Moulton Lectures
On
Electro-Acoustics
|
Lecture
28 Rocking Armature Transducer 1 |
Presented
by:
Dave
L Moulton
Assisted
by
David Poulten and James Burt
Location: Racal Acoustics Limited (UK)
Date: 21-April-2004
Background
Lecture 27 was intended to act as an introduction to
the fundamental principles behind the operation of a range of transducers which
function by means of the variable reluctance in a moving air gap. We have seen
in a magnetic circuit that an air gap also has an associated force across the
gap. By varying the flux density across the gap we can alter this force and
cause a mechanical structure such as a diaphragm to move in sympathy with the
varying flux density. This principal of operation will form the foundation of
this lecture, which aims to explain the physics behind what is known as a
rocking armature transducer. This type of transducer is also known as a
balanced armature or variable Reluctance transducer and is in widespread use
throughout the world in many types of audio ancillary. Such examples are the
earphone and microphone transducers found in many types of commercial and military telephone handsets. Although, it
has to be said as time moves on these transducers are becoming less common due
the advent of mobile phones and miniature lightweight headsets. These tend to
use moving coil earphones and miniature electret microphones as the main audio
transducers.
Along with many other companies throughout the world
Racal Acoustics still manufacture rocking armature transducers and have made
many process improvements over the years.
Assistance with this lecture
I have called upon the valued experience of two work colleagues David Poulten and James Burt, to assist with this lecture by showing examples of the rocking armature build process, and to give an insight into the various test and setting methods.
Content
This lecture will be dedicated to the
electro-magneto-mechanical construction of a Rocking Armature transducer. A
separate lecture will be held to explain some of the elecro-acoustic properties
of the transducer. Thus in this lecture I will cover the following.
·
Outline the electro-magneto-mechanical construction.
·
Show the equivalent magnetic circuit.
·
Explain the mechanical Force balance.
·
Explain a bit about the build process
·
Explain the criteria for setting the sensitivity
·
Show a typical frequency response
On With The Lecture
Firstly I would like to build up a simplified construction showing the fundamental elements of a rocking armature transducers.
28.1 Magnet and Yoke assembly
Consider a small rectangular magnet made from Alcomax, attached to a paramagnetic yoke assembly made from Permalloy.
Figure 28.1
In figure 28.1 we shall consider the magnet to be fluxed to saturation.
28.2 Wind coils around each of the two outer arms of the yoke assembly
Consider two coils wound around each of the arms of the yoke assembly as shown:
Figure 28.2
In this case the two coils are wound such that the ‘start’ of coil 1 rotates clockwise onto the yoke and then continues onto coil 2 in an anti-clockwise direction. The wire eventually exits coil 2 and is designated the ‘finish’. The coil wires are generally single stranded thin insulated copper wire. The actual number of turns wound on each coil (the same number for each coil) is set depending on the intended impedance of the transducer.
In reality, the two coils are wound by a machine onto a plastic bobbin. Several thousand turns of wire are often wound onto each coil. The completed bobbin assembly comes off the machine in a folded state. See figure 28.3 below:
Figure 28.3
The bobbin is unfolded through 90º and then fitted onto the magnet and yoke assembly as shown in figure 28.4.
Figure 28.4
The bobbin is sealed to the yoke and magnet assembly as shown in figure 28.5.
Figure 28.5
To get a more clear picture of the complete assembly see figure 28.6, which gives a three dimensional perspective.
Figure 28.6
28.3 Adding the armature
Ok, so we have now got a coil attached to the Yoke and magnet assembly. Let us now go back to the simplified example shown in figure 28.2 and add an armature. The armature is pivoted on top of the magnet and secured onto the yoke support arms. The result is an assembly which forms a gap at both of the yoke bobbin arms and is able rock about the magnet pivot point. The restoring force in the armature movement is provided by the torsion stiffness in the armature arms.
Figure 28.7
28.4 The mechanical-magneto circuit under steady state flux conditions
If we assume that the coils have no current flowing in them, then we can consider the system to be in a steady state condition, this is shown in figure 28.7. Assuming that the magnet has been pre fluxed to saturation we can draw a simple DC magnetic circuit for the system. In this case I will remove the coil for convenience. We will also make the valid statement that the Reluctance of the air gaps is far greater than the reluctance of the armature, yoke and magnet combined, thus:
Figure 28.8
Thinking back to what was learnt in Lecture 27, clearly we have two magnetic loop circuits, both with a flux derived from the permanent magnet. Thus. Assuming we have two gaps with reluctances R1 and R2, then we can write
Writing in terms of flux density we have:
The gap flux densities B1 and B2 result in forces occurring across both of the air gaps. These forces tend to attract the armature towards the yoke. This is analogous to two people on a swing. If the fulcrum is at the mid point of the swing and the two people are symmetrically spaced about the fulcrum, then the rotation will be in the direction of the heaviest person. i.e in the direction of the greatest gravitational force.
Figure 28.9
In order for the armature to remain perfectly balanced and not be biased either clockwise or anticlockwise, the forces across both gaps need to be balanced. Thus assuming constant gap areas, then the gap widths need to be set the same. Under this condition we can write:
Thus the flux density that produces the force can be expressed as:
The force across the cap is related to the ratio of the gap area to the magnet area and to the magnitude of the permanent magnet flux density. Typically the force across the gap is related to the flux density in the following way:
In order to ensure that both gaps are the same. The magnet and yoke assemblies are lapped and polished. The lapping process is a bit like skimming a motor car engine block before applying the head gasket. The polishing ensures that the lapped surface is perfectly smooth and free from contaminants. This process occurs at the early stages of construction, usually just after the magnet has been welded to the yoke.
Now there is another force that comes into play and that is the Torsion force within the fixed arms of the armature. This force can be related to a mechanical stiffness or restoring force. Thus for the armature to move the torque produced by the force in the air gap has to also overcome the torsion force in the armature.
The magnetic force will displace the armature towards the coil arm of the yoke by an amount depending on the armature restoring force. The reduction in the gap will result in an increase in the force across the gap. Thus if the forces cause the gap length to reduce by an amount Δl, then the force will increase by an amount:
If the armature is not correctly balanced it is possible for the magnetic force across the gap to exceed the torsion restoring force. When this occurs the armature will collapse onto the yoke arm on the side providing the strongest force. This is often referred to as ‘Pole Over’
In real life the torsion arms are adjusted for stiffness and the gap is set very accurately with a feeler gauge. This coupled with the strength of the permanent magnet results in a system of balanced forces. Thus the mechanical forces are balanced with the magnetic forces.
Figure 28.10
Typical gap settings are 0.089mm ± 0.013mm.
28.5 Adding
a diaphragm connecting pin
In order to get our system to translate mechanical movement into perceivable acoustic sound energy we need to house the assembly and add a diaphragm. Before we add the diaphragm we need a means of linking the end of one arm of the armature to the diaphragm. This is done using a small pin made from radial metal, welded to the armature. This process is usually done before the armature is welded to the yoke and magnet assembly.
Figure 28.11
28.6 Housing the Magnet,Yoke and Armature Assembly
The whole assembly is now housed in what is commonly known as the rear case, this is usually moulded from plastic. The magnet, yoke and armature assembly shown in figure 28.11 is commonly known as the ‘Motor Unit’.
Figure 28.12
Figure 28.13
28.7 Adding
the diaphragm
Unlike the situation depicted in lecture 27, where the diaphragm stiffness played an important part in the balance of forces. The stiffness of the diaphragm in a Rocking Armature transducer has no affect on the balancing of the armature. Thus a diaphragm made from very light material and with optimum acoustic radiating area can be used. This gives the Rocking armature the advantage of being very sensitive for its size. The low stiffness of the diaphragm provides a relatively low acoustic radiation impedance and thus radiates much more efficiently into a low impedance medium such as air.
The diaphragm is generally dished in shape. This reduces the affect of resonant modes across the diaphragm and allows for a relatively flat response from 300Hz up to 2.5kHz. The transducer response falls off very rapidly after 2.5kHz. (This is often increased yo 3.5kHz with the addition of secondary acoustically tuned front cap).
The diaphragm contains a small hole at its centre. The pin on the armature fits through this hole when the diaphragm is fitted on to the motor unit. A small controlled amount of resin is applied to the pin to secure it to the diaphragm. The amount of resin applied is critical as its volume can be significant in affecting the overall acoustic cavity compliance between the diaphragm and a primary front cap.
Note: it is important that the pin enters the diaphragm perpendicular to the armature. If the pin is bent. There will be a loss in mechanical transmission between the armature and the diaphragm. This is to be expected as a bent pin will only be able to transmit a component of the armature force to the diaphragm.
Where Ft is the resultant transmission force and θ is the angle of the pin relative to a true perpendicular to the armature.
Figure 28.14
28.8 Adding
a front cap
To complete the job, a front cap is fitted and required to do two things.
· Lock the edge of the diaphragm into place.
· Provide an acoustic loading for the diaphragm.
The front cap is secured into place by means of being crimped around the outer edge of the rear case. The front cap also has a series of laser controlled bleed holes located in equally spaced circular patterns on its face. These bleed holes provide acoustic resistance and inertance which forms part of the front cap filter characteristic. (Governed mainly by the cavity enclosed by the cap and the diaphragm). The detailed acoustics of this device will be discussed in another lecture.
Figure 28.15
28.9 Applying a signal to the transducer coils
Consider the situation where we apply a sinusoidal current through the transducer coils. Each of the two coils will generate an induced sinusoidal flux Φi which will add sinusoidally to the static flux Φg in each of the gaps. Thus the total flux in the left arm ΦglT can be described as:
It follows that:
The two coils are wound in such a way so that the induced flux seen in each of the two air gaps are 180º out of phase with each other. As the flux density in one air gap increases, the flux density in the other decreases. Thus, in our example, the equations for the right hand air gap will be.
And
The negative sign takes into account the 180º phase change due to the direction of the two coil windings.
The result is a complementary force across each of the air gaps, resulting in a turning moment (torque) across each of the armature arms. The armature is thus forced to rock in sympathy with the induced coil current.
The direction of the static flux through the air gaps is determined by the initial polarization of the magnet. In the case of figure 28.15 above, we have polarized the magnet so that the North face is in contact with the armature. In this situation the static flux lines circulate out of the North face and return through the south face. Thus, in this case the static flux lines are directed across the gap in a direction from the armature to the yoke.
Note: the direction of the two coils and the polarization of the magnet determine the phase of the transducer. If the transducer is wound in phase, then the peaks in the induced coil current will result in the diaphragm being moved upwards, creating a positive pressure. If the earphone is wound out of phase, then the induced current peaks will result in the diaphragm being pulled downwards, thus creating a negative pressure. The phase relationship between the induced current and the acoustic pressure will change with frequency, this is due to the reactive nature of the whole electro-mechanical-acoustic system.
It is normal practice to define the DC phase of the transducer by means of a mark on the appropriate electrical contact tag, thus identifying the start of the coil winding in a given direction.
28.10 System efficiency
Without going into a lot of mathematical detail, the whole system is at its most efficient when the mechanical forces across the air gaps are perfectly balanced by the torsion in the armature arms and the acoustic air mass loading on the diaphragm (Defined by the acoustic impedance seen by the diaphragm, made up of cavities, holes and air). By efficiency, I mean that for a given electrical input the acoustic output is at its optimum level. Thus, the transfer of input electrical energy to output acoustic energy is at a maximum when the system is perfectly tuned.
28.11 The risk of Pole Over
One of the risks with a balanced armature system is that one side of the armature could come into permanent contact with the yoke, thus closing the air gap. This situation is usually irreversible and renders the transducer in-operable, as described earlier, this event is often termed as ‘pole-over’. The risk of pole-over can be attributed to several factors:
· The static magnetic force is set greater than the armature torsion restoring force.
· The air gaps are set too short.
· The permanent magnet is fluxed too high.
· The torsion in the armature arms is set too low.
· The transducer suffers an impact (dropped on the floor)
One of the problems of tuning the transducer system to optimum efficiency is that the system can end up being close to the edge of pole-over stability. Maximum efficiency involves maximum armature movement and thus maximum variations in gap width. For very small gap settings, a high signal level drive to the coils could risk causing pole-over. More significantly at maximum efficiency the transducers are more prone to pole-over when dropped or knocked.
It is usual practice to tune the system so that it is backed off from optimum efficiency. This usually involves setting the mechanical restoring forces higher than the magnetic force across the gap. The result is a more stable, but reduced sensitivity transducer.
Figure 28.16
The magnetic forces are set by the level of flux in the permanent magnet. Hence, setting the correct hysteresis curve for the magnet can be used to set the efficiency of the system.
28.12 Setting the transducer sensitivity
Putting it simply, setting the sensitivity of a rocking armature transducer requires energizing the coils with a sinusoidal signal and monitoring the acoustic output from the diaphragm. The process then is to reflux the permanent magnet in situ until the desired acoustic output is achieved.
Rocking armature transducers come in several sizes, a particular size common to a lot of telephone handsets is known as the 3T size transducer.
This transducer is normally set to have an earphone sensitivity at 1kHz of 118dBSPL/mW. Thus the sensitivity is defined for a 1mW signal input power at 1kHz. This sensitivity is always the same regardless of the impedance of the transducer, the impedance is set by the number of coil turns. Typical impedances at 1kHz are 8Ω, 16Ω, 24Ω, 32Ω, 150Ω, 300Ω, 600Ω, 1kΩ, 2k4Ω.
The sensitivity is measured as part of a frequency response, which is done by locating the earphone into a suitable coupler and placing it on to an artificial ear (typically a Racal B4 coupler and a B&K 4153 artificial ear). The earphone coils can be driven with a constant power input and typically swept over a 100Hz to 10kHz frequency band.
Figure 28.17
A typical frequency response profile for a transducer measured in this way is shown below in figure 28.18 below.
Figure 28.18
As a microphone the sensitivity at 1kHz for a 300Ω transducer is normally
-56dBre1V/Pa (open circuit load).
More detail about the acoustic properties of rocking armature transducers will be covered in another lecture.
End of Lecture