Monday, 17 December 2018

Test your mic and avoid a wasted session.

by Peter Flory, Peersonic Limited 

It makes sense to test your bat recorder or bat detector equipment prior to use. Microphones can degrade with time, the main enemy being humidity, or worse still total saturation.

So far, I have found that over a 5 year period I have no customers with failed microphones where they are using handheld equipment. I am sure they are all still working fine, but worry that they might be degrading and the user is unaware.
Static equipment left out in the weather is a different matter. I have observed totally dead mic's, and again more concerning some that have started to lose the top end response after a year or so of use.

So here are some thoughts on how to test your equipment and be sure it is going to be effective, with cost in mind.

Jangling Keys:
If you are a casual user why not just jangle keys. A detector will give you a response, but you just don't know how well it is operating over the frequency range.
Jangling keys is better if you have a recorder, the stimulus can be recorded and the response viewed as a spectrogram.

Here is a spectrogram of key jangling,I tried with different bunches of keys and found that on the whole the stimulus was covering the spectrum well up to about 110 kHz. Although it is not a repeatable signal. 


Fixed Frequency or frequency sweep:
A better test would be to use a definite fixed frequency, generated from a signal generator, a specific sine wave , or series of waves played back through a tweeter can give a measurable stimulus. 
Use a sine wave rather than a square wave as the latter contains many harmonics and it would be nice to observe the response at each frequency step as far as possible.
Signal generators are standard electronic test equipment. They are usually around the £150-£200 area, but there are some lower cost kits too that might be as low as £50.

A low cost China manufactured sig gen at a bit less than £150.

Next you need to select a speaker/ tweeter. These are never designed to cover the full bat spectrum, so some trials and selections are required. However they are not costly. There are voice coil speakers which are fairly level over 20kHz to 90kHz, Or pietzo speakers that will go much higher, both will suffer from features over the band so this needs to be considered. - There is no perfect flat response.

Now it is possible to chose a selection of signals, at predefinied amplitude ( volume ) and set up your own regular test pattern.
Just test and record, and observe and trends in the recordings that suggest deterioration over time.
Monthly perhaps.

Calibration:
Having found that a microphone is deteriorating, typically at the higher frequencies. You might want to calibrate it. You can't.
A mic that is on the way out is best consigned to the dustbin. This should not be confused with the process of matching a mic to a recorder by equalisation. That is fine, but not a great idea if the mic is becoming unreliable.

What to do with a sick mic:
Manufacturers vary, if you are able to replace the internal mic element this should be cheap, if the mic is a unit, you will need complete replacement. So from the lower end ( replacing the insert) to replacing the unit, the cost would be anything from £12.00 to about £300.

Specific bat Ultrasound mic testers:
Rather than go for a test equipment based arrangement, you could do better by obtaining an ultrasonic mic tester as a unit, speaker included or integrated, with features specified in the suppliers data sheet or users manual. Should make life easier.

Some testers emit a spot frequency, 40 kHz, 60 kHz. This will certainly show something, but what about at 125kHz? Or the various points in between? Best thing then is to emit a series of stepped frequencies. Also a chirrup is useful as it gives a fast changing signal and your spectrogram will show up some missing points perhaps.
Stepped frequencies will allow you to check your heterodyne or FD equipment at each point, a full range of patterns is even more revealing for recording FS gear.

At Peersonic I have developed a tester that emits single steps over the range, plus patterns, and for fun some emulated familiar bat sounds.

Speaking to my competitors I find there are fixed frequency testers, and I note the Titley "chirp" equipment, which is similar , but moves on to offer a complete test bench that ensures the speaker and receiver are at a fixed distance from each other.


From left to right: a Peersonic multi-mode mic tester, an Elekon dual-mode tester and a Titley Anabat tester 

Test Environment:
If you are expecting to get highly repeatable calibrated measurements you had better hire a sound chamber, and probably some expensive transducer equipment. But, just testing if it's ok is really good enough for most of us. Here is some advice -

Try to keep the following the same each time for best results.

Use the same room.
Try to ensure that the furniture is in much the same position each time.
Do not stand next to the speaker, or in front, best to operate from behind the setup..
Avoid wearing wooly pullovers as they are absorbers of sound.
If you can, check that there is no electromagnetic interference, (a desktop PC with the sides off tends to transmit all sorts of rubbish).

Or test outside in an open space.

Conclusion:
Testing your mic is a bit like backing up your computer. Unless you are running a schedule, it really only happens when its too late.


No comments:

Post a comment