No. I change the inductor. And usually, on well designed XOs, there're resistors and capacitors.
Note: a first-order crossover is not well designed, it's crap.
7 strands (6 around 1) is a stable twist configuration, I use it in my work.
The advantage of litz, as abraxalito stated, is it reduces the proximity losses typical of inductors as the frequency goes up. Proximity losses in air core crossover inductors is easily measured within the audio band.( edit: for example, a .25 mH air core 18 gauge will double it's resistance by 5 Khz, from .18 ohms to .4 ohms) I don't measure Q, but that is a very good way of doing it. Rather, I measure the series resistance increase which is caused by proximity effect.
What neither measurement does is show exactly what the resistance increase is vs time. All the agilent meters I've used simply report a number, whereas the actual proximity induced resistance is time dependent. It modulates at twice the drive frequency.
Litz inductors are an attempt to eliminate that non-linearity. Whether or not it works or is audible, you're just gonna hafta find out yourself.
jn
Last edited:
No. That is what litz wire is, many insulated strands in a bundle. Standard litz has many more wires of smaller gauge, the count can go into the tens of thousands depending on what you want. The purpose of using it is to break up the proximity effects within the wires. I had some 5 mH coils made using 15 gauge magnet wire, and the proximity induced resistive rise climbed over 100 ohms before I reached 1 Khz. When I used litz with a few hundred strands equivalent to 15 gauge, the coils remain constant resistance out above 30 Khz. I could not get reliable readings above 30 Khz because of interwinding capacitance mucking up the measurements. In general, once the resistance has increased 2 orders of magnitude, you have to question the reliability of the readings; I typically stop when the readings go over 50 ohms or so.
The basic problem is that magnetics is not particularly easy to understand. When I learned about inductors back in the 70's, skin and proximity was taught, however, the tie in with non linear effects were never stressed. To this day, I find many misconceptions are still embraced. I had to learn a lot while working with superconductors, talk about skin depth..whoa.
jn
Last edited:
Nannoo, I'd suggest you just build it into your Xover and have a listen and make up your own mind if you think the few $s has been worth it, or not.
Sometimes, upgrading components makes errors &/or poor design, tolerances, etc more obvious and the sound 'goes backwards', where poorer quality components sometimes allows some 'masking' of the speakers and it can actually sound 'better'.
Sometimes, upgrading components makes errors &/or poor design, tolerances, etc more obvious and the sound 'goes backwards', where poorer quality components sometimes allows some 'masking' of the speakers and it can actually sound 'better'.
At 5 kHz, this inductor will have 7.85 Ohms. A usual driver has something 4 Ohms. See how I don't care about it?
If it is modulating then it is not time dependent. It is just a nonlinear effect, creating some 2nd order harmonic distortion. And then, again, how big is it?
If you are expecting it to be .18 ohms plus a pure inductance, then it will perform differently than expected. Since the data I mentioned was a very small inductor with a small wire, imagine a large inductor using 10 or 12 awg.
As I said, the resistance is modulating, and it is doing so as the absolute value of the slew rate. Given a sine excitation, the resistance will run as ABS(cos(t)), double frequency. An entity whose value is being modulated at some frequency is actually a time dependent thing. If the crossover is designed such that the woofer has decoupled and dropped off, no big deal as long as the second isn't getting to the midrange.
jn
Baseballbat, I assume from your above posted comments that this 'mumbo jumbo' stuff has no significance, or is of no practical interest to you, yes?
As 'jn said, nearly all inductor testing is done using sinewave signals - it's pretty much what most of us have available, except for the guys working in Universities, labs, etc, and is the industry standard.
Anyhow, the curious thing about inductors (and capacitors also, for that matter) used in the signal path is that there aren't any discrete sine waves in music at all so those very accurate Xover design tools only give theoretical sinewave laboratory results and are, at best, just an approximate starting point to the working filter circuit
If the programs could handle a signal of constantly varying complex transients mixed in with dc offsets, unequal signals, etc that is what is the musical signal, we might get a better Xover design off the computer - sadly, 'it ain't happening any time soon'!
With active electronic filters, some of the limitations of passive components can be reduced but it's still no guarantee of a 'better' final sound, despite the higher quality of the amplified signal - and this isn't really news to anyone, is it!
As 'jn said, nearly all inductor testing is done using sinewave signals - it's pretty much what most of us have available, except for the guys working in Universities, labs, etc, and is the industry standard.
Anyhow, the curious thing about inductors (and capacitors also, for that matter) used in the signal path is that there aren't any discrete sine waves in music at all so those very accurate Xover design tools only give theoretical sinewave laboratory results and are, at best, just an approximate starting point to the working filter circuit
If the programs could handle a signal of constantly varying complex transients mixed in with dc offsets, unequal signals, etc that is what is the musical signal, we might get a better Xover design off the computer - sadly, 'it ain't happening any time soon'!
With active electronic filters, some of the limitations of passive components can be reduced but it's still no guarantee of a 'better' final sound, despite the higher quality of the amplified signal - and this isn't really news to anyone, is it!
Actually, even at the labs we just use an agilent meter, which measures using sines. The meter separates the real and imaginary components of the inductor's response to a sine drive.
That will change eventually. Someday someone will write an article on this.
Historically speaking, that'll probably be two years off.
jn
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.