Please keep your distance !
How coils in crossovers rub off on each other
What was that again with the choke coils ?
They consist of enamel-coated, wound copper wire and reduce the alternating current flowing through them with increasing pitch or frequency. They owe their ability to more or less block overtones to magnetism. Every electrical conductor, including copper wire, generates a magnetic field around it when an electric current flows through it. And it is precisely this magnetic field that induces counter voltages in the conductor, which steadily increase as the frequency of the alternating current rises.
But now comes the "small print": the magnetic field generated by a coil also reaches neighboring inductors, in which it induces alternating voltages. During music playback, each coil of the crossover therefore transmits signals to other coils in its vicinity, which can affect the sound of loudspeakers. An experimental setup was designed to give an initial impression of whether this external induction is audible or negligible.
A
tone generator was the signal source and an amplifier sent its sine tones with one to two watts through a coil with an upstream load resistor. The coil served as a transmitter, so to speak, and a second one was placed nearby as a receiver. The ends of the wires were wired to the positive and negative terminals of a loudspeaker. It is hard to believe how clearly the sine tones could be heard through the loudspeaker right into the treble range - and that without any electrical connection between the amplifier and the receiver.
This experiment showed that the mutual influence of the coils on a crossover is by no means insignificant. But what to do?
Trial and error makes perfect sense
Changes to the coil positions showed that the extent of the interference depends primarily on two factors that can be controlled: the distance between the coils and their geometric alignment. And in order to be barely audible, an induction voltage of one thousandth of the original amplifier voltage proved to be suitable, which corresponds to a signal-to-noise ratio of 60 decibels. In other words: If the amplifier sends 10 volts into the transmitter coil, no more than 10 millivolts should be measurable at the receiver coil. The next step was to try out how to position the coils so that they would leave each other alone. For this purpose, a millivoltmeter replaced the loudspeaker in the figure above, and initially two different inductances served as test subjects: the different coil types available at Intertechnik with 4.7 mH as the transmitter and an air-core coil with 0.47 milliHenry as the receiver. Both are typical inductances for speakers with 8 Ohm impedance - the large one for woofers, the small one for midrange and tweeters.
In principle, the reverse distribution of roles is also conceivable, i.e. a small inductance as a transmitter and a large one as a receiver. In practice, however, this case is rather rare, as the highest currents generally flow through the large coils in front of woofers. The following diagram illustrates how much electrical energy reaches the individual loudspeaker branches of a three-way speaker.
If a crossover separates at 400 Hz and 3.5 kHz, the woofer will receive an average of around 56 % of the power output from the amplifier when playing music, the midrange driver around 34 % and the tweeter 10 %. The underlying red curve in the diagram is based on spectral analyses of various pop, rock, jazz and classical recordings.
The power distribution of pink noise or IEC and DIN noise should only serve as a comparison here, as these types of noise may be familiar to some.
Stray field and inductance
The experiments showed that the sensitivity of a (receiver) coil to externally induced interference voltages increases with its inductance. Even with optimum placement or alignment, a distance of 20 mm or more between neighboring coils is often necessary to reduce interference signals to -60 dB. This is particularly important for inductances in the mid-range in terms of sound, especially as woofer coils also scatter mid and higher frequency signals, which they largely prevent from reaching the woofer itself due to their inductance.
The tests revealed the following correlations: with double the inductance of the transmitter coil (10 mH instead of 4.7 mH), the minimum distance specified in the following tables was reduced by around 30 %. By contrast, with only half the inductance of the transmitter coil, the minimum distance increased by around 50 % compared to the values in the table.
The opposite was true for the receiver coils: if their inductance was doubled from 0.47 mH to 1.0 mH, a distance of around 50% greater than that of the transmitter coil was required. On the other hand, if the inductance of the receiver coil was only halved, its distance from the transmitter coil could be around 30% smaller than shown in the following tables.
Here are the minimum distances that were determined using the test setup described above, starting with round or roller-shaped coils.
Roller-shaped coil in pos. X A1 to A3: Alignment of the receiver coil to the transmitter coil .
A1 A2A3
| Pos.X | Transmitter coil | A1 Alignment LU32/26 | A2 Alignment LU32/26 | A3 Alignment LU32/26 |
| 1 | HQS 32/26 | 37 | 10 | 0 |
| 2 | COT 92/39 | 50 | 15 | 0 |
| 3 | DR 56/35 | 63 | 15 | 0 |
| 4 | HQP 56/35 | 55 | 20 | 0 |
| 5 | HQP 62/47 | 63 | 28 | 0 |
| 6 | HQ 58/46 | 67 | 36 | 0 |
| 7 | HQ 43/45 | 65 | 37 | 0 |
| 8 | DR 56/61 | 75 | 58 | 0 |
| 9 | TO 10 | 5 | 38 | 5 |
| 10 | LU 92/39 | 41 | 6 | 0 |
| 11 | LU 120/55 | 43 | 18 | 0 |
Minimum distance between the coils in millimeters for -60dB interference voltage.
Pos. X : Position or orientation of the transmitter coil with 4.7 mH
A1, A2, A3 : Alignment of the receiver coil (0.47 mH, LU32/26) to the neighboring transmitter coil
Coils whose designation begins with HQ, DR or P have ferrite cores of different types, COT stands for a Corrobar powder core coil, TO for one with a toroidal core and LU for air coils. Air coils without magnetizable materials in the vicinity of their windings (see items 10 and 11 in the table) generate a similarly strong magnetic stray field around them as core coils, and both variants require comparable distances from their colleagues on the crossover. The winding direction of the copper wire of neighboring coils has the greatest influence here: If it is the same for both, as in case A1, the coils must be given a large distance, whereas if the windings are perpendicular to each other, as in A3, the coils may even be able to come into close contact without interfering with each other. The toroidal core coil TO 10 (see item 9 in the table) seems to be a little out of line here, which is due to its special wire winding around the toroidal core.
What are the minimum distances when the position of the transmitter coil is changed?
Roller-shaped coil in pos. Y A1 to A3: Alignment of the receiver coil to the transmitter coil.
A1A2A3
| Pos.Y | Transmitter coil | A1 Alignment LU32/26 | A2 Alignment LU32/26 |
A3 Alignment LU32/26 |
|
1 |
HQS 32/26 | 16 | 64 | 0 |
| 2 | COT 92/39 | 55 | 95 | 3 |
| 3 | DR 56/35 | 45 | 105 | 0 |
| 4 | HQP 56/35 | 40 | 90 | 0 |
| 5 | HQP 62/47 | 43 | 97 | 7 |
| 6 | HQ 58/46 | 48 | 107 | 0 |
| 7 | HQ 43/45 | 36 | 100 | 0 |
| 8 | DR 56/61 | 50 | 105 | 0 |
| 9 | TO 10 | 13 | 13 | 10 |
| 10 | LU 92/39 | 56 | 90 | 0 |
| 11 | LU 120/55 | 60 | 100 | 0 |
Minimum distance between the coils in millimeters for -60dB interference voltage.
Pos. X: Position or orientation of the transmitter coil with 4.7 mH
A1, A2, A3: Alignment of the receiver coil (0.47 mH, LU32/26) to the neighboring transmitter coil
If the position of the transmitter coil is changed, the following also applies here: If the winding direction of the transmitter and receiver coils is the same, as is the case with A2 this time, a very large distance is required so that they do not affect each other audibly. Only if their windings are perpendicular to each other, as is the case with A3, can the coils come quite close to each other. The Torobar TO 10 toroidal core coil (item 9 in the table) is again an exception due to its special wire winding around the toroidal core.
Not all choke coils are round or roll-shaped. And especially those with a magnetizable core made of transformer or iron sheet proved to be somewhat peculiar with regard to their scattering, as the following measurement data show.
Transformer and I-core coil as transmitter A1 to A3: Alignment of the receiver coil
to the transmitter coil
Pos. X
Pos. Y
A1A2A3
Item Z
| Pos.X | Transmitter coil | A1 Alignment | A2 Alignment | A3 Alignment |
| 1 | I 78 | 5 | 85 | 3 |
| 2 | I 96 | 16 | 90 | 3 |
| 3 | I 130 | 34 | 105 | 0 |
| 4 | I 150 | 40 | 110 | 5 |
| 5 | FE 96 | 20 | 20 | 0 |
| 6 | FE 130 28 | 19 | 0 | 0 |
| Pos.Y | Transmitter coil | A1 Alignment | A2 Alignment | A3 Alignment |
| 1 | I 78 | 5 | 63 | 63 |
| 2 | I 96 | 7 | 77 | 5 |
| 3 | I 130 | 20 | 90 | 5 |
| 4 | I 150 | 29 | 98 | 5 |
| 5 | FE 96 | 0 | 33 | 25 |
| 6 | FE 130 | 0 | 0 | 35 |
| Pos.Z | Transmitter coil | A1 Alignment | A2 Alignment | A3 Alignment |
| 1 | I 78 | 0 | 0 | 72 |
| 2 | I 96 | 0 | 0 | 81 |
| 3 | I 130 | 0 | 8 | 90 |
| 4 | I 150 | 0 | 3 | 92 |
| 5 | FE 96 | 10 | 50 | 5 |
| 6 | FE 130 | 18 | 60 | 0 |
Minimum distance between the coils in millimeters for - 60dB interference voltage.
Pos. X to Z : Position of the transmitter coil with 4.7 mH
A1, A2, A3 : Alignment of the receiver coil (0.47 mH, LU 32/26) to the transmitter coil
The "I" in the coil designations stands for a bar-shaped core made of transformer sheet metal, the "FE" for standard transformer cores in the form of the letter E
That's right :
The illustrations below show the best way to position coils on crossovers in order to minimize mutual interference. Viewing angle from above.
Roller-shaped coil Roller-shaped coil
Winding direction
Adjacent coils have the least influence on each other if their windings are perpendicular to each other. The blue arrows illustrate the winding direction.
Transformer and roller coil I-core and roller coil
The winding direction of coils with transformer or I-core should also be perpendicular to that of the neighboring coil.
Not like this if possible!
Transmitter DR56/61 Receiver HQ40/30
with 4.7mHwith 2,7mH
268mm
The influence proved to be astonishingly large when the receiver coil (HQ 40/30 in this case) also had a high inductance and both coils were positioned virtually head to head. A distance of around 27 cm was required for -60 dB interference voltage!
Conclusion
When alternating current flows, the choke coils required for loudspeakers send magnetic energy into the other coils of the crossover and induce alternating voltages in them, which can impair the loudspeaker sound. This annoying property is more or less inherent in all types of coils, whether they have a magnetizable core or not.
However, the negative impact of this effect can be significantly reduced by positioning and aligning the individual coils so that their winding directions are perpendicular to each other and by spacing inductors a few centimeters apart in cases of doubt.
But what kind of doubtful cases are meant?
If, for example, several large inductors are used in the low-frequency range, which make optimum placement and alignment on the crossover board difficult simply because of their dimensions, this is such a case of doubt. The strong magnetic fields of closely neighboring bass coils can even modulate each other, which then leads to very peculiar interference noises.
And larger inductances (1.0 mH and more) in the midrange branch are likely to be particularly sensitive to interference from woofer coils because the sensitivity of the receiver coils increases with their inductance. In addition, the dispersion of even large inductances does not only contain low frequencies, as one might think, but almost the entire audible spectrum.
Coils in the tweeter branch, on the other hand, will react more sensitively to the series coils in front of midrange drivers than to the large inductances for the bass range. At least if they are aligned with the large colleagues as recommended above. This is because the larger the inductance of the transmitter coil and the smaller that of the receiver coil, the lower the electromagnetic scattering effect.
If the dimensions of the crossover board do not allow sufficient distance between the coils, separate boards are recommended, for example one for the bass and a second for the mid and treble range.
Berndt Stark
