The right Litz wire can make a slight difference at audio frequencies. Agree, inductors are the worst, but sometimes they're still the right choice. Remember that when used in a frequency selective network, the signal amplitude may be greatly reduced and the inductor performance over a wide range may not matter, so long as it doesn't misbehave so badly as to cause the response shape to change.
4 is not so bad, but I assume since both are resting on a planar surface, symmetry of orthogonality is broken. I'd avoid it also.
Jn
Last edited:
An article on skin effect is of little use with respect to a discussion of proximity effect when discussing multi turn inductors.
Jn
Ps. Look at the graphs in my gallery. R climbs and L drops well within the audio band, measured results. And those were small inductors light gauge.
It takes far more care to measure skin effect on a single conductor not in proximity to the return...you are trying to see a drop from 15 nH per foot buried in a loop inductance of tens of microhenries, and doing this in the hundreds of kilohtz.
It would be easier to measure the skin effect within a coax, as the inductance is the sum of the internal L of the core plus the flux in the dielectric. Unfortunately, the shield is also skinning to it's inner surface so R can become problematic, too many terms..
Last edited:
They look solid enough. Honestly, I would not make a distinction between potting materials, especially if the wax is hard enough. I assume those use hard wax, otherwise they'd fall apart.
Jn
In terms of losses and L shifts, I would expect litz to be more stable.
In terms of sound quality, that is a unknown because the xover is dialed to the drivers.
I recommend substituting to litz in one crossover, then test the soundstage using one amp channel, then two channels with a mono source.
If there is no significant difference, all of the image should remain in the center and stable. If it makes a difference, you will hear some of the music content shift right or left of the rest of the content.
Remember, this test is not a head in vice test, you are using non shifted program material as a spacial reference to look for shift of part of the program.
If you hear no difference or smearing, then the inductor choice made no difference. If you detect side smearing, then you have to figure out which inductor you prefer with your speakers.
Jn
Ps. There is obviously some kind of paper involved as well on those foil inductors, does anybody know what type?
In terms of sound quality, that is a unknown because the xover is dialed to the drivers.
I recommend substituting to litz in one crossover, then test the soundstage using one amp channel, then two channels with a mono source.
If there is no significant difference, all of the image should remain in the center and stable. If it makes a difference, you will hear some of the music content shift right or left of the rest of the content.
Remember, this test is not a head in vice test, you are using non shifted program material as a spacial reference to look for shift of part of the program.
If you hear no difference or smearing, then the inductor choice made no difference. If you detect side smearing, then you have to figure out which inductor you prefer with your speakers.
Jn
Ps. There is obviously some kind of paper involved as well on those foil inductors, does anybody know what type?
Last edited:
Thanks.
So the insulation is "special paper". Well, that cleared up a lot.
"Power distortion" is what they call what skin effect is. Ok.
They've no clue what proximity effect does to current distribution, they assume it remains uniform across the surface. Boy, they need an analysis package..
Oh, they use a polyurethane coat and then a polyamide bonding. Keep in mind, polyamide is nylon, polyimide is kapton. Major difference only if you want to go to 300C or 1.8 Kelvin. Room temp, no big deal.
Jn
In the sense that they are the least like an ideal component, yes. Resistors and capacitors are mostly much more ideal in their behaviour, at least at audio frequencies.merlin el mago said:
I agree that the resistance will rise with higher frequency. I disagree that there will be a non-linear component; you cannot make nonlinearity out of linear equations (Maxwell, Ohm). However, I believe we have had this argument before and neither of us managed to convince the other so I am happy to leave it at that: a statement of difference.jneutron said:
Take a power inductor of .1 ohm, put a sine current in of sufficient speed that proximity effect at zero crossing increases it to .2 ohms. At zero going positive, the R is .2 ohms, at 90 degrees R is .1, as it goes through zero negative R is again .2 ohms. At 270, it is .1. The resistance will be a function of the current slew, varying from .1 ohms to .2 ohms, .1 at max current, .2 at zero current, varying as some function of the sine.
Multiply the resistance by the current through 360 degrees.
You do agree that proximity effect is a direct result of current slew, right? As the slew is what causes the time varying magnetic field that causes eddy currents and consequent flux exclusion.
Jn
Edit: one confounder that has to be worked around is the inductive changes caused by slew rate. I think I have an experimental design to cancel that out in measurement. Unfortunately, the last 4 years, this stuff was really back burner..
Last edited:
Sorry to have stirred things up - I really wanted to know if people had preferences between wax and polypropylene foil inductors (also now waxed Litz as that's been thrown in) - in terms of what they hear ..but I feel like if been injected into the magnetic flux time vortex ;p
It's interesting stuff, and proves there's a lot to it and to learn, but unfortunately most of it is going over my head right now.
It's interesting stuff, and proves there's a lot to it and to learn, but unfortunately most of it is going over my head right now.
Most of it, even though fascinating, is not in any way related to the way inductors sound.
Sorry, i also cannot add anything interesting. A few years ago my crossovers got updated with Alphacore copper foils and the improvement was substantial, if not exactly staggering. A friend later on bought samples of each and every "audiophile" choke and pronounced the Duelund king of the chokes. He also really badmouthed my Alphacores
No, not exactly. Proximity effect (and the related skin effect) is caused indirectly by current skew. Essentially what you have to do for a non-sinusoidal waveform is calculate the effect separately for each frequency component - the effect depends purely on frequency, geometry and material properties. Then add up what you find. This whole process is linear, so the final outcome must be linear. No distortion, no new frequency components created. Your sine example varying from 0.1 to 0.2 ohms through the cycle is simply wrong - unless of course all the EM textbooks are wrong. I admit this is possible, but I believe it to be unlikely.jneutron said:
You can get nonlinearity creeping in if there is significant heating (which varies with current^2 - so nonlinear) or changes in geometry (e.g. due to magnetic fields - the force between turns will go like current^2). The simple ideal case of fixed geometry and linear materials cannot generate nonlinearity, and this is close enough to reality in most audio situations.
I don't want to get drawn into a long argument on this as we have been over this before and all we end up doing is restating our respective cases. This is unprofitable for us and boring for others.
Indirectly by current slew? (I assume you meant to say slew, my I pad continuously changes words on me).
Oddly enough, when we charge an inductor at 83 amps per second for 80 seconds, we can easily measure the changed resistance from proximity effect. Straight forward, easy to do. The resistance bumps to a new, constant value during the entire ramp. When we flat-top, the resistance drops to it's DC resistance.
When we ramp down, the resistance increases during the ramp down, and remains at the higher value until zero.
The proximity based resistance increase is a direct function of the absolute value of the rate of change of the magnetic field. If you are calling the absolute value function "linear", then we are in agreement.
Jn
Ps funny thing you mention e/m textbooks. One day I went to a friend down the hall to borrow one of his texts, Jackson (an obscure unknown author
), he said he didn't have any.He is one of the worlds foremost experts in magnets and e/m theory, (I always discuss technical issues with him, this exact conversation being one of them), which begged the question why no books?
His answer: I start with maxwells equations, and derive everything I need..
Again, why?
Answer... That way I know it's right.
The discussion we are having? He understood.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.