Interfacing OPAx387 with with Sensor VS45-H

[newbie_hs]

I am designing an acoustic emission sensor signal chain.

The sensor P/N is VS45-H. The signal chain is given below.

For buffer and amplifier I am planning to use OPAx387.

May I know your thoughts about this. Is OPAx387 is a good choice for this?

I have not decided any P/N for variable gain Amplifier.I think I can go with LMH6514SQ.

The DAQ may be an ADC present in the microcontroller(STM32) or will use a dedicated ADC and it's output will be processed by a uC.

May I know your thoughts about this entire signal chain.

 

[aliarifat794]

OPAx387 is a good choice. Confirm the input signal amplitude and frequency range of the VS45-H sensor to ensure the OPAx387's bandwidth and slew rate are sufficient.

 

[newbie_hs]

I received below information from sensor manufacturer.
"The VS45-H has no electronic inside, i.e. it will output the signal from the piezo inside the sensor. Typical noise value will be around 10 µV, the max. value for typical acoustic emission excitation up to 5V but it can be also higher:"

 

[KlausST]

Hi,

I did a lot of similar designs. And every single one was different.

If you go for low cost .. you would probably omit some stages.
If you go for low distortion .. you would place the BPF at the very beginning of the signal chain (this is to avoid the risk of overdriving the input stages of amplifiers by high frequency spikes)
If you go for low noise you would reduce the stages and make the signal paths low impedance, and put the BPF at the end of the chain.

What frequency range are you targeting?
And what are your requirements regarding, noise, distortion, dynamics ... ?
What information of the signal are you after?

You can´t get the optimum "for everything".

***
My recommendations:
(I´d say the first buffer stage can be omitted)
The first stage should be a high gain amplifier stage, with LPF/BPF included.
* LPF just to get rid of high dV/dt spikes
* BPF if you additionally want to get rid of DC
* high gain ensures high signal level making it less prone for (induced) noise

I´m not sure about the "gain control stage". Is the gain more expected to be >1 or <1 (like volume control pot). Why do you need it at all?

****
If you want us to give feedback about dedicated devices: post a link to the datasheets.

Klaus

 

[newbie_hs]

What frequency range are you targeting?

The bandwidth of acoustic emission signals from lithium-ion batteries typically ranges from low-frequency signals around 50–100 kHz to high-frequency signals up to 700 kHz.The amplitude of these signals ranges from a few microvolts to several millivolts

What information of the signal are you after?

Acoustic Emission signal from Li-ion battery. Don't know whether you will have access to the paper
Another link for same paper

The first stage should be a high gain amplifier stage, with LPF/BPF included.

You mean filter with some gain.Like sallen key filter.Please let me know if I am wrong

Data sheets of Major components are given below.

Sensor

Opamp

Variable gain Amp

 

[KlausST]

Don't know whether you will have access to the paper

No, I don´t need a paper. I need information from you.
Examples:
So when measuring acoustics .. one may calcualte a number: often one can hear the airplane outputs a noise level of 120dB(A). ..and maybe show the result in a numeric display.
--> thus I know you do the filtering (weighting) and get a normative value of the loudness.
OR you say you want to run an FFT to get the spectrum, amplitudes, phase information and show the results in a graphic display
--> this tells me you do the algorithms ... and so on
OR you say you just want to store the raw acoustic signals as a wave file on an SD-Card

My question rephrased: How do you process the data .. to get which result?
***

You mean filter with some gain.Like sallen key filter.Please let me know if I am wrong

First I need your informations ... then I can tell you if a sallen key filter is a suitable solution.

I wrote the three lines with the "*" ... to explain that it depends on the answers of my previous questions.

The first stage should be a high gain amplifier stage, with LPF/BPF included.
* LPF just to get rid of high dV/dt spikes
* BPF if you additionally want to get rid of DC
* high gain ensures high signal level making it less prone for (induced) noise

--> First things first. I asked my questions in a (hopefully) meaningful order. Most important first.
As long as I don´t know what your targets and requirements are ... I can not give detailed assistence. Thus please answer my questions first.

An example:
Let´s say I recommend a sallen key filter .. you design it according my recommendation. PCB is done. You are testing youtr application. At this point you give the information that you are interested in phase shift information...
Then you need to do a redesign since the (sallen key) filter manipulates your phase information. A maybe better solution then would be just to use a first order filter to get rid of frequencies above fs/2 (to avoid alias frequencies) and do the main signal filtering by software. As said: maybe.

Klaus

--- Updated Nov 15, 2024 ---

OPAMP:
if you are interested in signals up to 700kHz, then a 5.7MHz GBW OPAMP is a bit slow. The theoretical gain limit is about 8 ... while at this frequency it is not able to regulate out any distortion.

GAIN Amp: Here you have a huge bandwidth of 600MHz. It is not designed for audio at all, is adifferential input and output amplifier, draws a huge 100mA. I guess it is rather expensive .. and has bad availability.
I guess there are better choices.
It all depends on your needs ..

Klaus

 

[newbie_hs]

I will get the all required informations ASAP.

if you are interested in signals up to 700kHz, then a 5.7MHz GBW OPAMP is a bit slow. The theoretical gain limit is about 8 ... while at this frequency it is not able to regulate out any distortion.

What I understood is maximum gain I can provide to the amplifier is 8.Please correct me if I am wrong
May I know what you mean by regulate out any distortion

 

[KlausST]

What I understood is maximum gain I can provide to the amplifier is 8.Please correct me if I am wrong
May I know what you mean by regulate out any distortion

In simple words:
An Amplifier is a regulator. It continously "compares" desired output with real output. And in case of any difference it adjusts the output to better meet the desired value.

But when you are at the very limit .... it has no time to detect a difference and thus has no time to cancel out any errors.

Compare it with this:
You are driving with a car, suddenly there is an animal on the road. If you are driving slow (low frequency amplifier) you can avoid the animal. (You have time to see the "error" and you can adjust to it). But if you are at the top speed end of your and your car´s capabilities .. you have no timee to react, you can not correct the problem.

So with gain of 8 and a frequency of 700 kHz one can say the signal is amplified, but there is no useful feedback for any correction.

Klaus

 

[FvM]

I don't see a particular advantage of a chopper-stabilized ("zero offset") OP in this application. It's neither DC offset nor 1/f noise limited. Don't know if you are taretting to low level (noise limited) or high level (maximal magnitude handling critical) applications.