Al Klappenberger Posted January 31, 2006 Share Posted January 31, 2006 Guys, Several years ago I posted a comparison between several different popular types of inductors used in crossover networks. Here's a bit more detail. I drew "Q" curves showing the quality factor versus frequency of 4 inductors. Q is defined as the ratio of inductive reactance (2 * Pi * F * Henrys) over ESR. That is equivalent AC resistance, NOT DC resistance! That is: Q=X/R. The higher the Q the lower the losses. All the inductors are roughly 0.2 mHy or 200 uHy. The 4 coils are shown on the picture, left to right: Black curve: 217 uHy of #14 solid wire. Red curve: 202 uHy of #14 Litz wire (Solen Litz). Green curve: 194 uHy of #14 Foil. Blue curve: 207 uHy 2-layer #14 solid solenoid 1.7 Inch diameter. The best Q by far is the Foil inductor, BUT. It's best Q value is 30,000 Hz where is is of no use at all! The peak of the curve should be a close to the crossover frequency as possible. The Solen Litz has the best Q over the usable range of a tweeter and the best at the all-important crossover of 6000 Hz. The #14 solid wire inductors is to poor to mention! Al K. Plots next..... Quote Link to comment Share on other sites More sharing options...
Al Klappenberger Posted January 31, 2006 Author Share Posted January 31, 2006 Guys, Here's the plots.. Al K. Quote Link to comment Share on other sites More sharing options...
popbumper Posted January 31, 2006 Share Posted January 31, 2006 Al: NICE! This is the stuff I really wish I had more time for - trying out some different components to see which ones give the best response and accuracy in crossovers. Thanks for sharing, this is great stuff, especially with pics and graphs. Chris PS - anyone done this with capacitors? Quote Link to comment Share on other sites More sharing options...
Al Klappenberger Posted January 31, 2006 Author Share Posted January 31, 2006 BTW: One observation I forgot to mention. Solid wire inductors work best for woofer filters. Notice the solid wore coil is actuall best of the bunch down around 1000 Hz. Larger values will probably show the same thing. I'm to cheap to by a 2.5 mHy foild inductor to test to verify the theory though! I believe the DCR is the important factor down low.Al K. Quote Link to comment Share on other sites More sharing options...
jwc Posted January 31, 2006 Share Posted January 31, 2006 Al, that is very helpful. Question about the Solenoid? I don't know much about those. Is there a similarity between it and say the iron core inductors? Quote Link to comment Share on other sites More sharing options...
D-MAN Posted January 31, 2006 Share Posted January 31, 2006 Al, how can JUST wires have any effect on the signal?[] DM Quote Link to comment Share on other sites More sharing options...
sfogg Posted January 31, 2006 Share Posted January 31, 2006 " Al, how can JUST wires have any effect on the signal?" 'Cause it is a coil........ it isn't just wire anymore...... Weren't you the one claiming the type of wire in a coil makes no difference to the signal in rationalizing how a couple of inches of silver can have an effect when there are many many feet of copper in the coils of the crossovers and in the drivers? Shawn Quote Link to comment Share on other sites More sharing options...
sfogg Posted January 31, 2006 Share Posted January 31, 2006 Al, Have you modeled how the different Qs of the coils would effect the FR if they were substituted in the same crossover? Shawn Quote Link to comment Share on other sites More sharing options...
mark1101 Posted January 31, 2006 Share Posted January 31, 2006 Al, When you say "losses", do you mean audio signal loss through the network? Quote Link to comment Share on other sites More sharing options...
D-MAN Posted January 31, 2006 Share Posted January 31, 2006 Shawn, a coil is just wire. Make all the excuses you want, but anything that effects the performance of a coil - also directly effects a wire. They are, afterall, the same exact thing - one is just shorter than the other. If you want to pretend that short run wires have no electrical properties associated with them, fine. What Al K. posted above SHOULD be enough to indicate that there is a difference between Lietz type wire and solid wire and frequencies, diameter, structure, and metalurgy and their effects. DM Quote Link to comment Share on other sites More sharing options...
Al Klappenberger Posted January 31, 2006 Author Share Posted January 31, 2006 It looks like this is a interesting subject! Dana, I really don't understand why Litz wire has such a dramatic effect on the Q. I think it's because current is somehow caused to flow around a solid wire at higher frequency rather than down it's length. The Litz wire is made of 7 strands of finer wire each insulated from the rest. I suspect it brakes up this cross current the same way a laminated core in a transformer disrupts eddy currents. I wish I knew for sure though! Shawn, Low Q inductors will show up most dramatically right at the cutoff in a filter. That is, at the crossover frequency. That is why it is important to pick the inductor that has the best Q at 6000Hz in a tweeter filter. It's very difficult to do a computer analysis of a filters response including the actual curve of the Q. To do it requires a way for the computer to correlate the frequency point being analyzed with the same frequency point on a curve when the curve has to be a variable! All you can do is assume a constant Q and analyze over a small frequency range then change the assumed Q and analyze somewhere else! Another approach is assume some "curve fit" equation. I think Q=square root of frequency, or something like that is sometime used. What I do is assume the Q that I can expect at the crossover frequency and allow it to remain constant with frequency. It's close because the error in assumed Q has less effect as you move away from the "corner" of the filter up or down. Mark, Yes, "losses" are associated with the resistive component of the otherwise pure reactance provided by an inductor or capacitor. They turn your audio into heat instead of passing them on to your speaker! Al K. Quote Link to comment Share on other sites More sharing options...
Al Klappenberger Posted January 31, 2006 Author Share Posted January 31, 2006 Guys, Here's a sample run of a program I wrote years ago. It calculates the length of straight wire needed for a given inductance. The program was used to estimate the wire length for "hairpin" inductors used at frequencys around 1000 MegaHertz where a true coil would have too much inductance. I would generally go to hairpin inductor for anything below about 10 NanoHenrys. This says that 3.52 Inch of #24 wire is needed get 0.1 uHy inductance at 100 MHz. (That's MicroHenry at 100 MegaHertz). AL K. Quote Link to comment Share on other sites More sharing options...
sfogg Posted January 31, 2006 Share Posted January 31, 2006 "a coil is just wire. Make all the excuses you want, but anything that effects the performance of a coil - also directly effects a wire. They are, afterall, the same exact thing - one is just shorter than the other." I have a $100 that says if a coil is uncoiled it will measure differently straight then when it was coiled up. According to you they should measure the same since it is the 'same exact thing.' Wanna take the bet? Shawn Quote Link to comment Share on other sites More sharing options...
3dzapper Posted January 31, 2006 Share Posted January 31, 2006 http://www.allaboutcircuits.com/vol_2/chpt_3/5.html Quote Link to comment Share on other sites More sharing options...
Al Klappenberger Posted January 31, 2006 Author Share Posted January 31, 2006 Guys, BTW: If you have a given length of straight wire it will measure a much LOWER inductance than if you wound it into a coil. The total inductance is equal to L1 + L2 + 2*Lm where Lm is the mutual inductance between L1 and L2. Look at the big solonoid coil on the far right in my first picture. It is two layers EACH layer has 1/4 the inductance of the total. .05 mHy + .05 mHy + 2 * .05 mHy (the winding are on top each other) = 0.2 mHy. Just for kicks I used the straight wire program to figure how long a #14 wire would have to be to yield .2 mHy at 6000 Hz. The answer: 284 FEET! Al K. Quote Link to comment Share on other sites More sharing options...
D-MAN Posted January 31, 2006 Share Posted January 31, 2006 Al, one thing that I was taught in the Navy was that current flows on the surface of all conductors. The more surface area the better, and the multi-strand wiring (Lietz) has more available surface area for its overall diameter than a solid conductor of the same diameter as the combined Leitz wires. This tends to increase some resistance, though, as we all know what happens when current flow exceeds the ability to dissapate the resultant heat caused by resistance. Larger diameter multi-wind solid core wires tend to be used for high current applications, like a car battery cable or power transmission lines. Frequency is not a particular concern in those applications. Concerns about power handling followed by flexability tends to dictate what type of wire is appropriate where. Coils are not subject to flexability concerns, of course. In WW2 the Germans had a problem using their radar which relied upon cables, where the British radar used hard-structure wave guides (no, not horns!). The Germans could not get microwave tranmission over long cables for that reason because of the losses involved, where the British could. Wave guides are structured as hollow square or rectangular conduits with silver coating on the inside (the conductor) for high frequencies, especially microwaves. The size (cross-section) of the waveguide is dependent on the frequency being transmitted along its path. Silver was discovered to be the best conductor inside the waveguide as a sprayed-in coating. Sort of interesting because all of the physics involved combine some of the elements of horn technology with wires. The concern is exactly the same - getting the signal from here to there with as little loss as possible. DM Quote Link to comment Share on other sites More sharing options...
efzauner Posted January 31, 2006 Share Posted January 31, 2006 Al, All the inductors have the same inductance, just different series resistance, correct? Have you simulated cross overs with inductors with different series resistance? What if you make the ESR a function of frequency? Do you have data of ESR vs Frequency of the various inductors? That way you can compare the Cross over response with different inductors and see what it does to frequency response, phase, group delay etc. Or just measure the crossovers built with the different inductors? The high series resistance of the inductors at the cross over frequencies would surely shift those crossover frequencies just like different speaker impedances. Another dumb questions: When you design a cross over for a give frequency, you use the nominal impedance of the driver, say 4, or 8 ohms, whatever. Would it not be better to use the real complex impedance of the driver at the cross over frequency? Quote Link to comment Share on other sites More sharing options...
mark1101 Posted January 31, 2006 Share Posted January 31, 2006 Very interesting......OK, so which inductor actually sounds best? Did someone do that test? Quote Link to comment Share on other sites More sharing options...
Al Klappenberger Posted January 31, 2006 Author Share Posted January 31, 2006 Dana, There's a problem with the "skin effect" theory. At audio the current is distributed virtually evenly through the wire. Current does not start comming to the surface until WAY above audio frequency. Litz wire does help with skin effect but the distributed capacity between the strands starts to screw up the advante as you go higher. So that's not what's casuing the improved Q at audio frequency. There is definitly something else going on! Silver plating depends on skin effect. At audio the wires would need to be soild silver! A total wast of money! Waveguides are a form of transmission line. The theory of those is out of my area of experitce. I have lots of books on the subject here, I just never got directly into it. I was always an L-C filter guy! Al K. Quote Link to comment Share on other sites More sharing options...
D-MAN Posted January 31, 2006 Share Posted January 31, 2006 Guys, BTW: If you have a given length of straight wire it will measure a much LOWER inductance than if you wound it into a coil. The total inductance is equal to L1 + L2 + 2*Lm where Lm is the mutual inductance between L1 and L2. Look at the big solonoid coil on the far right in my first picture. It is two layers EACH layer has 1/4 the inductance of the total. .05 mHy + .05 mHy + 2 * .05 mHy (the winding are on top each other) = 0.2 mHy. Just for kicks I used the straight wire program to figure how long a #14 wire would have to be to yield .2 mHy at 6000 Hz. The answer: 284 FEET! Al K. However, would you agree that an arbitrary 16ft chunk would have a measurable amount of inductance (however small) and resistance (and capacitance)? It stands to reason that it would, and therefore, it is not simply a conductor with no electrical properties. Now what do these properties (however small) do to the signal being passed? Seems to me that for anyone to simply discount or ignore the properties of wires is discounting WAY too much. DM Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.