The Loudspeaker Crossover Part II: Measurements & Conclusions
The screen-shot below, shows the Clio input shorted to its output so we have a reference of the signal used for the magnetic vs air core choke test.
The noise floor of the Clio is below the -90db level at the bottom of the top screen
The screen-shot above is a dual view, with the top screen a spectrum analysis, amplitude vs frequency; and the bottom screen an oscillogram, or the voltage vs time view. Both views are two ways of examining the exact same data. If our choke and amplifier are perfect, the measurements will look exactly like the view above. Since our 4 ohm resistor is able to dissipate 900 watts RMS, and the bridged amplifier is rated to deliver 3000 watts into this 4 ohm load, we will limit the power into the load bank to 900 watts. To further limit the resistor heating, we will use a sine wave which is on for one second and off for two. With this arrangement we can take a stable measurement while reducing the long term (3-second avg.) power to only 1/3rd of the signal-on condition.
Now if my theory is correct, we do not go from "the choke is perfect", to "the choke is saturated" like a light gets switched on or off. If the magnetic core saturates, this effect should happen gradually as the alignable magnetic domains available become more rare, as the magnetic material runs out of them (this is saturation). We should be able to see this effect by viewing either the time or spectral data. When we examine the sine wave above, it reminds us that the AC signals are not constant, but vary with time. That 115 volts we plug our computers and televisions into goes from zero to 162 volts, to zero to minus 162 volts 60 times every second. If we attempt to drive the 4 ohm load with a 900 watts RMS sine, we will require a current of 15 amps RMS and 21.2 amps peak! Although the manufacturer does not list this parts saturation point on the data sheet, we should be able to find it for ourselves. Unlike a transistor amp without compression run into clipping, the shape of a magnetic cores hysteresis curve limits gradually as it approaches it saturation point, so there is no exact point of onset of saturation.
At the frequency of 120 Hz, the choke calculates out as a little more than 4 ohms. Since our load is also 4 ohms, and since music tends to have a lot of power occurring in this frequency range, we shall use 120 Hz as the test frequency. Under this condition, there will be approximately as much voltage across the inductance as the resistor. Choosing a much lower frequency will tend to take the choke out of the equation, and choosing a much higher frequency would seriously limit the amount of current flowing through the circuit or put very high demands on the voltage required of the amplifier. The signal is raised until the voltage across the 4 ohm resistor is just above 60 volts (61.6 rms), and the power absorbed by the load is 900 watts. This conditions requires about 95 volts from the amplifier. (In the perfect world, 87 V would do, but nothing is perfect).
The signal is measured and shown below:
High level of 3rd Harmonic Distortion due to the Magnetic core of the 5.6 mH inductor
With the Clio Meter maximized, the THD measures 6.9% and IMD measures 1.2%. As can be seen from the measurement above, the distortion is symmetrical, as expected, so the dominant distortion is 3rd harmonic, with 2nd harmonic low. The waveform is starting to take on the form of a triangle wave as a result of the changing impedance of the magnetic core choke.
Well, perhaps this distortion is a result of the amplifier being driven so high? Let's test that. Since I do not have an air-core choke this large, I construct one out of two 2.4 mH cokes, stacked one on top of the other. It ends up being 5.7 mH. (The reason two 2.4 mH chokes add up to more than 4.8 mH, is due to mutual, not self inductance). This is why you do not have a linear relationship between turns and inductance in a choke. (And why there are (6) different equations to solve for inductance on the Wikipedia page referred to earlier). If both of these chokes were wired in series and placed far apart, instead of stacked one on the other, they would in fact be 4.8 mH. Now, at 1.7 ohms, the DCR of our air core choke is considerably higher than the magnetic core choke which measures only 0.28 ohms. This increased resistance requires us to deliver 115 volts to the series circuit to achieve the same 60 volts (62 V) across the 4 ohm test resistor.
The measured response is below:
The distortion Products from the combination of Amplifier and 5.7 mH Air Core Choke
The distortion measured 0.952% THD. As can be seen above, the spectrum across the resistor looks quite different when driven through the air-core choke than it does when driven through the magnetic core choke. The oscillogram too looks free of distortion as well. So, let's ask ourselves, is this better because the DC resistance is higher by 1.5 ohms? Maybe the issue was current limitation in the amplifier. This amp is rated to deliver 3000 watts into 4 ohms, which requires 27.3 amps RMS. The 62 volts across 4 ohms only requires 15.5 amps RMS, just a little more than half what the amplifier should have available. Maybe the 45 degree phase angle between voltage and current at 120 Hz is straining the amp? What is the limit of this amplifier? Let's see. Without a series choke, with the 4 ohm load alone attached to the amplifier, I drive the amp to clipping, and measure 88 volts RMS across the resistor terminals.
The following measurement data was taken.
Amplifier in Clipping
As can be seen here, the sine wave is clipped, as the top of the waveform is flattened. Since we were able to deliver 115 volts unclipped to the higher load impedance, we have to assume this amp has run out of current. At 88 volts RMS, a 4 ohm load will draw 22 amps RMS. Close inspection of the amplifier specification claims 3000 watts @ 1% THD when bridged into 4 ohms but only at 1 kHz. Clearly at 120 Hz, this amplifier has clipped at 1936 watts. As had been said before, trust but verify... So, lets drop the input by 1db, and measure the signal across the resistor again.
Amplifier 1 db below Clipping - No Choke
This time we manage to get 80.6 volts across the resistor, with very low distortion. Clearly with a zero phase angle, this amp can deliver 20 amps RMS and 28.3 amps peak without clipping into 4 ohms. We can be pretty confident that the difference we have measured is the result of the magnetic core of the choke, and not the amplifier or signal generator. Suppose we drop the input drive to the magnetic choke to 6 db less than the 900 watts RMS delivered to the test load. Lets say we drop the voltage down to 28.5 volts RMS, or 203 watts RMS. What kind of distortion will we measure? Lets see..
6 db below clipping with Magnetic Core Choke
Even at 6 db below clipping, the Magnetic core choke results in more than 1% THD
While the waveform looks pretty clean, we see the same distinctive line up of harmonics, with 3rd being very dominant. When metered, the distortion at 25.5 volts RMS measures 1.3%. This is still higher than 0.952% THD measured with 62 volts across the resistor, and 115 volts from the amplifier using the air core choke. At this point we are well below the limitations of the amplifier, and delivering a current of 7.125 amps RMS, less than half as much current (1/4 the power) as was delivered through the air core choke and yet despite our reducing the power to 203 watts RMS, we find we have more THD in our resistor than what was present at 961 watts when driven through the air core choke.
The use of magnetic cores in chokes is a two edged sword. While magnetic materials in chokes will reduce size, cost, and the DCR of the choke, it comes with a disadvantage in terms of distortion. While amplifier manufacturers are busy reducing their distortions to numbers well below 0.01%, loudspeaker manufacturers still struggling to build systems which can play at or below 1% distortion anywhere at or near their output limits. To add an additional 1% distortion by using magnetic core inductors in series with the speaker is a price I am personally not willing to pay.
So, what conclusions are to be drawn from this article?
- Temperature is an enemy of all three components, and temperature has real and measurable effects upon the transfer function of the crossover network (and your music!)
- Using Nichrome as a resistor material provides a huge advantage over copper or aluminum by virtue of its relative constancy with temperature fluctuation
- Electrolytic Capacitors suffer hysteresis effects which can negatively impact audio
- Bypass caps can be used to improve performance of Electrolytics by lowering ESL
- Electrolytic Capacitors suffer from high dissipation, which cannot be minimized without using multiple capacitors in parallel
- PCB layout is an important consideration and effects the overall crossover power handling and performance, and is especially important in terms of heat and inductive coupling
- Saturation is not a digital phenomenon - It happens gradually and the magnetic core choke increases distortions at levels considerably below those at which the cores will "Saturate"
- Magnetic Cores can reduce crossover cost, DCR and size by increasing inductance
- Magnetic core Inductors are inherently nonlinear, even before saturating
• Air cores chokes are preferable, but suffer from high losses and high cost
While this is not an exhaustive study of chokes, or of every type of magnetic core material, we have seen how the use of a non-linear material can affect the sound of the music we hear. If we are looking to improve our systems, and purchasing upgrade crossovers, we need not only be informed about the specifications of the parts used in them, we also need to consider the PCB layout as well. While perhaps the simplest electrical circuits in use in audio today, crossovers can play a major role in the sound we hear coming from our speakers, and our speakers are the most important link in the audio chain in the vast majority of high performance home theater systems. The sum of the parts and their quality truly does matter, especially if you're a golden ear and desire the very best performance.
Paul Apollonio is a Southern California based consulting engineer, with 25 years of experience in the audio industry designing loudspeakers and crossovers. He may be reached at papollonio@yahoo.com and his resume and profile viewed online at http://www.linkedin.com/in/paulapollonio.
I wasn't able to post my series links because my post count is only one since I just joined. Too bad. I've got better things to do than be malicious. Let me know and I'll email them to you if you'd like.
Quite frankly with my latest project the crossover sounded great first try. After experimenting with parallel designs for decades, I find the series much easier to implement and test, with only four components for a three-way and two for a two-way. I've even mixed Zeta .5 and Zeta .7 with good results. I personally shy away from Zeta 1 or lower for the same reason as a parallel slow slope, too much overlap, and most drivers don't have that wide a response to handle it. All my testing is by ear with difficult program material, voice, piano, instrumental combos and percussion, mainly classical and jazz. I just forge ahead right or wrong. It all started when I inherited a pair of Dahlquist DQ-10s. One look at that crossover schematic convinced me that almost anything goes, and they sound great. It works for me, and after all, serendipity rules. Shed those prejudices! Hope this is all helpful.
highfigh;723200
Do you have any links to info on these Zeta crossovers?
Here is a circuit of a series crossover.
Series crossovers [speakerbuilder.net] present a lot of formidable problems, as both sections of the crossover interact.
My rear backs, use series crossover in the passive part of the crossover. These speakers started in 1984, but the crossover did not get to its final form until 1994!
They are very difficult to perfect.
This was a good article, and reiterates points I have made many times over. The sad fact is most commercial designs do use miserable chokes with iron cores and wire of too thin a gauge, and electrolytic caps abound. A speaker is then severely compromised right out of the stating gate then, and I mean severely compromised. A speaker with decent components in the crossover will not be cheap. This leaves money on the table, for active speakers, which I believe with modern production methods, could give much better performance per dollar.
I really believe it is the receiver obsession that is so limiting. If you think about it, it is absurd that a pre pro generally cost more than a receiver. The only reason is production numbers. Burying the receiver is long, long over due now.
audioantique;723194
Paul,
Very informative and straight to the point, thanks. I suggest that a way to bypass the inductor series losses is to build a series crossover, where there are no components in series with the drivers. One exception to this would be a padding network for a tweeter, but that is only series and parallel resistors. I realize that this solution isn't for everyone, but I've had great success lately with Zeta .7 quasi-second order crossovers. A 3-way is only two caps and two chokes all parallel to the drivers, and the, for me, essential Zobel for the midrange is also parallel to the driver. Of course just like with a parallel crossover, it's essential to use drivers in their usable range and not push their envelopes. I'm not a series crossover zealot, but extensive listening seems to validate my efforts, at least to my ears.
Good luck, all.
Do you have any links to info on these Zeta crossovers?
Very informative and straight to the point, thanks. I suggest that a way to bypass the inductor series losses is to build a series crossover, where there are no components in series with the drivers. One exception to this would be a padding network for a tweeter, but that is only series and parallel resistors. I realize that this solution isn't for everyone, but I've had great success lately with Zeta .7 quasi-second order crossovers. A 3-way is only two caps and two chokes all parallel to the drivers, and the, for me, essential Zobel for the midrange is also parallel to the driver. Of course just like with a parallel crossover, it's essential to use drivers in their usable range and not push their envelopes. I'm not a series crossover zealot, but extensive listening seems to validate my efforts, at least to my ears.
Good luck, all.