Art and Electronics

After having been to the class-meeting I told you about, I came again in contact with my former arts teacher.

38yrs. ago we had been doing a kybernetic sculpture together, consisting of a wave-shaped long mirror with 7 flash tubes in the focus points of it. A small cutout of a piano keyboard(1 octave) was used to trigger those tubes by the spectators. Now he asked me to restore the original condition again.

I had the idea to make another interface working with a mike array. The sound of people standing in front is picked up, processed and triggers the respective bulbs. Bandpass filters with moderate Q(5) resonate between 200 and 800Hz (adjustable). The idea is to get the people humming/singing and if they hit the right pitch the bulb starts flashing faster up to 5/s. If there are several people in harmony more bulbs will start flashing, you get it.

To avoid people screaming too loud I want to make a AGC circuit, that keeps the level constant over a range of 60dB. Here is my AGC being the most critical block:

60dB AUTOMATIC GAIN CONTROL 1/4 SSM2164 /|----. o / | \\/ | ___ -5V | .---< | /\\ o--o-|___|-o threshold.-. | \\ | | | 15k .--->| | .-. \\|-o--' .-. | | | | | | | | | '-' 15k| | | | |510 | | '-' || | '-' | GND o-----||-o | | | 1n|| | --- |0.003..3V___ | |\\ | ---560p | .----|___|--o--|-\\ | | | | 10k | >--o === .-o---o------. .--|+/ | GND | fullwave | | |/ | | Rectifier| === | ___ | Averager | GND | .-|___|-. '-----o------' .-o-|\\ | 15k | | ___ | \\/| \\ | |\\ | o----o---------|___|-o--o /\\| >-o-|-\\ | || Input 15k | | | / | >--o---||---o to .-. '---|/ .-|+/ || filter bank | | | |/ 510| | | '-' === | GND --- 560p --- 1/4 SSM2164 | === GND (created by AACircuit v1.28 beta 10/06/04

--
ciao Ban
Bordighera, Italy

Seems way too complicated to me. If all you are looking for is the presence and relative levels of frequencies over the voice range, something more class-C-like would work since we aren't trying to preserve the audio envelope here for further use.

Ban wrote: 60dB AUTOMATIC GAIN CONTROL

1/4 SSM2164 /|----. / | \\/ | ___ -5V .---< | /\\ o--o-|___|-o | \\ | | | 150k .-. \\|-o--' .-. | | | | | 15k| | | | |510 '-' || | '-' o-----||-o | | 1n|| | --- 0.003..3V___ | |\\ | ---560p .----|___|--o--|-\\ | | | 10k | >--o === |\\ .-----o------. .--|+/ | GND | \\ | Fullwave | | |/ '---------------|-1>-. | Rectifier| === | / | | Averager | GND |/ | '------o-----' .--------------------' | | ___ | | .-|___|-. | .-o-|\\ | 150k | | ___ | \\/| \\ | |\\ | o----o-------|___|--o--o /\\| >-o-|-\\ | || 300mV | | | / | >--o---||---o to .-. '---|/ .-|+/ || filter bank | | | |/ 510| | | '-' === | GND --- 560p --- 1/4 SSM2164 | === GND (created by AACircuit v1.28 beta 10/06/04

corrected schematic, I forgot the inverter. The circuit works as follows:

3V[3mV] rectifier out, the VCA will have to do +20dB[-40dB] to compensate for the input current which results in a control voltage of -33mV*+20[-40]= -660[+1320]mV The inverter makes +660[-1320] +660 = +1320[-660]mV resulting in -40dB[+20] main gain and 330mV out. This way there is also a possibility to adjust the threshold with a linear pot.

--
ciao Ban
Bordighera, Italy

Or you could look at using something like a THAT4301

Available through UK-based ProfusionPLC for US$18 with MOQ of 2, or about $9 each in tubes of 25. A bit pricy, but a well-proven single-chip solution with excellent audio quality (not that the quality matters in this application).

But you know, the classic technique for a levelling VCA is to use an optoisolator, like this:

Okay, maybe overkill, especially with the tubes :-)

OK, you just have one channel but do you really need 2 VCAs?

| .-. level adj. ___ | |

Hello Ban,

If you want to build a really nice one that preserves as much of the spectrum as possible look around in ham radio publications (or on sites like

Be patient, some of those sites are slow because they must run at a budget close to zero. There these limiters are usually called "clippers".

Anyway, another option and a simpler circuit could probably be built around the old MC1496 which is still around.

Regards, Joerg

Sorry the delayed response, but I was visiting a beautiful electronic music festival near Berlin

The resistor was misdrawn and is not there at all and the value belongs to the input resistor below.

Here is another part with an adjustable bandpass of Q=2.5...5; Fullwave rectifier and an integrator driving a comparator to get a repetition frequency between 0Hz and 5Hz depending on the input level of 3mV to 3V. The output drives a relais (20ms on) and also the analog switch(1NO, 1NC) . Power is +/-5V. I make all sections dual for increased reliability and will handwire the different sections together.

|| .-||---------. | ||56n | ___ | .. ___ | ___ ___ ___ o-|___|-o-||-o-|___|-o-----|___|---------|___|-o-|___|-o 10k | || | 100k | 22k 22k | 1M -5V .-.56n| |\\ | | | | '-|-\\ | ___ ___ ___ | .-----. | |1k | >--o-|___|-o-|___|--o--|___|-o | | '-' .-|+/ 10k | 10k | 22k | | .-. | | |/ | |\\ - o /o | | .-. === AD8625 .--o-|-\\ ^ 4148 / | |10k -->| | GND | | >---+ ADG623 / '-' | 1k| | | -|+/ - o 1u || | | '-' | GND|/ ^ 4148 +-----||--o | | | | | |\\ || | '----o '-----------' '--|-\\ | | | >---o === GND-|+/ | GND |/ | Bandpass adjustable Fullwave Rectifier Integrator | .----------------' | switch | driver | |\\ AD8625 | '-|-\\ | ___ | >--o------------' +5Vo-|___|--o--|+/ | 33k | |/ | ___ | ___ | -5Vo-|___|--o--|___|-' 33k 39k

Comparator (created by AACircuit v1.28 beta 10/06/04

--
ciao Ban
Bordighera, Italy

3rd part Mike preamp

Electret +5V o Microphone Preamp | 150u 40dB gain | +[/ __ ___ .---o--[|-|___|-o-|___|-. | | [\\ 10 | 1k | | .-. | || | --- | | o----||-o --- | |4k7 | |\\ || |15n 3n3| '-' '-|-\\ | ___ out | | || | >--o-|___|--o o---o----||---o---|+/ 220 | | || | |/| SE5534A _ | .-. 47n .-. o | / \\--' | | | | +/-5V |(Mic) | |10k | |33k | \\_/--. '-' '-' | | | 140Hz-7kHz o---o === | | GND | / \\--' .-. |(Mic) | | | \\_/--. | |10k | '-' | | '---o | === GND (created by AACircuit v1.28 beta 10/06/04

--
ciao Ban
Bordighera, Italy

Hello Ban,

The feedback resistors 10/1k seem a bit low. Can't you increase those, lower the 15nF and then get away with a smaller electrolytic than that

150uF? In my experience the smaller electrolytics last longer, they don't dry up as fast. Assuming that this kind of stuff might be left out baking in the sun or in a hot car trunk at times.

Regards, Joerg