SparkFun Forums

[mprowe]

Hi,

I would like to use the Sound Detector (SEN-12642) for an in-car project. However due to the high amounts of spurious audio noise, I would like to replace the electret microphone with a wholly electrical path.

The input to the Sound Detector is to be one of the speaker outputs from a small Bluetooth Audio amp. I am currently experimenting with this one: http://www.ebay.co.uk/itm/TDA7492P-50Wx ... SwPcVV2ORX. But this could be changed.

My question is this: How do I couple the high-power output from the BT amp to the very sensitive input on the Sparkfun Sound Detector? I wonder if a small transformer or opto-isolator would do the job? In addition to the electrical constraints, size is also important to me.

Your comments and advice would be greatly appreciated.

Regards, Martin

[jremington]

The schematic diagram for the Sound Detector shows that you can change the gain of the input stage by adding a resistor at pads for R17. Adding a 1K resistor will result in a gain of a bit less than 1, which might be OK if the audio input levels are typical of line input (about 1 volt RMS). It would be best to use a line input source.

For the speaker output of the BT amp, use a multimeter to estimate the amplitude of the output signal at the volume you will be using (with a speaker connected), and make up a voltage divider to reduce that voltage to line input range for the detector. It would be around 20 VRMS for 50 watts into 8 ohm speakers (if you like it loud), so roughly 220K:10K resistors.

[mprowe]

... and make up a voltage divider to reduce that voltage to line input range for the detector...

Ah.. the simplest is the best! Thank you.
What about R1? Should I cut the track to remove the the DC bias?

Regards, Martin

[jremington]

Yes, cut R1 - it is power for the microphone.

If all you want is a logic indicator that there is audio coming from the BT amplifier, just rectify the speaker output.

[mprowe]

If all you want is a logic indicator that there is audio coming from the BT amplifier, just rectify the speaker output.

Yes, that is all I want to know - that there is audio from the amp...
Let me just check my understanding of your suggestions?
Are you saying that I can do away with the Sound Detector by just connecting one of the speaker outputs to a bridge rectifier. If I then clip the max (rectified) output to 5v with a zener diode, I could read that state on a ATMega input.

Regards, Martin

[jremington]

by just connecting one of the speaker outputs to a bridge rectifier

Yep. You should isolate the speaker outputs with a small cap on each input leg of the bridge rectifier (say 100 nF).

The audio level would have to peak at 4-5 V to give a valid logic signal. If much smaller, you could use a single transistor amplifier as a switch.

[mprowe]

Yep.

Thanks J(?),

I'm away for a few weeks now, but when I'm back, I'll breadbord this and see what I get.

Once again, many thanks, Martin

[mprowe]

Hi again,
Still struggling with this "Squiggly analogue" stuff!!
On a slightly different tack, I now know that my test amplifier has a TDA7429P chip. Which on investigation, I find is a D-Class digital amplifier.
WooHoo... I thought. Back to "on's 'n off's" - I can cope with that!
I thought that all I need to do is: XOR the amp output plus & minus (+/-) signals and when I get a zero (as a result of a 50% duty cycle), I've got my silence?
Nope! Not working as expected!

I only have a Picoscope PC oscilloscope, so I'm not absolutely sure that I am able to "see" what is actually going on, so I'm relying on a very basic understanding of PWM techniques. Of course, this chip may be using a different modulation scheme anyway!!

Any thoughts or suggestions will be most welcome.
Regards, Martin