[phpBB Debug] PHP Warning: in file /mnt/forum-shared/ext/sitesplat/BBCore/event/listener.php on line 70: in_array() expects parameter 2 to be array, boolean given
Wireless microphone - SparkFun Electronics

SparkFun Forums 

Where electronics enthusiasts find answers.

All things pertaining to wireless and RF links
By Mes

What would be the best configuration to implement a wireless microphone data gathering system for industrial use? I'm looking for cheap price, not HIFI.

I'm gathering machine sound data using a computer, digitizer board and several microphones, but the wires are a problem.


By Guest
Hi Mes,

Depends on what bitrates and range you have to transmit. If you say 'cheap', does that mean that development of dedicated hardware is not possible? If not possible, just build a wireless IEEE802.11 network with standard laptops / network cards. If you can, think of the architecture: make a wireless microphone network, or make a box that the microphone wires run to and go through the air from there.

The noise and steel in the industrial environment generally limit the obtainable range or data rate considerably (factor of 2..3 loss of range).

Keep in mind that wireless means: an unreliable physical layer in the network. Packets get lost, can be disturbed, can be listened to by unauthorised parties, can be 'faked', etc.

If the application is 'machine monitoring', this may work. If you want to use the data for a real-time feedback loop, think well about what happens if measurement data are NOT (correctly) received.

Best regards,

By Guest
Thanks for your interest Marco!

I'm sampling at least 20kHz. Development of dedicated hardware is possible, but the design should be very simple in the microphone end. The range would be under 10 meters (30 feet?) from a computer. The problem is that it should run several years without maintenance. Security is no problem. Do you think that sending in analog form would be a bad idea?

Best regards,

By Guest
It depends on the accuracy (resolution, linearity, signal/noise) that you need.

If you want to do run FFTs on the signals, you will need a very clean signal. That is: the cleaner the signal, the more accurate (OR the faster) you can monitor the machine condition.

Numbers like 50 dB (0.3%, roughly 8 bits/sample) signal / noise ratio, and 1% distortion (non-linearity) are common in less expensive FM systems.

If that is not good enough, I would go digital from the microphone on. I think it would fit well within about 2 square inches of pcb area.

No maintenance? Just use good components (no adjustable coils, potmeters, and electrolytic capacitors) and you're safe.

Best regards,