Hi Chris
This reminds me of a sensor I built years ago with piezoelectric disks. It looked like a giant childs "jack".It was constructed of acrylic sheets and tubing.The center was a glue together cube. It had 6 ~1" diameter tubes about 6" long which radiated out from the center in the 6 perpendicular directions.The end of each of the 6 tubes had a glued on 1" piezo disk.I filled the whole thing with glyserine. I was going to connect the opposing disks in a half wheatstone bridge. The thought was I would have a triaxial accelerometer by measuring the pressure diferences from one side of the liquid mass to the other. When I found out what the response was from a radio shack mag vs frequency plot for the disk I thought I would measure no low frequencies and put the contraption aside. Maybe I was incorrect and had chosen the wrong disk. I can still blow the dust off of it.:)
Regards
Barry
ChrisAtUpw@....... wrote:
You can get greatly extended LF range and reduced noise using a 2" electret piezo speaker as the sensor. They have a capacity of ~65 nF, so with a 47 M Ohm input impedance, you can get a Tc of 3 sec, corresponding to a period of 19 sec, 0.05Hz.
The straight sided 1L "food" flasks can be modified as low temperature drift capacities. I use 100 M Ohm to get to 40 sec period with an OPA604 opamp.
It is an advantage to fill the capacity with the lowest density polyurethane foam. This ensures a near isothermal pressure response as opposed to an adiabatic / slump response and gives about 40% more output. The pressure / temperature relaxation time constant of this system seems to be about 0.1 sec.
The thermal drift can be reduced if necessary by adding candle wax to the inside of the flask to increase the 'thermal mass'. The thermal time constant of an empty flask for ambient temperature changes may be about 1/2 hr.
You can also use a screw top glass jar with a restriction tube as a simple series low pass RC input filter to limit the pulse / noise response.
You need to use a ~constant disk temperature to avoid LARGE voltage drifts.
As with all sensitive LF microphones, you do need to limit the wind noise, preferably with a loop of porous irrigation hose, maybe 10 to 30 m diameter, with solid cross connecting tubes to the sensor in the centre.
Panasonic Electret Condenser Microphone WM-034DM
The Sellers web address:
Hi-Tech Enterprises, Inc. http://www.videoequipment.com
(Note they are a discontinued line)
You can download the data on these WM-034 microphones from:
http://www.panasonic.com/industrial/components/audio/aud_mic.htm
If you buy the CURRENT Panasoic WM61A, you will get a higher sensitivity. I suspect that the VLF response is likely to be similar. You can also use these for intruder detection. You can't open a window or a door without getting quite a large signal.
There is a comprehensive web page giving details of several types of infrasonic sensor at http://earth.unh.edu/johnson/MICROPHONES/microphone_list.html
This is worth serious attention for anyone interested in infrasound.
Regards,
Chris Chapman
Hi Chris This reminds me of a sensor I built years ago with piezoelectric disks. It looked like a giant childs "jack".It was constructed of acrylic sheets and tubing.The center was a glue together cube. It had 6 ~1" diameter tubes about 6" long which radiated out from the center in the 6 perpendicular directions.The end of each of the 6 tubes had a glued on 1" piezo disk.I filled the whole thing with glyserine. I was going to connect the opposing disks in a half wheatstone bridge. The thought was I would have a triaxial accelerometer by measuring the pressure diferences from one side of the liquid mass to the other. When I found out what the response was from a radio shack mag vs frequency plot for the disk I thought I would measure no low frequencies and put the contraption aside. Maybe I was incorrect and had chosen the wrong disk. I can still blow the dust off of
it.:)
Regards Barry
ChrisAtUpw@....... wrote: You can get greatly extended LF range and reduced noise using a 2" electret piezo speaker as the sensor. They have a capacity of ~65 nF, so with a 47 M Ohm input impedance, you can get a Tc of 3 sec, corresponding to a period of 19 sec, 0.05Hz. The straight sided 1L "food" flasks can be modified as low temperature drift capacities. I use 100 M Ohm to get to 40 sec period with an OPA604 opamp. It is an advantage to fill the capacity with the lowest density polyurethane foam. This ensures a near isothermal pressure response as opposed to an adiabatic / slump
response and gives about 40% more output. The pressure / temperature relaxation time constant of this system seems to be about 0.1 sec. The thermal drift can be reduced if necessary by adding candle wax to the inside of the flask to increase the 'thermal mass'. The thermal time constant of an empty flask for ambient temperature changes may be about 1/2 hr. You can also use a screw top glass jar with a restriction tube as a simple series low pass RC input filter to limit the pulse / noise response. You need to use a ~constant disk temperature to avoid LARGE voltage drifts. As with all sensitive LF microphones, you do need to limit the wind noise, preferably with a loop of porous irrigation hose, maybe 10 to 30 m diameter, with solid cross connecting tubes to the sensor in the centre.
Panasonic Electret Condenser Microphone WM-034DM
The Sellers web address:
Hi-Tech Enterprises, Inc. http://www.videoequipment.com
(Note they are a discontinued line)
You can download the data on these WM-034 microphones from:
http://www.panasonic.com/industrial/components/audio/aud_mic.htm If you buy the CURRENT Panasoic WM61A, you will get a higher sensitivity. I suspect that the VLF response is likely to be similar. You can also use these for intruder detection. You can't open a window or a door without getting quite a large signal.