Choke coils for loudspeakers
If you want to keep high tones or frequencies away from mid-range and woofers because they should no longer transmit them, then a coil is switched into their signal path. Such choke coils consist of tightly wound copper wire coated with insulating varnish and reduce the alternating current that passes through depending on its frequency. The increasing resistance at high frequencies is caused by a magnetic field that every electrical conductor builds up around itself when current flows through it. When the current flow changes, the magnetic field induces an opposing voltage in the conductor, such as the coil wire. If alternating current flows through it, its flow changes in time with the frequency, and as the frequency increases, the induced counter-voltage increases and with it the resistance of the conductor or coil wire.
This property, known as inductance, increases in coils with the number of windings. The unit of measurement for inductance, named after an American physicist, is Henry (H), but for loudspeakers only thousandths of this is required and is therefore referred to as milli Henry (mH). In order to limit the application range of mid-range or woofers to higher frequencies, the required inductance is calculated using the following equation:
mH
The L stands for the required coil inductance, Z for the impedance of the loudspeaker in ohms, Pi for the circle number 3.14 and fc for the desired crossover frequency in Hertz, at which the level attenuation is already 3 dB.
Example: A woofer with 8 Ohm impedance or nominal resistance, which is only to be used up to 300 Hz, therefore requires a coil with approx. 4.3 milli Henry.
Typical effect of a choke coil on the level (in decibels) and the AC resistance (in ohms) as a function of the frequency of the AC current (in Hertz).
However, their inductance is only one of the effects of choke coils, and in order to find the optimum coil for the respective purpose, it helps to know the properties that can affect the load capacity, sound and price of the coils.
Coil impedance
The total impedance of a coil is mainly made up of the DC and AC resistance, with the latter increasing with increasing frequency and coil inductance. The alternating current or
reactance of a coil is not a fixed and easily calculable variable due to phase shifts between current and voltage. Rather, it is determined by all resistances in the measuring chain, i.e. by the voice coils of the loudspeaker chassis, by fixed resistances in the crossover and by the loss resistances of the cable connections between the amplifier and loudspeakers.
Direct current resistance
The ohmic and non-inductive resistance of the long copper wire of a choke coil reduces the current conducted to the loudspeaker regardless of the frequency and leads to a loss of power. If the DC resistance is 20% of the loudspeaker impedance, a third of the amplifier power supplied is literally burned up in the wire of the coil. Out of 100 watts, only around 65 watts remain, which is clearly audible. To remain inaudible, the DC resistance of a coil should be no more than 10% of the speaker impedance, i.e. 0.8 Ohm when using 8 Ohm speakers. In addition, however, there are the losses of the meter-long cable between the amplifier and loudspeaker as well as the usually unnoticed transition resistances at the terminals of the amplifier outputs, the oxidized cable ends or plugs and the terminals of the loudspeakers.
If you really want to minimize losses, you should clean the cable connections and contacts from time to time and only allow a DC resistance of perhaps 0.4 ohms in the choke coils in front of the woofers. But how do you achieve this?
Thicker wire reduces the DC resistance of a coil. Using thicknesses between 1.6 and 0.8 mm, diagram (2) illustrates the relationship using the example of a choke coil with 1.0 mH inductance. The lower curve already indicates that so-called core coils have far lower loss resistances.
Large inductances require more coil windings and a longer wire than small ones, because to double the inductance of the same coil type, the number of windings must be increased by the square root of 2, i.e. 1.41 times. Conventional air-core coils with 2 mm thick copper wire only undercut the 0.4 ohms targeted here up to 4 mH inductance and are still quite large, heavy and expensive.
Core coils, on the other hand, manage with a shorter wire length and therefore have lower losses than air-core coils, as diagram (3) shows.
The special features of core coils
Magnetizable material inside or around the coil increases its inductance many times over, which is why the wire can be shorter than that of an air-core coil to achieve the same inductance. There are basically three different types of core:
Ferrite cores (ferrobar) made of magnetizable but electrically non-conductive ceramic,
Powder cores made of iron dust, which are permanently held in shape by binding agents,
Corobar coresmade from a crystalline powder mixture
Transformer cores made of laminated sheet metal strips, which consist mainly of iron. Magnetically, transformer cores are much more effective than those made of ferrite, but they are also heavier and 
more expensive.
One weakness of core coils lies in their so-called remanence, a kind of memory effect of magnetizable materials: If the copper wire coil through which the current flows has magnetized the coil core, the latter retains its magnetization to some extent even after the excitation has ceased. This causes non-linear distortions, especially in the form of K3, which means that core coils 
add a percentage of three times the frequency to the
alternating current signals flowing through them. Although loudspeakers also produce K3, especially in the bass range, which can climb above 5% at high levels and very low frequencies, it rarely exceeds 1% in the upper bass range. Choke coils should not add further distortion to this if possible. At amplifier power levels 
of less than 50 watts, typical ferrite core coils are modest with hardly more than 0.1 % fundamental distortion, transformer core coils usually with around 0.3 % and powder core coils lie somewhere in between. Such values are not alarming. Only transformer core coils with so-called EI cores trimmed to low loss (zero ohms) produce an unacceptable K3 of up to 1% even at the lowest power levels. One percent K3 can already be audible and represents the limit of reasonableness - coil distortion should never rise higher than this, but it will if currents are too high. The limited ability of core coils to process arbitrarily high currents without complaint is related to the phenomenon of saturation.
Core saturation
As the current flow increases, the magnetic field generated around the coils intensifies. However, core materials only allow a maximum magnetic field strength until they are saturated and can no longer keep up. Iron alloys can withstand field strengths five to ten times higher than ferrites. Beyond the saturation limit, the distortions of many core coils skyrocket.
Saturation and impedance
Speakers with an impedance of 4 ohms draw twice as much current from the amplifier as 8 ohm speakers with the same volume control setting. However, choke coils in front of woofers with 4 ohms saturate at half the power. Of course, this also applies to two bass speakers connected in parallel, each with 8 ohms. Depending on the crossover design, the impedance can fall far below 8 ohms even with only one 8 ohm woofer in the bass and require a more powerful coil. On the other hand, with half the speaker impedance, only half the inductance is required to achieve the same attenuation of higher frequencies, for example 2.2 to 2.7 mH instead of 4.7 mH. The saturation limit of the coils as the maximum permissible amplifier power in watts for 1% K3 ultimately remains the same if the impedance and inductance are halved or doubled together. As a reminder: 1% K3 means that one percent of three times the frequency is added to the original signal, which may well be audible.
Saturation and frequency
Mechanical coil resonances
The wound coil wire, but also cores made of ferrite or iron, can begin to vibrate mechanically and lead to coil resonances, usually at medium to high frequencies. Gluing, bonding or impregnating in synthetic resin can significantly reduce unwanted vibrations in the windings.
Microphony
The pressure waves generated by the loudspeaker in the enclosure cause the crossover and its components to vibrate, which can have a negative impact on signal processing and sound. A sub-housing for the crossover, i.e. a separate small housing in the actual speaker cabinet, as well as gluing, bonding and screwing together the components and the circuit board can help to counteract this.
Eddy currents
In electrically conductive materials, such as iron, magnetic fields generate electrical currents, so-called eddy currents. Unfortunately, these currents influence the magnetic properties of the material. In coils, this would lead to losses and distortions. To counteract this, transformer core coils are made by layering thin strips of sheet metal that are electrically insulated from each other. With powder cores (Corobar, Ferrobar), on the other hand, the binder insulates the iron particles electrically from each other and reduces the formation of eddy currents. Ferrite cores (Ferrobar), on the other hand, are electrically non-conductive and do not produce eddy current distortions.
How many watts can they withstand?
To find out how much power you can expect from the various core coils, you can measure their saturation limit for 1% K3. To do this, we used various versions with 4.7 mH, a typical inductance for bass speakers in three-way loudspeakers with 8 Ohm impedance. Diagram 8 illustrates the result.
The left-hand scale shows the sound pressure in decibels to be expected from a hi-fi speaker at a distance of one meter - at the usual listening distance of three to four meters, the sound pressure in the living room is of course much lower.
The scale on the right shows the calculated saturation power in watts at 8 ohms and the coil type.
Coils whose designation begins with HQ, DR or P have ferrobar cores of different types, CO stands for a corrobar powder core coil, TO for a torobar with toroidal core and the other coils with the highest load capacity use cores made of transformer sheet.
For those who want to know more, Table 1 shows the voltage, current and amplifier power at 8 ohms that the core coils in Diagram 8 can handle with little distortion at loud impulses.
Table 1 Saturation limits of core coils with 4.7 mH
|
Coil type |
Rdc Coil |
Z at 100 Hz (1) |
Voltage for 1% K3 (1) |
Current for 1% K3 (2) |
Pmax on 8 Ω (2) |
|
Ferrite core coils |
|||||
|
HQS 32/26 |
2,31 Ω |
10,83 Ω |
21,5 V |
2,0 A |
32 watts |
|
HQG 52/36 |
0,40 Ω |
8,99 Ω |
29,1 V |
3,3 A |
87 watts |
|
HQ 43/45 |
0,47 Ω |
9,02 Ω |
33,6 V |
3,7 A |
110 watts |
|
HQ 58/46 |
0,19 Ω |
8,81 Ω |
37,5 V |
4,3 A |
148 watts |
|
DR 56/35 |
0,33 Ω |
8,86 Ω |
43,0 V |
4,9 A |
192 watts |
|
HQP 56/35 |
0,47 Ω |
9,02 Ω |
50,6 V |
5,6 A |
251 watts |
|
HQP 62/47 |
0,32 Ω |
9,08 Ω |
50,5 V |
5,6 A |
251 watts |
|
DR 56/61 |
0,21 Ω |
8,97 Ω |
52,5 V |
5,9 A |
278 Watt |
|
Corrobar powder core coils |
|||||
|
COT 92/39 |
0,55 Ω |
9,31 Ω |
139,4 V |
15,0 A |
1,800 watts |
|
COT 92/39 |
0,52 Ω |
9,10 Ω |
139,4 V |
15,3 A |
1,877 watts |
|
Torrobar toroidal coil |
|||||
|
T 010 |
0,20 Ω |
8,72 Ω |
94,4 V |
10,8 A |
933 watts |
|
Transformer sheet coils |
|||||
|
I 78 |
0,76 Ω |
9,52 Ω |
74,4 V |
7,8 A |
487 watts |
|
I 96 |
0,58 Ω |
9,08 Ω |
96,4 V |
10,6 A |
899 watts |
|
I 130 |
0,26 Ω |
9,02 Ω |
125,8 V |
13,9 A |
1,546 watts |
|
I 150 |
0,20 Ω |
8,96 Ω |
>141,4 V |
>15,8 A |
>2,000 watts |
|
FE 96 |
0,18 Ω |
8,82 Ω |
141,4 V |
16,0 A |
>2,000 watts |
|
FE 130 |
0,08 Ω |
8,81 Ω |
>141,4 V |
>16,0 A |
>2,000 watts |
(1) Measured values including 8 Ω load resistance
(2) Values calculated from measured data
Even inexpensive ferrite core coils with up to 4.7 mH seem to be able to withstand a lot of power, but it should be noted that it is not the continuous power but the pulse power of the connected amplifier that should be taken into account here. Amplifiers can deliver around twice their continuous or rated power for short periods. Bass loudspeakers can handle such power peaks, and suitable chokes should also be able to do so.
Procedure :
To determine the saturation limits, each coil was connected in series to a resistor of 8 ohms and then connected to the amplifier output. The input of the amplifier received a sinusoidal signal at 100Hz and the output voltage was increased until the K3 was exactly 1.0%. The complex impedance of the coil and fixed resistor was then measured at 100 Hz. The saturation current and the corresponding amplifier power at 8 ohms were calculated from the measured voltage and impedance. For 4 Ohm loudspeakers, however, this power must be halved.
The following Table 2 helps to roughly transfer the power limit of the coils listed in Diagram 8 with 4.7 mH to other inductors of the same type.
| |
|
|
| |
|
|
| Table 2: Saturation and inductance of the HQP 56/35 | |
|
| |
|
|
| 1.0 mH | 4.70 x Pmax1 | 1,175 watts |
| 1.5 mH | 3.13 x Pmax1 | 782 watts |
| 2.2 mH | 2.14 x Pmax1 | 535 watts |
| 2.7 mH | 1.74 x Pmax1 | 435 watts |
| 3.3 mH | 1.42 x Pmax1 | 355 watts |
| 3.9 mH | 1.21 x Pmax1 | 302 watts |
| 4.7 mH | ---- x Pmax1 |
250 watts |
| 5.6 mH | 0.84 x Pmax1 | 210 watts |
| 6.8 mH | 0.69 x Pmax1 | 172 watts |
| 8.2 mH | 0.57 x Pmax1 | 142 watts |
| 10.0 mH | 0.47 x Pmax1 | 117 watts |
| 15.0 mH | 0.31 x Pmax1 | 77 watts |
| 22.0 mH | 0.21 x Pmax1 | 52 watts |
| 27.0 mH | 0.17 x Pmax1 | 42 watts |
| 33.0 mH | 0.14 x Pmax1 | 35 watts |
Example :
For the coil type HQP 56/35 with 4.7 mH, a maximum power of 250 watts for 1% K3 was determined at 8 ohms. With only 2.7 mH, 435 watts would be possible, with 8.2 mH only 142 watts. Coils with significantly lower inductance than the measured inductance (here 4.7 mH) often tolerate slightly more power than calculated, while ferrite core coils with higher inductances (e.g. 10 mH) usually saturate at lower power than expected.
Excursus: Inductance versus saturation
If the saturation current of an inductor is measured, the maximum current (Is2) for 1% K3 of other inductors of the same coil type can be determined approximately as follows:
Is2 = -
Is1 ; Ampere
The inductance of the measured coil is to be entered for L1 and its determined saturation current for Is1, while L2 stands for any inductance of the same coil type.
If the inductance of a coil is halved, its saturation current increases by the square root of 2 to 1.41 times the original value. However, practice shows that there can be a difference of more than 10% between measurements and calculations. Coils with a ferrite core can usually withstand less current than calculated with increasing inductance, while those with a powder core or core made of transformer sheet can often tolerate slightly higher currents than calculated until they reach saturation.
The power limit of a smaller or larger inductor can be roughly calculated as follows:
Pmax2 = 
- Pmax1 ; Watt
The inductance of the measured coil is to be used for L1 and Pmax1 is its determined power limit, while L2 stands for any inductance of the same coil type. Consequently, the power for 1% distortion is halved when the inductance is doubled.
In addition to the inductors with 4.7 mH listed in diagram 8, larger and smaller inductors as well as numerous coils from other manufacturers were also tested. In some cases with astonishing results: despite similar optical and haptic properties, some could only handle a fifth of the power of the coils listed here. The main reason for this is probably the core materials, as the magnetic properties of the various ferrites, iron powders and transformer sheets can be worlds apart, which is not visible to the naked eye. Even coils from renowned German manufacturers showed weaknesses, e.g. in that their copper wire was simply wound onto rod-shaped cores without a stabilizing plastic carrier, so that it can loosen over time due to vibrations in the speaker cabinet. And ferrous cores were not always coated with lacquer to prevent rusting.
Coil recommendations
Audible losses can be avoided by regularly removing layers of oxide and dirt from all contact points between the amplifier and the individual speaker chassis and by selecting choke coils for the woofers whose DC resistance does not exceed 5% of the speaker impedance. Low-pass filters for bass speakers in three-way loudspeakers benefit from core coils because they are smaller and cheaper than low-loss air-core coils. However, this only applies if the core coils are not completely oversized. If, on the other hand, they are undersized, audible distortion will result with powerful pulses. Table 3 shows which core coil type from Intertechnik is best suited for 8-ohm woofers in terms of dimensions and price, depending on the inductance and power handling. For speakers with an impedance of 4 ohms, however, the specified power should again be halved.
Table 3: Optimum coils with an Rdc of max. 0,40 Ω
|
Inductance |
Pulse load capacity min.200 watts at 8 ohms |
Pulse load capacity min. 500 watts into 8 ohms |
|
1.5 mH |
HQ 40/30/095 |
I 78/150/095 |
|
2.2 mH |
HQ 40/30/095 |
I 78/220/085 |
|
2.7 mH |
HQ 40/30/095 |
I 78/270/085 |
|
3.3 mH |
HQP 56/3.3/118 |
I 130/3.3/132 |
|
3.9 mH |
P 62/390/140 |
I 130/3.9/132 |
|
4.7 mH |
P 62/470/140 |
I 130/4.7/132 |
|
5.6 mH |
DR 56/5.6/118 |
I 130/5.6/132 |
|
6.8 mH |
TO 10/6.8 |
I 150-6,8-160 |
|
8.2 mH |
TO 10/8.2 |
I 150-8,2-160 |
|
10.0 mH |
TO 10/10 |
I 150-10-160 |
Coils for other loudspeaker applications
A large inductance is not always connected in series with the woofer, but sometimes also in parallel with it in order to smooth the impedance curve in the bass. This usually requires inductances of over 10 mH, the saturation limit of which should be similar to that of coils connected in series with the woofer.
However, a low DC resistance is hardly needed for this, which is why the very resilient, slightly higher impedance and, above all, cheaper CO44 and CO55 coil types are recommended for the impedance linearization of woofers.
High-pass filters for mid-range drivers also often require quite large inductances in parallel with the speaker chassis. However, upstream capacitors reduce the supplied bass energy so that such midrange coils only have to cope with around half as much as those for woofers. In addition, their DC resistance can easily exceed 5% of the speaker impedance, which makes coil selection much easier. The situation is similar with crossovers for tweeters, in which inductors connected in parallel to the speaker only have to handle low currents, for which small air-core coils are usually sufficient. However, the higher the DC resistance, the larger the design of the coil should be in order to minimize the risk of heat build-up and subsequent melting of the insulating varnish on the copper wire. In contrast to the inductive (alternating current) resistor, the ohmic (direct current) resistor literally burns up part of the amplifier power flowing through it.
The practically distortion-free air-core coils are also ideal for damping high frequencies in mid-range drivers. For bass-midrange drivers in two-way systems, however, the wire of air coils used as low-pass filters should be as thick and low-loss as possible so that their DC resistance is no more than 5% of the speaker impedance, because on the one hand bass drivers are usually 6 dB quieter in the bass range than in the midrange if they are placed in a normal cabinet instead of on a huge wall, as preferred by manufacturers for brochure frequency responses. And secondly, an ohmic series resistor with 10% of the speaker impedance already reduces the available amplifier power by a fifth, one with 20% even by a third.
Berndt Stark
