Messure ac volume rising speed

[framlin]

Servus,
I want to create a circuit, that can messure, how fast the volume of an ac signal rises.
If the volume reaches a defined level (starting by 0) too fast (eg under 200 ms)  a red led should go on.
If the volume has not reached the critical level at the messure-point in time, a green led should go on.
With volume the abs(volume) is meant.
The ac signal is an audio-signal (voice) genarated ba a microphone and an amplifier. So what I want to messure is, if the voice is going too loud, too fast.

My current idea is, to do this with op-amps, biased as differentiators or an op-amp biased as comparator together with a 555, to messeure the elapsed time.

I am an electronics noob, coming from the software-side of the technical world.
If anybody has any ideas or hints, how I can do this voice-volume-speed-mesurement, I would appreciate this very much.

[michaeliv]

A purely analog circuit for this would be difficult to construct if you don't have a trigger in that would tell you when to start measuring.
Do you have a trigger in? As in do you press a button and after 200ms the device will either flash red or green depending on whether the volume increased too fast in the previous 200ms ?
An easier solution would be an arduino. The built-in ADC of an arduino can sample at about 10k/s. This should be enough to give you good accuracy. And it will be much easier to implement the logic in software.

[framlin]

@michaeliv: thanks for your answer.
I wanted to have a short silence-period as a trigger. So there is no button, but if there is silence followed by a small volume of the voice, the messurement should start.
This is another challenge I have to distinct vioce from environmental noise.
If I would do that in software ( I actally have done this 20 years ago with delphy on a pc) I would use a fft to filter human voice.
I know, that I have to filter a band between 100 and 200 Hz.
So I know, how to do that in Software, but I want to do it in hardware, to have a low-powered wearable device.
At the end, I want to have something, that is mounted on a necklace, that is vibrating, if I speak too fast to loud ....
So my fear is, that an atmega328 is not powefull enough and needs to much power.
I did some experiments with a stm32f04, but I would like to have an analog circuit at most, if it would be possible.

[danadak]

You might consider doing a rms converter to extract rms value of voice, rather than some
arbitrary value. The triggering is simple and consistent.

Consider a PSOC, it has the analog stuff, and a DFB (digital filter block), and can achieve
low power. Quite a few videos on their web site for training, in fact there is a project that
does a complete signal path filter from input pin to output pin.

http://www.cypress.com/products/32-bit-arm-cortex-m3-psoc-5lp

There is also PSOC Bluetooth family, PSOC 4 based, but filter would have to be
done analog or SW based FFT, and there has been an FFT project somewhere in
their forums, on PSOC 4.

http://www.cypress.com/products/psoc-4-ble-bluetooth-smart

Regards, Dana.

[framlin]

@Dana: Tank you for this PSoC-hint, I wll have a look at it.