Audio Sense schematic needed

[MCH]

Hi,
I am considering building a power amplifier and want it to have audio sense. I have searched a while, but could not find any solution suitable for my project.

I do know Rod Elliot's Signal detecting circuit but this is a fully analog circuit that does not fit my needs:

Signal Detecting Auto Power-On Mk 2

ESP Project Pages - simple circuit switches on power when it receives an audio signal - also works as sound activated switch

sound-au.com

What i need is a circuit that can pull up (or down) a gpio of a microcontroller. It will be the microcontroller that will switch on/off the amplifier and takes care of the timing (elapsed time after the last sense before turning off the amp). The audio sense circuit itself does not need to have any timing capabilities, just send a signal whenever audio is detected.
Yes, i know i can use Elliot's circuit to pull up and down a signal, but i want to have full control of the timing via software + want the circuit to be as simple as possible.

Is anyone aware of how modern audio sense circuits work and most importantly, any schematic available?. I assume that in devices that digitalize the imput everything is done in the digital domain, but i hope i don't need to go that way, as it will make it too complex.

PS: i am interested only on building my own circuit, i am not interested in commercial finished products.

thanks.

 

[MCH]

After thinking about it, and given i will have 12V in the circuit anyways, i think i am going to go for Rod Elliot's design. It is designed for audio, tried and tested, and the explanation in the website is so good that even i can understand it. I will just change the relay for an optocoupler to pair with the microcontroller.

In the meantime i also found this other one, much simpler and with just a few parts, that is what i was looking for initially. But i was thinking that by the time it will take me to make it work as i want, i will have the more complex one built and running.

Analog circuit to check if an audio signal is present?

On the project I am working on, I have 3 line-level audio signals, and I need to determine on an arduino if a non-silent signal exists on each line and act appropriately (in my case, "unplugging" a

electronics.stackexchange.com

Still, if anyone knows of a more convenient circuit or IC, please let me know.

 

[mcdn]

Hey @MCH, I too have been frustrated by the complexity and component count of music detector circuits. Also a lot of them have low input impedances.

So, if you're using a microcontroller, maybe this one I designed could work? It connects to an ADC input rather than the GPIO, to enable signal level readout: https://www.circuitlab.com/circuit/pcy9wraj5mt4/music-detector-active/

It's basically a 20dB gain/buffer followed by a half-wave precision rectifier and LP filter, so will work with both SE and differential signals. It does require balanced power inputs to the op amps, but I'm sure someone better at circuit design than me could make it work with a single ended power supply.

I'd loe to get feedback/improvements done so we could "solve" this, as it's clearly a common use case. @Rick Sykora might have some views too.

 

[Rick Sykora]

Tbh, was holding back a bit as am a little too close to Buckeye’s design and need to be discrete. It feeds a MCU too. @mcdn’s looks similar but will say some refinement had to be done to avoid adding distortion. As @MCH is likely right that Rod’s is more complex but some of that is dealing with running at 12v and the output relay.

 

[mcdn]

Tbh, was holding back a bit as am a little too close to Buckeye’s design and need to be discrete. It feeds an MCU too. @mcdn’s looks similar but will say some refinement had to be done to avoid adding distortion. As @MCH is likely right that Rod’s is more complex but some of that is dealing with running at 12v and the output relay.

Thanks for chiming in @Rick Sykora . Any thoughts you can share on distortion issues would be appreciated, is it not sufficient to simply have a 100k input impedance and small decoupling cap? The actual power amp buffer runs in parallel to this.

 

[MCH]

Hey @MCH, I too have been frustrated by the complexity and component count of music detector circuits. Also a lot of them have low input impedances.

So, if you're using a microcontroller, maybe this one I designed could work? It connects to an ADC input rather than the GPIO, to enable signal level readout: https://www.circuitlab.com/circuit/pcy9wraj5mt4/music-detector-active/

It's basically a 20dB gain/buffer followed by a half-wave precision rectifier and LP filter, so will work with both SE and differential signals. It does require balanced power inputs to the op amps, but I'm sure someone better at circuit design than me could make it work with a single ended power supply.

I'd loe to get feedback/improvements done so we could "solve" this, as it's clearly a common use case. @Rick Sykora might have some views too.

That is great, thanks for sharing!
In my particular case, signal degradation doesn't matter, my DAC is a motu ultralite mk5 so i have channels enough to have one dedicated to the signal sensing input summing all other channels via software in this single output (additional advantages: detection independent of volume setting as this channel will be fixed and possibility to "dial" the detection sensibility changing the gain). This is for a multichannel amp and this is what I was going to do anyways. But the balanced power supply is indeed a limitation.... hope that someone chimes in.
In any case the discussion about how to avoid additional distortion is appreciated.
No worries @Rick Sykora I have a Boxem amp -this is where the inspiration came from, it works fantastic - and wasn't expecting Fred to chime in

 

[mcdn]

That is great, thanks for sharing!
In my particular case, signal degradation doesn't matter, my DAC is a motu ultralite mk5 so i have channels enough to have one dedicated to the signal sensing input summing all other channels via software in this single output (additional advantages: detection independent of volume setting as this channel will be fixed and possibility to "dial" the detection sensibility changing the gain). This is for a multichannel amp and this is what I was going to do anyways. But the balanced power supply is indeed a limitation.... hope that someone chimes in.
In any case the discussion about how to avoid additional distortion is appreciated.
No worries @Rick Sykora I have a Boxem amp -this is where the inspiration came from, it works fantastic - and wasn't expecting Fred to chime in

If this is for your multichannel TPA3255 project, then I have a schematic and PCB design and control software (and some built PCBs) for a 4-module TPA3255 controller using a Pi Pico or Pico W. That's why the sense circuit has 20dB gain, because the Pico ADC only has 8 bit resolution so gets swamped by noise. It's a bit rough around the edges so I haven't made it public on Github yet, but if you PM me with your Github ID I can give you access.

 

[MCH]

If this is for your multichannel TPA3255 project, then I have a schematic and PCB design and control software (and some built PCBs) for a 4-module TPA3255 controller using a Pi Pico or Pico W. That's why the sense circuit has 20dB gain, because the Pico ADC only has 8 bit resolution so gets swamped by noise. It's a bit rough around the edges so I haven't made it public on Github yet, but if you PM me with your Github ID I can give you access.

Yes it is for my heat pipe cooled project
Sure, will send you a PM, I am very curious to see your multichannel amp. Thanks!

 

[MCH]

Hi guys, sorry to come back to this again. I know the following questions are very stupid but remember i am very noob and just want to make sure i am not missing anything. My apologies in advance:

I am mostly done with my project but when i was about to send the PCBs to print, I had some second thoughts.
I really want this audio sense circuit to work well and i want to make sure i chose the best way to interface it to the amp modules.

I have eventually chosen to go with Rod Elliot's design, and use an optocoupler to interface with the amplifier, the interface part looks like this:

What turns the amplifier ON/OFF is a voltage supervisor (not shown in the picture). There are two conditions for the voltage supervisor to turn the amplifier on:
- That the voltage supply to the amplifier reaches a thereshold: there is a voltage divider that brings down the power supply voltage to VDD_supervisor, connected to the supervisor input. When VDD_supervisor reaches >2.93 Volts, the supervisor turns the amplifier on.
- That the "MR" of the voltage supervisor is not low.

The idea is to keep the MR pin low until the audio signal starts, no matter if the other condition (power supply) is met or not.

I am using an optocoupler that will turn on when audio signal is detected. As optocouplers are normally open switches, i am using a mosfet that connects MR to ground. I am using VDD_supervisor to keep the mosfet on. I use VDD_supervisor for this task for no particular reason. VDD_supervisor is ca. 4V.

But when i look at the result, it seems to me way overcomplicated, and i think that i don't need the optocoupler at all and that this simplified version would do the job exactly the same:

But for some reason i have the feeling that i am missing something in this last option.... what would you guys do in this case?