automatic gain control!

[the cracken]

i need an automatic gain control for an amp.
i was thinking about using one of those digital pots to alter the gain of an opamp. but how can i drive the digital pot. if i use its (the digpots) threshold then it would constantly cycle up one and down one and it think this would induce alot of distortion to the audio. anyone have any ideas?

i want to keep it as simple as possible so no pic's or anything because ill get confused

heres what i was thinking.

this is only the feed back part of the circuit. The digital pot go's to the gain resistor R1 which isn't shown.

**broken link removed**


does the Vcc to the inverting input need to be the same as the Vcc power supply of the opamp??

Most importantly DO YOU think this will work and any ideas would be gratefully received!!

thanks very much!!

 

[bychon]

The Vcc part is right. And yes, this will tick up then down all day long when no adjustment is needed. If the change in gain is small enough and slow enough, you won't notice it. Human ears aren't all that accurate. Be sure not to let it make decisions at audio frequencies. That is, keep the 555 running slower than 20 Hz. A lot slower if you can.

There is a thing called a transconductance amp..single chip like an op-amp. Its gain is controlled by how much current is flowing through a circuit in the chip so it's easier to adjust. Too much loudness? Reduce the multiplier current. Completely analog. No incrementing, no timer needed. Look 'em up and read all about it. See if that's what you need.

 

[the cracken]

The Vcc part is right. And yes, this will tick up then down all day long when no adjustment is needed. If the change in gain is small enough and slow enough, you won't notice it. Human ears aren't all that accurate. Be sure not to let it make decisions at audio frequencies. That is, keep the 555 running slower than 20 Hz. A lot slower if you can.

There is a thing called a transconductance amp..single chip like an op-amp. Its gain is controlled by how much current is flowing through a circuit in the chip so it's easier to adjust. Too much loudness? Reduce the multiplier current. Completely analog. No incrementing, no timer needed. Look 'em up and read all about it. See if that's what you need.

Thanks for the great tip. i was going to make an analog gain control but i didn't know how to make a variable voltage into a variable current!! know i do ill change the circuit! here's what i think ill do instead!
**broken link removed**

do you think this will work?? thanks for the help!

 

[the cracken]

Thanks for the great tip. i was going to make an analog gain control but i didn't know how to make a variable voltage into a variable current!! know i do ill change the circuit! here's what i think ill do instead!
**broken link removed**

do you think this will work?? thanks for the help!

i don't know anything about transconductance op amps do you know of one that i can use for this application???
i want to keep the AC in at about ~17V peak to peak. so rectified = ~8.5v x 0.707 = ~6v ill need a bias of ~-6.1
the led works at max brightness at 20mA Iout=Vin/R1 therefor for a 0.1V input ill need R1= 5Ω to give a 9V output i need a resistor of 450 ohms (R2=Vout/I)

does any one have anything to add????
thanks for the help!

ignore the circuit above this is the real one
**broken link removed**

 

[audioguru]

Some FM and TV stations in my city have horrible automatic-gain-control systems.
The first syllable of every word comes blasting through like a hammer before the slow circuit reduces the gain.
The first background sound blasts through distorted and much too loud then later the very slow circuit reduces the gain.
If there is a loud momentary sound then the AGC cuts all sounds too low for a few seconds.

With music the horrible stations have the background music volume pumping up and down with the "beat".

 

[MrAl]

Hi there,


One little note here...

In one of your circuits it looks like the feedback is positive when it should be negative. In other words, it looks like when the audio input goes higher then the gain goes higher too, which would be opposite of what you really need. You might want to look at that and make sure that when the signal goes higher the gain goes lower.

As audioguru was saying more or less, these things are a little problematic in that they dont work exactly as perfectly as we would expect.
One problem is sometimes called "breathing", where the sound output goes up and down so that it sounds like some person breathing in and out. It sounds rather crummy. Probably the only way to get a near perfect system is to digitize everything coming in and store it, examine the data, develop a signal and use that to compensate the output signal. That would require some AD and DA converters and memory or just some DSP microcontroller or possibly just a delay line (with more sundry circuitry) or similar, all working in real time.

 

[bychon]

You can count on these guys to program a microcontroller to operate a flashlight!

Are you experimenting or building a radio station that is better than the ones in England?

Oh yeah, Signetics makes a NE570/571/SA571chip called a compandor. Compressor/expandor, and AUTOMATIC LEVEL CONTROL. Good place to look up breathing problems and see how application note AN174 tells it.

 

[Space Varmint]

You can count on these guys to program a microcontroller to operate a flashlight!

Are you experimenting or building a radio station that is better than the ones in England?

Oh yeah, Signetics makes a NE570/571/SA571chip called a compandor. Compressor/expandor, and AUTOMATIC LEVEL CONTROL. Good place to look up breathing problems and see how application note AN174 tells it.

Motorola makes one too but a compressor is not an AGC circuit. A compressor drives the signal more positive so that the signal peaks saturate thereby keeping them within a particular amplitude level and is usually used in FM to decrease deviation of a transmitted signal. An AGC goes negative to reduce the biase so that if a signal is too loud and / or the AGC loop is hyper-responsive it will cutoff the transistor output.

On the Op's circuit, I have never seen an AGC that used a full wave rectifier. Usually only one diode is used and yes check the direction like MrAl said.

 

[bychon]

Are you suffering from the belief that I don't know the difference between a compressor and an agc circuit?

 

[Space Varmint]

Are you suffering from the belief that I don't know the difference between a compressor and an agc circuit?

Maybe I was...lol. Sorry man, it was the way you brought it up that sort of through me off track.

 

[k7elp60]

You may want to look at a SA571 dual, compandor.