Microphone Selection, Vehicle Testing

[MachineryWatch]

I will be doing some sound and vibration work on horse trailers soon to compare and contrast the effects of different construction methods. Can any of you NVH guys recommend an appropriate microphone for such an application? I appreciate your help.

http://www.machinerywatch.com

 

[GregLocock]

For vehicle work we generally use type 1 instrumentation, A weighted, as we are interested in frequencies down to 30 Hz, but they are so loud they overshadow everything else (there is a good argument that we should use B weighting).

Do you mean microphone , or SLM, or microphone+preamp?

AEither way have a look at B&K

http://www.bksv.com/1130.asp

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[MachineryWatch]

Greg,

Thanks for the reply. I was thinking I would work with unweighted data on this because we are doing this with the horse's hearing in mind, as opposed to humans. My research shows horses hear higher frequencies than humans and there does not appear to be too much information about which frequencies, or what range of frequencies are more sensitive.

I guess microphone type is more what I was interested in rather than specific model numbers, but that is helpful, too.

Skip Hartman

http://www.machinerywatch.com

 

[Rob45]

Skip:
A-weighting might still be good for horses, since it primarily de-emphasizes the low frequencies.

As to mikes themselves, I'd recommend looking at the Larson-Davis ICP mikes, since you'll save a significant amount of money, and the L-Ds are quite durable. I used mine day-in and day-out, and they're very relaible.
Cabling for them is cheap too.

You aren't in North Texas, by any chance? Lot of horse trailer mfrs around here.

 

[GregLocock]

LD cartridges (ie the diaphragm and grill assembly) seemed to be less robust than the B&K equivalent when we used both.

That may have changed.

Skip, you probably want a 1/2" free field cartridge.
For high frequency work you might get away with type 2 instrumentation, but I don't know how much money that would save you.

You should still use A weighting, as vehicle work is dominated by the low frequency content, unless you can guarantee a S/N ratio of >60 dB throughout your instrumentation chain.

So, have you got an actual sensitivity plot for a horse? What's the upper limit for frequency?




Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[Rob45]

"Paging Mr. Ed, telepone call for Mr. Ed! Please pick up the white courtesy phone!"

But seriously - you're not selling trailers to horses, you're selling trailers to horse OWNERS. So reducing trailer noise is a marketing tool.

You make a sound recording in a typical trailer, then in your trailer, and let them hear the difference as well as measuring the level difference.

 

[Rob45]

And BTW, Type 1 pre-polarized mikes are readily available, if that's worth anything to you.

 

[MachineryWatch]

Thanks guys. I moved away from TX in the late 70s!

I understand that A-weighting may still be a good way to go. I may go ahead and try comparing ICP Type 1 and ICP Type 2 mikes during this test. I have a Type 1 1/2" free field cartridge mike.

I will be making simultaneous vibration measurements and was planning on looking at coherence. I guess I need to think through that process to figure out if there is a way to look at coherence and apply the filtering to the sound measurements.

Skip Hartman

http://www.machinerywatch.com

 

[GregLocock]

Sure, put a sensor on each path or source and then use the partial coherence to establish the coherent energy at the mic due to each path.

Should only take about 8 weeks!

It might work up to 250 hz for road noise.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[MachineryWatch]

I have found some information that shows an average sensitivity plot for horses. It is fairly similar to a human's. The main difference is the extension of the ability to detect sound at frequencies up to about 35 kHz. Sensitivity drops off significantly above 16 kHz.

Skip Hartman

http://www.machinerywatch.com

 

[MachineryWatch]

Any tips on how to mount microphones that I will be using in the horse trailers? Do they need to isolated from floor vibration?

Skip Hartman

http://www.machinerywatch.com

 

[Rob45]

It wouldn't hurt to provide some isolation: Hang the mike(s) from bungee cords?

Or take a block of foam rubber and tape the microphone down so it hangs off the edge - that's what I use for the "sleeper's ear" position in the bunk of a Class 8 truck - then attach the foam to a structure that would let you simulate the horse's ear.
Or whatever part of the horse...

Good luck!

 

[GregLocock]

Absolutely essential to mount them properly. Always listen to your in-vehicle data.

One way is to clamp them in place. You can use a test tube holder as the final clamp, with a bit of rubber sheet in the jaw to prevent scratches and clunks. This tends to give the best results.

The other option is a true compliant mount, using duct tape, foam, elastic bands etc etc.

That can also work, but listen for clonks and so on when you go round corners.

Make sure your leads are secured, make sure the mic position is known and repeatable.


Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[Rob45]

Yes, indeed: one of the very best things you can do is to actually listen to your data.

Between your ears resides one of the best analysis and source identification computers there is.

- R

 

[MachineryWatch]

Thanks for the suggestions. I agree with the advice to listen to the data. A lot of things that we can hear do not seem to show up clearly in time waveforms and spectra. It is amazing what we are capable of when it comes to extracting things from the data with our senses.

Skip Hartman

http://www.machinerywatch.com

 

[Rob45]

Skip:
One more tip is that you might find your greatest noise source is tire noise if the tread pattern on the trailer tires has a uniform pattern or tread block spacing all the way around the tire, so you get a pure tone at the tread-block passing frequency. You can check for this if you find a high level in maybe the 300 - 500 Hz range at say 60 MPH; check the tire revs/mile and measure the block spacing at several places around a tire. If it's the same everywhere, it's likely to be a problem.
Passenger car tires vary this (for fairly obvious reasons), truck tires do not (out of ignorance, I think), and I've no idea what trailer tires do.