Boundary-layer dish microphone

Project ideas for a directional microphone assembly.

Introduction.
The Dish.
The Microphone capsule.
The Power supply.
The Pre-amplifier.
Positioning the capsule.
Handling-noise.
The stereo dish.
Further developments.
1. Better capacitors.
2. Dealing with wind-noise.
Component suppliers.
References.
Further reading.

Introduction

Being something of a tightwad, I'm always trying to do things as cheaply as possible. So when I wanted to do some wildlife recording it seemed only natural (to me at least) that I should attempt to do it myself from scratch.

I wanted to have a microphone with good directional and gain characteristics. A parabolic microphone would have been ideal, but I really didn't want the complexity of producing an accurate dish profile. So I started to look around for alternatives.

The most likely candidate was a dish shaped microphone assembly using the boundary-layer principle, which I first became aware of from The PZM Boundary Booklet: A Basic Primer and Experimenter's Guide (PZM® Crown International, Inc.). One section of the booklet describes the use of a boundary-layer dish, among many others. This booklet contains many more suggestions for boundary-layer microphone assemblies which are eminently suitable as projects for the home constructor.

Whilst not giving the ultimate in either gain or sound quality, the dish microphone project nevertheless gives a useful amount of both. At the same time this project requires very little outlay, uses readily available materials and is simple to construct.

Having built the dish-microphone, and also a stereo version described later, I can say that both performed well for me, and proved to be very interesting projects.

Back to Contents.

The Dish

The ideal material to use as the boundary material is Perspex® (Plexiglas®), but as I said I'm tight so I went with the cheapest option... a polypropylene mixing bowl, which can be bought in any High Street. The problem with these things is that they are not truly dished, and often have both a wide, flat base and steeply sloping sides. The only thing to do if you have a go at this yourself is to hunt round and find the one with the best dish profile. It shouldn't matter too much for the purposes of experimenting but a good profile should ensure that your project doesn't look too much like a refugee from the kitchen cupboard.

You can use any size of bowl that you can lay hands on, but the larger the bowl the lower the frequency for which the microphone remains directional. If you can obtain something larger than a mixing bowl that can be used in this application then by all means use it, but bear in mind that a very large dish will be rather cumbersome and heavy if used 'in the field'.

One alternative that springs to mind as an alternative is one of those Perspex® skylights that can be found in some public buildings. The larger sizes may well be too heavy to be practical, but if you can find one of appropriate size, then this would probably give superior performance to the humble mixing bowl. I'm sure there are many such alternatives available, all that is required is that you apply a little lateral thinking to the selection process.

fig.1 Boundary-layer dish microphone
D

Key to fig.1

A: Domestic polypropylene mixing bowl
B: Microphone power-supply
C: Electret microphone capsule
D: Optional hand-grip

Back to Contents.

The Microphone capsule

The microphone capsule that you use for this project needs to be an electret unit with an omnidirectional polar pattern. The particular ones I used for this project were generic cassette-recorder replacement microphones which I obtained from Maplin , but are likely to be available from any electronics retailer or repair shop. These are extremely cheap and really only intended for voice applications, having a limited response in line with that requirement. Nevertheless they do give adequate performance for assessment purposes. I wouldn't advise using anything too expensive at this stage, because you are likely to be fiddling with the design quite a bit. A more expensive capsule is almost certain to sound better, and give improved gain characteristics, so you could substitute a higher performance capsule once you are sure that the project is going to be appropriate for your requirements.

Back to Contents.

The Power supply

Being electret microphones these require some form of power supply, the recommended minimum being 1.5V. With this project I chose to use a 9V battery supply for two main reasons; firstly, the higher voltage allows the capsule to operate with reduced noise-floor, and secondly, these small batteries are easy to obtain and install with many mounting options available. In addition, I planned to experiment with a stereo version of the dish which meant that I could power each capsule at it's optimum 4.5VDC. I found no problems whatever with running these capsules from a 9V battery and would recommend that you do the same, provided that the capsule you select is rated to operate at this voltage.

The circuit used to bias the capsule from the battery was a generic circuit (taken from the capsule's catalogue page) representing the minimum requirement to provide correct biasing.

Fig.2 Circuit diagram of generic electret microphone capsule biasing
D

Typical specification of generic electret microphone capsule

Response: omnidirectional
Frequency response: 50 Hz to 13 kHz
Sensitivity: -60 dB ±3 dB (0 dB = 1 V/mbar at 1 kHz, Vcc = 4.5 V, RL = 1 k ohm)
Impedance: 1k ohm max.
S/N ratio: < 40 dB
Sound pressure level: 120dB max.
Power supply: 1.5 V to 10 V DC
Recommended voltage: 4.5 V (optimum performance)
Current drain: < 0.5 mA

Back to Contents.

The Pre-amplifier

Confession time I'm afraid. I didn't construct a pre-amp for this project, relying instead on the built in pre-amps of my DAT recorder.

So why bother listing it as a topic if I didn't actually build one? Simply put, the design of a pre-amplifier with an appropriately large signal-to-noise ratio is beyond the scope of this article. My only aim is to put forward a few ideas and to highlight areas where the project may be improved.

Indeed, there is considerable scope for improvement to the performance of the project by building a dedicated pre-amp. By combining the biasing and a low-noise pre-amplifier adjacent to the capsule itself there is likely to be significant improvement in the signal-to-noise ratio and in the overall frequency response.

Back to Contents.

Positioning the capsule

When positioning the capsule bear in mind that a central location should be avoided as this will produce the largest distortions of the frequency response. Diffraction from the edges of the dish will be identical from any part of the edge, and so the interference will combine to produce more exaggerated effects.

If the capsule is positioned off-centre then the coincidence of distortions will be reduced, and the overall frequency response will be less affected.

Back to Contents.

Handling-noise

If you use a polypropylene mixing bowl as a boundary material and use the microphone as a hand-held device then handling-noise is likely to be a problem due to the transmission of vibration from the hand-grip. Other materials may prove to be better in this respect than polypropylene.

It is possible to reduce the level of vibration by the use of various materials to damp the hand-grip, but the effect of this damping is unlikely to be wholly effective. Hence, when used with a hand-grip, care must be taken whilst recording to avoid excessive handling-noise.

A better alternative is to arrange for the dish to be mounted on a stand of some kind, so that the operator does not have to be in physical contact with it. Whilst this may not be practical in all circumstances it is the simplest and most effective method of eliminating handling noise.

The most complicated arrangement is to incorporate some kind of shock-absorbing mount integral to the hand-grip. This kind of arrangement is potentially most effective for the usage of this kind of microphone, but producing an effective shock-mount for a device of this size is likely to be quite difficult.

In my own project I used a solid wooden hand-grip, covered in foam tape (as used on the handlebars of road-racing cycles). Handling-noise for this arrangement was quite high, and to minimise noise a light hold of the hand-grip was required. With care this was adequate for most situations except those where focusing in a particular direction for an extended period, in which case a microphone stand was used.

Back to Contents.

The stereo dish

As I mentioned earlier, I also constructed a stereo version of the dish microphone. This used the same components as the mono version (but with two microphone capsules and bias circuits, naturally).

As you might expect simply arranging two capsules on the base of the dish gives little stereo effect. However the use of an acoustic screen,or baffle, between the capsules produces a much improved stereo performance.

The baffle material that I used was a small sheet of Perspex® that I had just lying around. This was cut to fit the internal profile of the dish, and then shaped to give a smooth curve on it's outer edge.

Fig.3 Stereo dish microphone
D

Those of you with experience of stereo microphone assemblies will be able to see that the arrangement of microphone capsules shown effectively form a coincident-pair, thus ensuring good monophonic compatibility for on-axis sounds. This is offset to certain degree by the time delay between the channels for off-axis sound sources, which is due to diffraction around the central baffle.

The performance of this stereo arrangement was found to be better than initially expected. One unexpected bonus was a subtle and attractive binaural effect when monitoring through headphones.

Back to Contents.

Further developments

Better capacitors

One possibility for improving the distortion characteristics of the microphone is to use a polypropylene capacitor in the biasing circuit. Though there are lower distortion capacitors available, I feel that polypropylene provides the biggest bang-per-buck in most audio projects.

The distortion characteristics of various capacitor types is a well documented area, though I have no idea if this information is available on the web at the moment. So if you know of a site carrying this data then please drop me a line so that I can point folks in the right direction.

If you have a go at building a low-noise pre-amp for the project then it might be a good idea to use good quality capacitors for that too, particularly for critical positions in the circuit.

If you are wondering what constitutes a 'critical position' then you'll find lots of good advice at one of the many Hi-Fi related sites, but please don't spend a fortune on them unless you are sure that the performance gain is worth the cost. After all this is a real 'kitchen-table' project, and we tightwads need to save our cash.

Dealing with wind-noise

If you happen to live somewhere where the weather is always fine, and the wind never does more than gently brush through the tree's then you can skip this section. This bit is only for the rest of us poor unfortunates for whom the weather turns from calm to gale-force as soon as we step outside. Hopefully you'll be lucky and wind-noise will never plague you, but if it does then you might as well be prepared for it.

Boundary-layer microphones are not overly sensitive to wind-noise, but in the case of the dish microphone, at least, it can sometime cause the odd problem. Overcoming it is not simply a matter of forcing one of those foam wind-shield things over the dish. I doubt that you'll be able to find one big enough in any case.

For my dish microphone I applied the time-honoured rule of tight-fistedness and simply 'borrowed' a pair of woollen tights (I think they're called panty-hose in the U.S.). In use this arrangement, though crude, worked very well. Problems with wind-noise were reduced quite noticeably. At the same time little effect on the frequency response was noted.

If you would rather do a more professional job, then it may be possible to obtain a sheet of open cell foam that can be used. Be aware that most of the foam that is cheaply available is really too acoustically dense for this job. If you wish to experiment with different types of foam, or indeed any other material, for use as a wind-shield, always test it thoroughly to ensure that it isn't attenuating higher frequencies to an unacceptable degree.

Back to Contents.

Component Suppliers

Maplin Electronics.
Tandy Corporation. ( RadioShack. )

Back to Contents.

References

The PZM Boundary Booklet... is reprinted in
Microphone Manual (Design and Application)
by David Miles Huber
Focal Press (Butterworth-Heinemann)
ISBN 0-240-80141-5

Back to Contents.