cost effective to use bundled smaller solid conductors, rather than go full tilt boogie to Litz wire.
So, I've been scratching my head and wondering, should I use foil inductors in some special area's in crossovers, maybe in things like the notch filter for the RD50's, where it's in the series path for the full range signal.
The typical inductor I use is always air core, and usually wound with 14 or 15AWG wire. Let's use the 14 AWG for an example for skin effect calculations, as it's larger in diameter and more likely to have problems.
Since I can't run MathCAD on the web, I'm going to past in some screen shots. Since we're talking about audio, we dont' have to worry about the harmonic spectrum of square wave converters- let's just take a stab at this at 10 kHz, OK? Ideally, we don't want any skin effect, because although it's affect is just like inductance, it's a secondary effect which adds into the inductance, and increases losses in the coil, changes the Q of the circuit, and changes the rate of roll off.
Hmmmm, I think we can say with some safety, that at 10 kHz, and with a wire no thicker than AWG 14, skin effect isn't yet becoming an issue. BTW, if you redo this for 100 kHz, it's another story altogether. So maybe for your ionic super tweeter crossover....
Now, this is just the top level of the document, the functions definitions are implemented below this, from standard "text book" equations. (I'm not inventing anything here, folks). Here's the gory details for those of you with that interest...
This jibes with the practice I've had in SMPS, where above 30-50 kHz I generally prefer small gauges or specialty wire. For SMPS converters, it's the actual power loss which is measurable in terms of the heating affect on the inductor.
Would a foil inductor of the same DCR be audibily better in a 5-10 kHz crossover? I'm not saying it can't be, but at this time, I'm likely to be another guy from Missouri- show me. At a 1.25 kHz crossover, I'm sceptical. :roll:
Be on the lookout- the better foil inductors have about the same DCR as good wire inductors- but I've seen some advertising a given AWG that don't have as low a resistance- there's something odd going on there. Match the DCR to the application- there's not point using a hot shot inductor in an LCR zobel, when all the more expensive part is going to be doing is making the vendor some more money.
Best regards,
Jon
Earth First!
_______________________________
We'll screw up the other planets later....
Well, first off, I'll be the first one to say, "Well, try it, listen to them in your own projects....see what you think..."
But some folks have been noticing there aren't any foil coils in the projects I've "published" to date, and wonder if it's something I'd consider, and when.
Now, a lot of you probably realize that my day job is technical, having something to do with power semiconductors. All true. More to the point, my background is SMPS design (prior to joining Siemens Semiconductors, now known as Infineon). In Switch Mode Power Supplies, we use coils (inductors), and sometimes at fairly high frequencies. How high? Well, some of my past projects, even 10 years ago, were operating in the 500 kHz region Quite a bit higher than CD Audio, in fact, that's almost the bottom of the AM Radio band; and those are square wave converters, with rise and fall times measured in a few nanoseconds. So, for the sake of arguement, let's say I have some familiarity with HF power conversion.
OK, the arguement for foil inductors is to reduce skin effect. Hmmm, good idea. I worry about that in power conversion, and in fact, usually use custom litz wire conductors or planar foil conductors in high frequency applications for inductors and transformers- generally, above 100 kHz. In fact, I use a simple MathCAD calculation to compare the resistance of the wire with the total impedance at a given frequency due to the combination of resistance and skin effect. That's not hard to estimate for circular conductors at all, like conventional magnet wire. It's a good idea to do, because often, in a given frequency range, it may be more cost effective to use bundled smaller solid conductors, rather than go full tilt boogie to Litz wire.
So, I've been scratching my head and wondering, should I use foil inductors in some special area's in crossovers, maybe in things like the notch filter for the RD50's, where it's in the series path for the full range signal.
The typical inductor I use is always air core, and usually wound with 14 or 15AWG wire. Let's use the 14 AWG for an example for skin effect calculations, as it's larger in diameter and more likely to have problems.
Since I can't run MathCAD on the web, I'm going to past in some screen shots. Since we're talking about audio, we dont' have to worry about the harmonic spectrum of square wave converters- let's just take a stab at this at 10 kHz, OK? Ideally, we don't want any skin effect, because although it's affect is just like inductance, it's a secondary effect which adds into the inductance, and increases losses in the coil, changes the Q of the circuit, and changes the rate of roll off.
Hmmmm, I think we can say with some safety, that at 10 kHz, and with a wire no thicker than AWG 14, skin effect isn't yet becoming an issue. BTW, if you redo this for 100 kHz, it's another story altogether. So maybe for your ionic super tweeter crossover....
Now, this is just the top level of the document, the functions definitions are implemented below this, from standard "text book" equations. (I'm not inventing anything here, folks). Here's the gory details for those of you with that interest...
This jibes with the practice I've had in SMPS, where above 30-50 kHz I generally prefer small gauges or specialty wire. For SMPS converters, it's the actual power loss which is measurable in terms of the heating affect on the inductor.
Would a foil inductor of the same DCR be audibily better in a 5-10 kHz crossover? I'm not saying it can't be, but at this time, I'm likely to be another guy from Missouri- show me. At a 1.25 kHz crossover, I'm sceptical. :roll:
Be on the lookout- the better foil inductors have about the same DCR as good wire inductors- but I've seen some advertising a given AWG that don't have as low a resistance- there's something odd going on there. Match the DCR to the application- there's not point using a hot shot inductor in an LCR zobel, when all the more expensive part is going to be doing is making the vendor some more money.
Best regards,
Jon
Earth First!
_______________________________
We'll screw up the other planets later....
Comment