Saturday, March 30, 2019

History and technical development of the ribbon microphones [1932-2016]

The ribbon microphone was patented in 1932 by Harry F. Olson. It consists of an extremely thin aluminium foil ribbon suspended in a magnetic field. Pressure gradients in the air cause the ribbon to move. The movement of the ribbon in the magnetic field generates a small electromotive force (emf) which can be amplified and recorded. The ribbon microphone quickly gained popularity with audio engineers for its uniform frequency response.



When magnetic tape became the dominant recording media, ribbon microphones became less popular and condenser microphones took over. With recordings and sound mixing processes making use of magnetic media there is always a slight loss of high frequencies. This problem could be remedied by large capsule condenser microphones. These microphones have a number of resonances in the 8 kHz to 12 kHz range that enhances the high frequencies before recording. The capsules of condenser microphones are tensioned tightly, causing the high-frequency resonances. 

The aluminium foil element of the ribbon microphone is only lightly tensioned, causing resonance at very low frequencies. When digital recording became the order of the day, ribbon microphones made a major comeback because high-frequency transfer loss was no longer an issue. Its ability to record fast transients accurately without adding upper-range resonances became again a very positive attribute.

Articles about the ribbon microphone dates back as far as 1931. Most of the early publications were authored by Harry F. Olson who patented the ribbon microphone. Few publications besides those by Olson can be found about the ribbon microphone, but plenty of patents relating to the ribbon microphone are freely available. A selection of those patents is discussed in the paragraphs to follow.

Olson 1932 

Harry Olson filed the first patent for the ribbon microphone in 1931 and the patent was awarded on 25 October 1932 with the title “Apparatus for converting sound vibrations into electrical variations”.

The patent illustrations are of very sturdy mechanical designs as illustrated in picture below:


It describes the ribbon as a relatively small item that is supported in such a way that it resembles the motion of a particle in free air. A device of this nature is classified as a velocity microphone. The combination of mechanical parts surrounding the ribbon is called a baffle. The size of the baffle around the ribbon is calculated according to the highest frequency that the microphone is designed for. The baffle should be designed so that the path length from the front of the ribbon to the rear of the ribbon is half the wavelength of the required frequency. Picture below provides a graph from the patent to illustrate the effect of the baffle size on the frequency response of the microphone. 


The ribbon is made of conducting material that is light in weight and that has low elasticity, for example aluminium foil. The ribbon must not be stretched tightly between its supports. It is crimped in order to suspend it rather loosely between its supports to promote flexibility along its whole length. The supports are made of non-ferro-magnetic conducting material, but it is electrically isolated from the ribbon by non-conductive material. The two signal wire leads are electrically connected to the two ends of the ribbon. The light weight and small restoring force of the ribbon causes its natural vibration frequency to be below the audible range. Tests that were done before the patent submission had shown that a natural vibration frequency of approximately 10Hz produced the most desirable results. The patent states that “When a diaphragm of small mass is suspended in this manner its mechanical reactance is small compared to the impedance of the air. In other words its mass reactance is negligible over a large frequency range compared to the acoustic resistance of the air which it displaces.” 

The ribbon is suspended in the air gap between two poles of a magnet in such an orientation that its surfaces are parallel with the magnetic force lines. The magnet can be a permanent magnet or an electromagnet. The gap between the ribbon and the magnet poles are kept to a minimum to prevent the leakage of air around the ribbon, but the gaps must still be sufficient to prevent frictional contact between the ribbon and the magnet. The patent suggests an air gap of 5.6mm with the ribbon slightly narrower. The ribbon cuts the magnetic field lines while moving in the air gap between the poles because of the sound pressure variations across it. This causes an electromotive force that is proportional to its movement. The electromotive force can be amplified with suitable electronic equipment.  

The magnetic pole pieces with its supporting structure forms the baffle. The baffle increases the path length from the front to the back of the ribbon. The length of this path has an influence on the response of the microphone. The paths around the top and bottom supports of the ribbon are shorter than the paths around the baffle, but its effect is relatively small because it influences only a small part of the ribbon. The clamping structure that secures each end of the ribbon is made of non-magnetic material (ex. copper or brass if it is made from metal). Although it is desirable to make the ribbon as light as possible, it is sometimes necessary to vary its thickness in order to increase its efficiency within the particular baffle setup. The movement of the ribbon is caused by the phase difference of the sound wave between the front and back of the ribbon. The phase difference is determined by the distance that the sound wave has to travel around the baffle from the front to the back of the ribbon. The greatest phase difference occurs when the path length around the baffle is half a wavelength of the sound wave under question. Olson provides a helpful visual representation of this effect in his patent. Picture below shows a horizontal cut through the microphone and a sinusoidal representation of a sound wave. If the path length A-B around the baffle is plotted as E-F on the axis of the sound wave, then the pressure difference between points A and B on the ribbon will be equal to the sum of C-E plus D-F.


The sound intensity at the opposite sides of the baffle is virtually the same for all wavelengths that are longer than twice the distance around the baffle. At wavelengths shorter than this, the intensity on the approaching side of the ribbon increases and the intensity at the retreating side decreases. The reasoning follows that the pressure difference across the ribbon is proportional to the frequency as long as the distance around the baffle is less than half the wave length. Due to the nature of the design the ribbon microphone exhibits very directional characteristics. Sound waves coming from an angle will produce less of a pressure difference across the ribbon. A sound wave directly from the side will produce virtually zero pressure difference. The pressure difference can be calculated with simple trigonometry rules. An alternative design is also provided in the patent. The alternative design does not make use of a ribbon, but of a lightweight diaphragm. This design is however overly complicated and will not be discussed further. 

Olson & Weinberger 1933 
In 1933 Olson, in cooperation with Julius Weinberger, filed a patent that made use of a combination of the pressure gradient (velocity) ribbon microphone and a pressure component microphone to achieve unidirectional operation. Unidirectional operation is desirable in order to improve the ratio of the sound source relative to the sound reflections in the room. The patent achieved its purpose through the use of a normal ribbon (as in the 1931 patent) assembled in series with a modified ribbon. The modified ribbon was adapted to act as a pressure microphone by enclosing the back of the ribbon. The normal and modified ribbons are working in phase for sounds generated in front of the microphone, but are out of phase for sounds originating from the rear side of the microphone. This patent illustrates the influence of the baffle in the extreme case of making the baffle infinitely large by enclosing the rear side of the ribbon completely. 

Anderson 1937 
Leslie Anderson added electronics to Olson and Weinberger‟s unidirectional microphone to file a patent in 1937. The electronics made it possible to adjust the phase differences between the two ribbons, thereby making it possible to adjust the directionality of the microphone from the mixing desk.

Ruttenberg 1938
In 1938 Samuel Ruttenberg filed a patent to address one of the imperfections of the ribbon microphone. When speaking close to the microphone, the low frequencies are over emphasised because the higher frequencies are attenuated. The high frequency attenuation happens because different sections of the ribbon move out of phase. Thus one section cancels out the electrical current that is generated during vibration of another section. The sections move out of phase due to the fact that the ribbon is longer than the wavelength of the higher frequency sound waves. Ruttenberg addressed this problem by designing a special housing for the microphone which closes up the rear of the microphone with an adjustable shutter. When closing up the rear of the microphone, its characteristics are changed from that of a velocity microphone to that of a pressure microphone because the sound wave does not have access to the rear of the ribbon. This idea clearly borrows from the same principles than Olson and Weinberger‟s patent on the unidirectional microphone. The microphone‟s operation in pressure mode tends to minimise the effect of high frequencies being attenuated. This patent clearly illustrates that tampering with the physical surroundings of the ribbon microphone, does have a distinct influence on its operation.

Bostwick 1938
Telephone conferencing has been around for longer than one would expect. In 1938 Lee Bostwick already addressed the problem of feedback between speaker and microphone during teleconferencing. He filed a patent for a device making use of two ribbon microphones and a loudspeaker. The device is constructed with a loudspeaker facing downwards onto a deflector underneath that reflects the sound waves in a horizontal direction toward the conference attendees. Two ribbon microphones are fitted perpendicular on top of each other, both on top of the loudspeaker. Because of its directionality, the ribbon microphones are insensitive to the sound waves emanating from below it, but are sensitive to the voices of the conference attendees around the table.

Bostwick‟s teleconference solution
Olson 1940
In 1940, Harry Olson filed a patent for an improved version of the 1933 unidirectional microphone . Olson discovered that he can design a unidirectional microphone with a single ribbon instead of the dual-ribbon method used previously. Furthermore, this microphone could be easily changed from unidirectional to bidirectional, to non-directional operation. The method of achieving this was to place a pipe structure (resemblance of a smoking pipe) behind a single ribbon. The pipe ends in a labyrinth that is filled with a soft fabric acting as acoustic resistance. The pipe has shutters that can be opened or closed to achieve the desired results, i.e. shutters open allows bidirectional operation, shutters partly closed enables unidirectional operation and shutters completely closed constrains it to non-directional operation. This patent illustrates once again the influence of the surroundings of the ribbon on its operation.  

Anderson 1942 
Leslie Anderson built upon Olson‟s 1933 patent by filing a patent in 1942 about the magnetic equalization of sensitivity in a unidirectional microphone. It is fundamentally an improvement on Olson and Weinberger‟s design by adding a mechanism to adjust the magnetic fields for the two ribbons and by changing the shape of the pole pieces in such a way that it is possible to vary the flux density of one air gap relative to the other. This patent indicates support for Olson and Weinberger‟s idea of modifying the baffle in order to change the operation of the microphone for a specific purpose.

Rogers 1942 
Ernest Rogers filed a patent in 1942 for a microphone with selective discrimination between sound sources . He claimed that the microphone can be used to determine the direction of a sound source. The microphone receives sound waves approaching the microphone straight-on, but attenuates sound waves approaching the microphone from an angular displaced direction. The construction of the microphone is basically four ribbon microphones assembled in an X-pattern. The ribbons are connected to a mixing circuit in such a way that the phase relationship between the ribbons can be adjusted. By adjusting the phase relationship, the microphone can be tuned so that the phases of sound waves from a certain direction cancel each other out, while the phases of sounds waves from another direction will add to each other. 

Olson 1946
The ribbon microphone inherently has a directivity pattern. Olson filed a patent in 1946 to combine two ribbon microphones in a perpendicularly fashion in order to get a microphone that has a 360° pattern on the horizontal plane. This microphone would be ideal for use in orchestras for example, recording all the instruments around it, but attenuating sound waves reflecting from the ceiling and from the floor. Part of the idea is borrowed from Bostwick‟s teleconferencing microphone discussed earlier. 

Anderson 1947 
Anderson‟s patent application in 1947 was filed as an improvement on Olson‟s 1931 design. According to Anderson, Olson‟s design displayed a considerable drop in output when the distance around baffle approaches one fourth of the sound wave‟s wavelength. Anderson‟s design claimed to have a more uniform response over the operating range of the microphone and also to have an enhanced high frequency response compared to the conventional design. This was accomplished by mounting one or more semi-circular bands behind the ribbon. The bands provide cavities that are resonant at the frequencies where the enhancements are deemed necessary.  


Olson 1950 
Olson and Preston applied for a patent in 1950 that introduced a new magnet design for the ribbon microphone. The new design brought the magnet closer to the ribbon by doing away with the pole pieces and using a magnet structure consisting of two large magnets, one on each side of the ribbon, with an oval hole in each magnet. The holes in the magnets are there to reduce the path length from the front to the back of the ribbon, thereby extending its upper frequency response. Olson‟s motivation for the design is that the efficiency of the magnetic structure increases as the magnetic source is placed closer to the air gap because the amount of leakage flux decreases. It is also an attempt to minimise the effect of the pole pieces on the difference in sound pressure between the front and the back of the ribbon.

Anderson 1954
In 1954 Anderson filed a patent concerning the magnetic circuit of the ribbon microphone. This patent addressed the problem of the air gap between the ribbon and the magnetic structure. If the gap is too small, the ribbon might touch the structure as it moves. If the gap is made larger, then the sensitivity of the microphone is adversely affected. Therefore the gap must be adjusted to a very specific distance. The patent provides no recommendations about the optimum size of the air gap or anything about the magnetics itself. It is purely an assembly method to mount the magnetic yoke pieces accurately and reliably in a repeatable manner. 

Fisher 1965
Charles Fisher addressed the upper frequency limit of the ribbon microphone in his 1965 patent. The patent illustrates the design of a ribbon microphone with an exceptionally large horseshoe magnet at the base with conical pole pieces attached to the magnet. The conical pole pieces are constructed such that it does not only taper off to the top, but also makes the opening for the ribbon narrower as it protrudes away from the magnet. The pole pieces and the ribbon are thus broad at the end closest to the magnet and narrow at the end furthest from the magnet. The front to rear path around the baffle at the furthest end is thus much shorter than that of the conventional design. The patent claims good response and directivity up to 20 kHz, but no frequency response graphs are provided to support the claim. 

Royer 1999 
For three and a half decades (mainly the period that the condenser microphone dominated) the US patent office did not receive any submissions concerning the ribbon microphone. David Royer and Richard Perrotta broke the silence in 1999 with their patent to modify the original ribbon microphone. The digital conversion of sound recordings was the motivation behind their invention. Non-ribbon microphones produce large frequency dips and phase distortions that get falsely interpreted by the digital equipment as valid data. Ribbon microphones do not have this problem. Royer and Perrotta deemed it necessary to modify the ribbon microphone so that it could be used at higher volume levels. To achieve the desired effect they designed a ribbon microphone with magnetic pole pieces which are much wider from front to back and with the ribbon positioned not in the centre, but one quarter of the pole width from the front of the microphone. In the original ribbon microphone the pole pieces are tapered to become narrower towards the ribbon. In Royer and Perrotta‟s design the pole pieces are purposely not tapered. The frequency response graph in the patent application shows a very flat response from 40 Hz up to 15 kHz. Frequencies lower than 40 Hz are attenuated by as much as 5 dB (at 20 Hz) and frequencies above 15 kHz are attenuated by as much as 7 dB (at 19 kHz). No graph is provided to compare the volume levels of the original (prior art) with the new design. This design increases the baffle size slightly with the increased thickness of the magnetic poles. Olson‟s 1931 patent only shows the frequency response between 100 Hz and 7 kHz, so no direct comparison can be made between the original and the new designs concerning the effect of the baffle size on the frequency response.  

Royer 2004
In 2004 Royer and Perrotta filed two more patents. They filed two patents for the same microphone on the same day, but each application emphasised a different aspect of their invention; thereby slightly changing the classification of each patent. One patent emphasises an angled magnet structure to improve the sensitivity and frequency response of the microphone. The other patent emphasises the specially designed transformer for said microphone. The applicants incorporated some of the principals of their previous patent, but this microphone was designed with pole pieces that taper narrower towards the ribbon (unlike the thicker magnetic poles of the 1999 patent). The angled magnet design effectively reduces the front-to-rear distance around the baffle. This is in accordance with Olson‟s rule that a shorter distance around the baffle will produce a higher frequency range. The angled magnet design aids in the sensitivity of the microphone because it maximises the amount of magnetic flux lines running perpendicular to the ribbon. Totally perpendicular magnets would have the highest impact on the sensitivity, but that would increase the path length around the baffle. The patent‟s design with magnets at an angle less than 80º, provides a good compromise between the baffle length and flux direction in order to get good sensitivity while also improving the frequency response. Royer and Perrotta made use of Neodymium magnets in this design. The Neodymium magnets are much more powerful than the magnets used in previous designs. The alloy for the pole pieces in this patent is Permendur or Hyperco 90 which has a much higher magnetic permeability than iron. The magnetic structure has thus been improved in more than one way. Detailed dimensions for the air gap and the ribbon are provided in the patent. This particular design also helped to reduce the size and weight of the microphone. Much other detail is also provided, but no graphs are provided to support the claims of the improved sensitivity and frequency response.

Royer‟s modern version of the ribbon microphone

The special purpose transformer is a complex toroidal transformer design. It has four primary coils, is a tape-wound-core transformer, has turn ratios of 100:1 and 200:1, includes at least three interleaved secondary coils, and/or includes at least three interleaved primary coils. The primary and secondary windings are also interleaved with each other. A buffer circuit is also enclosed within the outer casing and is electrically coupled to one of the secondary coils. The input impedance of the buffer circuit is at least 10 MΩ including a field-effect transistor. The core of the transformer consists of a “tape wound” configuration instead of compressed ferrite or composite materials. The “tape wound” core is constructed from thin M-6 nickel iron strips that are rolled up very tightly into the shape of a doughnut. These two patents are both noted for their thorough description of the workings of a ribbon microphone and its illustration of the improvements that are possible with modern technology.

Akino 2005 
Hiroshi Akino addressed the design of the ribbon (not the microphone motor) in his patent application of 2005. He patented the idea of depositing a thin layer of gold over the traditional aluminium ribbon. Gold is also deposited to the ends of the electrode plates. The reason for the gold plating onto the aluminium ribbon is to prevent oxidation of the ribbon. The electrode plates are also gold plated to prevent corrosion due to intermetallic electrolytic reactions. Preventing corrosion at the ribbon-electrode-junction ensures that the impedance of the junction remains as low as possible, thereby preventing the loss in sensitivity that would usually occur over time. 

Crowley 2005 
During the same year Robert Crowley filed a lengthy patent with twenty different claims. The patent aims to enhance quality and repeatability during the manufacturing process of the ribbon microphone. Crowley stresses the importance of the ribbon in his patent. He addresses the tradeoff between mass and resistance of the ribbon by using composite materials. The ribbon in this patent is claimed to consist of carbon fibre nanotube filaments attached to a layer of conductive material. 

Tripp 2007
With the main focus still on the ribbon, Tripp and Crowley filed a patent in 2007 for the invention of a polymer ribbon. Their patent makes us of a polymer ribbon that exhibits high toughness, high conductivity and good shape-memory. The ribbon‟s corrugated structure is formed by compressing the ribbon between two dies and then heating and cooling it to set it permanently into this shape. The polymer ribbon is coated with a conductive coating. The polyethylene terephthalate (PET) that it consists of does not become brittle with age because it contains no plasticisers. The invention claims that the ribbon will not be damaged by phantom power (48VDC applied by mixing boards) because the PET that is used is about eight times stronger than a normal aluminium ribbon. Another advantage of the high strength of the PET is the fact that the ribbon can be made much thinner and thus lighter, making it more responsive to low intensity sound waves. 

Akino 2009
In 2009 Akino patented another ribbon design. In any ribbon microphone design the ribbon must be set to a resonance frequency that is below the lowest frequency to be recorded. To achieve this, the tension on the ribbon must be very low. The corrugated form of the ribbon helps to realise this low tension. The corrugated ribbon is manufactured by feeding the ribbon through a set of spur gears. According to the patent the flat areas between the corrugations cause a high mechanical resonance within the required frequency range to be recorded. To solve this problem a ribbon was invented onto which a special pattern is transferred by means of a roller with a transfer mould. The selective embossing of the ribbon produces a more rigid ribbon that suppresses the sharp resonances. On the frequency response graphs in the patent it can be seen that the specially formed ribbon suffers less from high frequency resonances (noticeably in the region above 10 kHz), but it suffers from lower sensitivity above 5kHz. 

Horng 2009
The National Chung-Hsing University in Taiwan recognised an opportunity to address the design of the ribbon with the aid of micro electro-mechanical systems (MEMS) technology. The patent claims that the frequency response of the microphone will be better if its sensitivity is improved and that the sensitivity can be improved by using a lighter ribbon. The weight and the size of the ribbon are however restricted by conventional manufacturing processes. According to this patent it is the size and weight restrictions of the ribbon that causes the attenuation of low and high frequencies. By applying the techniques that are used in semiconductor fabrication and in MEMS, the patent describes the manufacturing of a smaller diaphragm and voice coil. No dimensions are provided in the patent to judge the reduction in size and no frequency response characteristics are given to illustrate any improvements. 

Cloud 2010 
With new designs for the ribbon itself being thoroughly covered since 2005, Cloud and Sank‟s patent focussed on other aspects of the ribbon microphone, namely the magnet motor assembly, a back-wave chamber and electronics. The patent introduces round magnet poles to the design. The rounded poles cause sound waves to reflect of it in a diverging pattern, thereby diffusing the reflected sound waves. Therefore the ribbon reacts to the sound with minimal interference from sound waves reflecting off the pole pieces. The result is a flatter frequency response compared to conventional flat magnets. No frequency response graphs are provided in the patent to support this claim. Another part of the patent illustrates a different design with a horizontal ribbon and a blast filter in front of the ribbon to protect it against high pressure sound waves. The same configuration also has a chamber at the back of the ribbon which prevents reflected sounds from entering the microphone from the rear. This design is completely different form the normal upright configuration of the ribbon microphone, looking more like the ice-cream cone design of an ordinary condenser microphone. The design is complemented by the addition of a differential cascade JFET amplifier that is powered by phantom power from the recording desk. The patent claims a signal boost of +20 dB with its JFET amplifier. 

Akino 2012
In his 2012 patent, Akino claims that the ribbon can be protected against impact by short-circuiting the ribbon‟s electric circuit when it is not in use. He claims that the back-emf generated in the ribbon will suppress its movement significantly due to electromagnetic damping. No calculations or data is provided to quantify his claim. Although the back-emf effect is a scientific fact, the author is of the opinion that the electromagnetic damping will not be significant enough to qualify the patent as anything more than a marketing exercise. This patent presents a good simulation challenge for students to determine how substantial Akino‟s claim is. 

Other Publications 
Most literature that can be found on the ribbon microphone is merely concerned with patents and advertising. The few publications that could be found with academic content on the ribbon microphone are summarised below.

Ke 2009 
Ke, et al. designed a symmetrical voice coil using MEMS technology. The aim of the MEMS design was to increase the effective length of the coil while reducing its physical dimensions. They succeeded to manufacture a voice coil with a 77.5 mm effective length while limiting the overall length of the ribbon to 17 mm. Compared to a standard 50 mm ribbon, it is a gain of 55% in effective length with a 66% size reduction.Picture below provides a drawing from Ke‟s paper to illustrate the design. The voice coil is designed in such a way that half of the windings run through the magnetic field (all in the same direction) while the returning half of the windings run outside of the magnetic field (in the opposite direction). If the returning conductors were also inside the magnetic field, they would have generated an emf in the opposite direction. By having the return path outside of the magnetic field, the total length of the electric conductor in the magnetic field is increased without the returning conductors cancelling the emf. The paper includes an overview of the design and a COMSOL® simulation of the magnetic flux density between the magnets. No frequency response information is provided. The frequency repose of this novel design would provide valuable information since the voice coil is fixed to the magnets and not allowed to move freely in the magnetic field like a conventional ribbon.

MEMS ribbon

Horng 2011
Horng presented a conference paper in 2011 after further study on the miniature ribbon microphone design that Ke published in 2009 under Horng‟s guidance. The conference paper expands on the previous publication by illustrating more detail on the MEMS manufacturing process and providing test results on the displacement of the voice coil, the velocity of the diaphragm, and the microphone‟s frequency response. It was shown that the displacement of the bare diaphragm (without coil) is slightly larger than the displacement of the diaphragm after printing of the coil. The reason is the additional weight of the coil. The displacement of the voice coil is almost linear with increasing sound pressure over the measurement range of 10 Pa to 100 Pa. It is not clear what equipment was used for the displacement measurements. The velocity measurements were done with a laser Doppler vibration system. Although the displacement tests showed a minor difference, the velocity tests showed a more prominent difference between the bare diaphragm and the diaphragm with coil. The measurement results of displacement and velocity cannot be directly compared since the displacement measurements were performed from 10 Pa upwards and the velocity measurements were done at a sound pressure level of 1 Pa. A displacement of 0.5 µm was measured for the diaphragm (including coil) at 10 Pa. At a pressure of 100 Pa a displacement of 6.0 um was measured. The velocity at 1 kHz with a sound pressure of 1 Pa is only 1.3 µm/s. This provides the reader with an understanding of just how small these measurements are (and thus prone to measurement errors). Horng‟s frequency response measurements show an increasing sensitivity from 100 Hz (-64 dBV/Pa) up to 20 kHz (- 50 dBV/Pa). It is worth noting that the microphone‟s sensitivity is not dropping at high frequencies (> 10 kHz) as one would expect from a closed baffle design. 

Kwon 2016
Kwon and Honorato performed out of the ordinary tests with a ribbon microphone at the Naval Postgraduate School in Monterey, California. They tested the ribbon microphone as an underwater transducer. They observed that its directional characteristics approach that of an ideal dipole with its two lobes becoming narrower with increasing frequency. 





No comments:

Post a Comment