digital audio recorder olympus DIY microphone

MICROPHONES and RECORDERS	homepage
Olympus DS-40 Audio recorder and player

		This Olympus recorder, at the highest quality setting, saves files in WMA format at 128kbps with a sampling frequency of 44.1 kHz, and has a frequency response of 50 to 19kHz. It will play WMA and MP3 files, has a 512MB memory, and has a small speaker on the back of the unit. This is a neat, good quality product. http://www.olympus-europa.com/consumer/2581_DS-40.htm The recorder has a plug-in stereo microphone which is pictured here. The mis are fed 2 volts from the recorder, via the jack plug. This microphone has a fairly limited low frequency response. The low frequency roll-off of this microphone is similar to that obtainable with the selectable low frequency cut option on the recorder. A particular colouration in the sound of this microphone (characterised by a strong response to sibilants) is the result of the casing and the internal positioning of the electret capsules. This colouration disappears when the casing is removed. Perhaps the designers have deliberately engineered a particular response to emphasise voice characteristics. The casing is effective in reducing wind noise. The microphone setup is shown here, with, and without the casing. The circuit board has been moved a little out of its locating guides. There appears to be a simple parallel RC network across the microphone positive and ground. The capsules are set at 90° to each other and the microphone can be pionted towards a sound source whilst retaining some stereo separation. Above, the frequency response graph published by a supplier of this capsule. Left, the microphone capsule. It looks very similar to the Panasonic WM-55A capsule. The identification around the side reads: "CZ034GU". CZ034GU Specifications Audio sample of the Olympus voiceguide filesWMA 48kbps Sample recording: Sounds in the garden with a light breeze Original 128 kbps 44.1kHz WMA recording using spaced Panasonic WM-61A electret capsules. I have replaced the inbuilt mono microphone with a Panasonic WM-61A capsule. This capsule is the same 6mm diameter as the original, and slightly deeper. The original capsule was a 'CZ034F' and seems very similar to the Panasonic WM-62A. This new capsule is more sensitive, and has a lower noise figure. I found that the capsule position does not correspond well with the aperture on the top of the recorder casing (see animation, left). This mismatch spoils the sound, and transmits too much case handling noise to the microphone. The animation here shows the offset. The magenta dot indicates the approximate position and size of the aperture of the capsule. I have since made a small modification to the top plate to make an easier path for sound reaching the mic.

Other Audio Recorders
The recorder shown at the left here is the Ripflash Pro. It was marketed under names such as Nextway and DCube. It has a 128MB memory and records in MP3 format at bit rates up to 224kbps. It does not have a mic input, so an external preamp is necessary, and I use the homemade preamp shown on this page. It makes excellent recordings, and in 2006, they were available at bargain prices. I bought several of these, as they were so inexpensive.

Shown right is the is the Sharp MS-200 minidisc recorder. It does a good job, but is relatively heavy and bulky. The mechanism makes noise, and so one can't directly plug in a microphone without picking up some of this noise. It uses a custom lithium Ion battery. I prefer a standard AA or AAA system because of its flexibility.

Home constructed stereo T-mic May 2007

I designed and built this microphone specifically for the Olympus recorder shown above. I wanted to use this in place of the Olympus mic because its plastic casing causes over-emphasis at certain frequencies.

My microphone uses two directional Panasonic WM-55a electret capsules. The mass of these capsules is low and I took particular care to support the capsules very lightly in order to minimise handling noise. The capsules in this design face 180° apart from each other, and so there is the maximum of stereo separation. The steel mesh is fairly light, but should be adequate if the microphone is handled with care. It is backed with the black fibrous paper-like material commonly found in microphones.

The body is constructed from nickel plated brass tubing. Small rubber lugs were added to locate with the depressions in the top of the Olympus recorder, to ensure correct alignment when plugged into the recorder. The 3.5mm plug is offset on the body to align with the position of the mic socket on the Olympus recorder.

Home constructed stereo microphone

This microphone employs two 6mm diameter Panasonic WM-61A omni directional electret capsules fitted into each end of a 390mm aluminium tube. The capsules are very sensitive, and have a fairly low noise figure. To reduce resonance, the tube is packed with putty. The rubber rings on the tube help to locate foam windshields (shown right), and are beneficial when laying the microphone down. The design relies on powering from a recorder. Handling noise is acceptably low, and the microphone produce a nice quality of sound.

Sounds in the garden with a light breeze Sample audio.128 kbps 44.1kHz WMA. Recorded with the Olympus DS-40

The stereo image is created by the differential between the outputs of the two capsules. Amplitude differences arise from the capsules' mild directional characteristics at higher frequencies, and from the relative proximity of sound sources. Phase differences occur where sound waves experience a time difference in reaching each capsule. The degree of phase shift is determined by the spacing of the capsules and the sound's angle of approach. This design provides a sense of spaciousness - best reproduced with earphones - which the T mic design shown above does not.

Outdoor microphones, good for birdsong and ambient noise from the great outdoors
I have set two Panasonic WM-61A electret microphone capsules into the wall of my house. The front of the capsules is more or less level with the surface of the walls, and held in place with soft putty. The capsules are tiny, so I did not need to drill much of a hole. I covered them with metal mesh and foam earpads from earphones, and this offers some protection from wind and the elements in general. I'm not sure just how well these capsules stand up to moisture, but I've had them there for a while now, and they're still working fine. They are very sensitive, low noise and give a high output.

The capsules are set about 1ft apart, on adjacent walls, and this provides excellent stereo sound. There is the time / phase difference in the signals from each capsule arising from the spacing between them. In addition, the positioning lends directional properties to the capsules because of the shielding effect of the corner of the building. Unlike more exposed elements, wind noise is not a great problem in this setup.

Microphone self-noise and sensitivity

The Panasonic WM-61A capsules have a S/N ratio of 62dB. I have experimented using these capsules with the Sharp MD-MS200 minidisc recorder and the Ripflash Pro recorder, and the results are very similar to the Olympus DS-40 in terms of noise when recording low level sounds. The capsules work well with the Olympus. The combination is highly sensitive (too high at times) and the overall noise level is low enough for recording low level sounds.

The above recorders cannot take full advantage of a very low noise microphone unless it also has a high output. One would need a very low noise preamp in order to take full advantage of a very low noise microphone. The Panasonic WM-55A capsules that I used for the T mic design (shown above) have a quoted noise figure close to that of the WM-61A, but they have a lower sensitivity figure of -47dB. They produce a recording which seems poorer in terms of noise than when using the WM-61A capsule.

Interesting material on using electret mics in parallel to increase output level and improve S/N ratio
More on this subject from Rane
Comparison of minidisc and microphone noise

Microphone Preamp Design December 2004
I needed a very small preamp which would operate from low voltage small batteries. I found nothing suitable available commercially, so I had to set about designing and building a preamp device myself. I wanted to use this in conjunction with a small digital audio recorder which had no microphone input. This unit was designed for use in conjunction with the Ripflash Plus. That recorder was capable of making high quality recordings at up to 224kb/s at a 44kHz sampling rate using the line input. I no longer have that recorder, and I replaced it with the Ripflash Pro (or DCube / Nextway 310), which, again, has only a line input. After much research, I identified the TEXAS INSTRUMENTS TLV2461 and a simple preamp circuit on their data sheet. I based my design on this circuit. changing the feedback component values to provide a gain level of 35dB. This IC will operate with a supply down to 2.8V, with low current drain, and has a low noise figure. I established that three AG12, 386 button cells would supply 2mA for a useful length of time. A bias voltage is supplied at the input for electret microphones. I fabricated a case by cutting apart a plastic box and reassembling one with final dimensions about 45X15X13mm, completing it after the components were assembled. I decided not to use a circuit board, and instead position each component to make best use of the limited space available, and wire the components point to point. I had some difficulty obtaining the tiny 0.25 watt resistors, and proceeded with the larger size, giving myself some extra problems fitting and soldering the components in such a small space. The series combination of 1kohm and 4.7mF in the feedback should give some low frequency roll-off. The -6dB point is about 33Hz. The device seems to work well, coping with a wide range of levels despite there being no adjustable gain level anywhere in the system. The unit was designed specifically to plug directly into the Ripflash Plus, but it can equally well be used in-line. Right, my preamp design: the unit used in-line, and the preamp during construction.

Microphone Preamp Design December 2004

I needed a very small preamp which would operate from low voltage small batteries. I found nothing suitable available commercially, so I had to set about designing and building a preamp device myself. I wanted to use this in conjunction with a small digital audio recorder which had no microphone input. This unit was designed for use in conjunction with the Ripflash Plus. That recorder was capable of making high quality recordings at up to 224kb/s at a 44kHz sampling rate using the line input. I no longer have that recorder, and I replaced it with the Ripflash Pro (or DCube / Nextway 310), which, again, has only a line input.

After much research, I identified the TEXAS INSTRUMENTS TLV2461 and a simple preamp circuit on their data sheet. I based my design on this circuit. changing the feedback component values to provide a gain level of 35dB. This IC will operate with a supply down to 2.8V, with low current drain, and has a low noise figure. I established that three AG12, 386 button cells would supply 2mA for a useful length of time. A bias voltage is supplied at the input for electret microphones.

I fabricated a case by cutting apart a plastic box and reassembling one with final dimensions about 45X15X13mm, completing it after the components were assembled. I decided not to use a circuit board, and instead position each component to make best use of the limited space available, and wire the components point to point. I had some difficulty obtaining the tiny 0.25 watt resistors, and proceeded with the larger size, giving myself some extra problems fitting and soldering the components in such a small space. The series combination of 1kohm and 4.7mF in the feedback should give some low frequency roll-off. The -6dB point is about 33Hz.

The device seems to work well, coping with a wide range of levels despite there being no adjustable gain level anywhere in the system. The unit was designed specifically to plug directly into the Ripflash Plus, but it can equally well be used in-line.

Right, my preamp design: the unit used in-line, and the preamp during construction.

TLV 2460 Data sheet: http://focus.ti.com/lit/ds/symlink/tlv2461-q1.pdf	Left, a circuit design by Texas Instruments. I used the TLV2461, which is the same,, but with no low power shutdown option. Below, my design, showing one channel.

Panasonic NV-GS60 camcorder ZOOM MICROPHONE December 2008

The Panasonic literature describes this as a 'zoom mic'. It is said to follow the zoom lens in its angle of acceptance. This is a stereo microphone which appears to change its character to a mono cardioid pattern as the focal length of the lens increases. There seems to also be a general increase in sensitivity - at least in the forward direction, as the lens zooms forward. I would think that the effect is produced electronically, rather than any physical or mechanical change to the microphone capsules setup. I imagine the mono cardioid effect is got by comparing the phase and amplitude of signals from the left and right microphones, and filtering out signals which are different. Sounds from dead ahead will be identical in each channel, and these can be extracted. I have come across an integrated circuit designed for camcorder microphones.
I wonder if this is the sort of arrangement used by Panasonic.

I am curious about how the camcorder creates a directional stereo signal with microphone capsules set into the case as the picture shows. I don't want to start opening up the camera - it's too new. I am tempted, however, because I would like to create a connection for an external microphone. First impressions are that this mic is fairly effective and sensitive with distant sources. The gain increases at longer focal lengths. Mechanical noise pickup from the motor is very low, and generally the audio quality is very good, although stereo separation is limited.

EM150 Microphone capsule April 2008

I await delivery of these capsules from the United States.
http://www.frogloggers.com/BT%20EM158.htm
http://www.primomic.de/ausdruck.php?id=62&en=1
These have a sensitivity of -32dB. and a S/N ratio of 75dB. I should be able to take advantage of this low self-noise because of the very high sensitivity of the microphones.

Find this interesting? email haldanes@ntlworld.com

homepage