Record voltage output at 100kHz

Thread Starter

yjoeyj

Joined Dec 10, 2022
5
I'm interested in the frequency spectrum of the voltage output of a vibration sensor (Kemet VS-BV203-B) at frequencies up to about 100 kHz (so either the fft or a timeseries). Unfortunately I don't have the budget for very expensive loggers/analyzers and was wondering if anyone has any suggestions in how to obtain this for a budget below say 750?
The vibration sensor has three wires, power at 3.3V, GND and signal output, which is biased by half the power voltage. I though perhaps an audio recorder may work? Are there any oscilloscopes or spectral analyzers for that budget that work too?
 

MrSalts

Joined Apr 2, 2020
2,767
A 15 second investment into searching and reading the datasheet for the vibration sensor confirmed my thoughts that this sensor is rated at 15kHz max so, no circuit will help your goal of 100kHz.
 

nsaspook

Joined Aug 27, 2009
12,998
A 15 second investment into searching and reading the datasheet for the vibration sensor confirmed my thoughts that this sensor is rated at 15kHz max so, no circuit will help your goal of 100kHz.
It won't brickwall at 15kHz so I would expect some soft of nonlinear signal response to at least the resonant frequency of the sensor at 30kHz.
 

Thread Starter

yjoeyj

Joined Dec 10, 2022
5
For the audio range (to ~20,000Hz) , a PC sound card should work as a cheap capture device if the signal is AC coupled using a capacitor to isolate the DC bias..

https://www.edn.com/low-cost-vibration-analysis-using-pc-audio-and-diy-motion-sensors/
Many thanks for response and suggestion. I'm not that knowledgeable in the field, but I'll have a look at how to remove the DC bias using a capacitor. Once the bias is removed, can use a simple 3.5mm jack to connect it to a computer?
 

Thread Starter

yjoeyj

Joined Dec 10, 2022
5
A 15 second investment into searching and reading the datasheet for the vibration sensor confirmed my thoughts that this sensor is rated at 15kHz max so, no circuit will help your goal of 100kHz.
Indeed. I should have been more clear with my details. I plan to use the aforementioned sensor for testing, but will purchase a different sensor (with same wiring) that goes to 384kHz when I've got it working.
 

MrSalts

Joined Apr 2, 2020
2,767
Do you only want frequency or are you looking for amplitude as well? And record or just respond when a certain frequency range is hit? More detail would be helpful.
 

nsaspook

Joined Aug 27, 2009
12,998
Many thanks for response and suggestion. I'm not that knowledgeable in the field, but I'll have a look at how to remove the DC bias using a capacitor. Once the bias is removed, can use a simple 3.5mm jack to connect it to a computer?
It's likely the sound card already has a DC blocking capacitor but yes, a normal audio jack would work but there might be an amplification requirement to get the needed signal sensitivity needed for testing.
 

ronsimpson

Joined Oct 7, 2019
2,954
I don't see how the sensor will have an output to 100khz. See data sheet below.
I use a digital scope to record waveforms and them, move them to a PC.
The sound card should work for 10 to 15khz. You might have a problem at 10hz. That can be fixed.
1670772598871.png
 

Thread Starter

yjoeyj

Joined Dec 10, 2022
5
It's likely the sound card already has a DC blocking capacitor but yes, a normal audio jack would work but there might be an amplification requirement to get the needed signal sensitivity needed for testing.
Cool, I should have a Creative Sound Blaser Play!3 lying around somewhere.
 

Thread Starter

yjoeyj

Joined Dec 10, 2022
5
Do you only want frequency or are you looking for amplitude as well? And record or just respond when a certain frequency range is hit? More detail would be helpful.
Amplitude as well. I can calculate it myself if I have the full time series of the signal, but I think some components may actually give the amplitude directly, such as the MSEGEQ7 (Graphic Equalizer Display Filter). Unfortunately, the output frequencies are quite limited for that one.
 

MrSalts

Joined Apr 2, 2020
2,767
Amplitude as well. I can calculate it myself if I have the full time series of the signal, but I think some components may actually give the amplitude directly, such as the MSEGEQ7 (Graphic Equalizer Display Filter). Unfortunately, the output frequencies are quite limited for that one.
A vfc32 or other frequency to voltage converter can give you some very accurate voltages for a dc to 100kHz signal but no direct measure of amplitude. If you go this route, you'll need a separate sensor circuit to monitor a amplitude.
 

Alec_t

Joined Sep 17, 2013
14,263
Once the bias is removed, can use a simple 3.5mm jack to connect it to a computer?
Cool, I should have a Creative Sound Blaser Play!3 lying around somewhere.
When I measured the frequency response of a Creative Sound Blaster sound card some years ago, I found it dropped off very steeply above 20kHz. I doubt things have changed in the intervening years.
 

MrSalts

Joined Apr 2, 2020
2,767
When I measured the frequency response of a Creative Sound Blaster sound card some years ago, I found it dropped off very steeply above 20kHz. I doubt things have changed in the intervening years.
If I'd design a sound card, I would filter above 20kHz. Then I'd sell a version without the filter and call it a DAQ board and charge more. Seems to be the norm.
 

ronsimpson

Joined Oct 7, 2019
2,954
I found it dropped off very steeply above 20kHz
The sensor's top end is 15khz so having a "brick wall" at 20khz is not a problem.
If you move to high end sound cards, they have different sampling rates. 44.1kHz, 48kHz, and 96kHz. I have some digital recorders that can be set to 96khz.

There are many audio effects software what analyze sound. You can take the audio file and write you own software. Using an audio card to record data is common.
 
Top