SparkFun Forums 

Where electronics enthusiasts find answers.

General project discussion / help
Did you make a robotic coffee pot which implements HTCPCP and decafs unauthorized users? Show it off here!
By Ving
#199324
Hi, I need to measure the speed of sound through solids like wood etc. Which components do I need? What is the cheapest way to do this? Sample thickness varies from 1/2 inch to 12 inches.
By jbreizh
#199336
hum interesting subjet, so i read this wikipedia page : https://en.wikipedia.org/wiki/Speed_of_ ... eous_media
It seems a non trivial problem, so without some indication of what you want to archive it is not possible to answer.
-form of the sample (rod, 3d form)
-what change between sample (material or just thickness)
-what do you want to make with this mesure (extrapolation to what?)
-how many mesurement (just a few or you need a field instrument)
i think you must think of your final need, then the protocol, then the material, then how to do it cheap. Because if it's just for a few approximate mesure, it could be done in lab with a an oscilloscope and some microphone, but for a field instrument, that's not the same.
By Ving
#199338
Thanks for the reply, I really appreciate it. I'm expecting the sound velocity to be between 1000 and 2000 m/s for the smallest thickness of 1/2 inch (.012 m). It means that my time of flight can be as low as 6e-6. I guess I'm confused as to what will give a sampling rate as small as 1e-6.

Here's more info that you are referring to:
- form of sample is rods and bats (yes 3d)
- material type, thickness and direction changes between measurements
- I'm just exploring these measurements
- I need about 20-30 measurements to begin with. These will eventually be done using a field instrument, but to start with I will use lab.

An oscilloscope feels a bit expensive to buy. I was wondering if there's any cheaper way to do this.. For example if I hooked up ultrasonic transducers to an arduino, would the arduino have enough processing power to do the job? Is the oscilloscope the only way to do this? What else can i attach to the microphones that can process the time delay as small as 1e-6 or better 1e-7.
By jbreizh
#199344
honnestly i don't know ???
I think an old oscilloscope can be find cheap
If i have understand the theorie, sound speed is the same as shock speed in rod. I find the video in youtube https://www.youtube.com/watch?v=WfzdeUw8mxY&t=5s. But i don't know if it can be use for shock absorbing and insulating material like wood.
An other remark : using common speaker and microphone seem to be difficult, because how to be sure that the sound travel through your material and not in the surrending air.
sorry i am not very helpfull
By jbreizh
#199345
hum i had maybe a lead for arduino use. You can use piezo knock sensor https://www.efxkits.us/piezoelectric-se ... g-arduino/.
I had you a wonderfull schematic of my thinking :
Image
so you have piezo 1 and 2 both side of your sample, you shock side 1. The piezo 1 mesure the initial shock and the piezo 2 the shock through your material. In theory, one soustraction and you are good. In reality, i don't know if it would work (precision....) and your shock wave will bounce one end to another.
Just an idea, your work to do now.
By Ving
#199346
Hi jbreizh, thank you for your replies, this is really helpful. I hadn't known about the logit device, but it looks like another form of a medium resolution oscilloscope. Seems really handy! The piezo sensor instead of mic sounds a great idea, though I'm still not sure if Windows and Arduino can process such small time spans. Looks like the oscilloscope might be my best bet!

Thanks for taking the time jbreizh, I really appreciate it.
By Valen
#199375
The Arduino's ADC cannot measure voltages as quick as required (microseconds). Few microcontrollers actually can. Iirc ADC clockrate for the Uno/Micro/Nano's is at best 1/26th of the clock rate. And then there is the issue of storing and analysing the data that takes even more time. But even worse their analog response only allow low frequencies (1- 20kHz) to be measured. High frequencies do not come through well enough due to the sampling limitations.

You are indeed better of getting a second hand analog oscilloscope of a few 10s Megahertz. With repeated pulses you can view the response on the screen. Storing the waveform might be a challenge though. Unless you find a way to photograph based on a trigger signal. Otherwise there are USB oscilloscopes with similar bandwidths that allow storing it more easily to a computer.