automatic gain control circuit

Hello,

I'm trying to come up with an automatic gain control circuit that can norma lize a 50Hz sine wave of varying amplitude (50mV to 5V), to an output of th e same sine wave but with an amplitude of 4.8V. (I'm looking at a max of 1k Hz)

My first thought is to try a voltage controlled amplifier with a peak detec tor used in it's feedback circuit, the peak detector would feed an op-amp w ith a 4.8V reference that represents the highest peak output I would like a nd that op-amp would adjust the voltage controlled amplifier until that pea k voltage is reached.

I'm trying to simulate this using TI's VCA810, but I haven't been able to g et it to work.

Can anyone recommend a a circuit to achieve this AGC design? Does my approa ch make sense?

much thanks!

Reply to
Fibo
Loading thread data ...

For a slowly varying input you can probably get it to work. You might be better off with a CdS photocoupler in the feedback path of an OpAmp. ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
 Click to see the full signature
Reply to
Jim Thompson

That should work if you close the loop carefully.

A better detector would help. An active full-wave rectifier followed by an active lowpass filter would be a good start. Then, basically, a PID controller. Working down to 50 Hz, your loop will be slow, ballpark 1 Hz maybe.

Reply to
John Larkin

Found this....

If honey-do chores permit ;-) I'll see if I can roll a Spice model from this information. ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
 Click to see the full signature
Reply to
Jim Thompson

I worked on a board that used a photocell gain control circuit. The AGC loop was closed through software. The first pass oscillated like crazy until we realized the coupler had ms of delay. Of course it oscillated! We had to restrict the max gain setting to make it stable.

I'm not sure why a photocell is better than the VCA.

--

Rick
Reply to
rickman

Wouldn't the LM13700 be pretty well suited to this? I think it has AGC circuits right in the datasheet.

At these low frequencies the easiest way to build a hard peak-limiter would be to have the rectifier/filter feed a a coupled LED/LDR in the feedback loop of an opamp. Just wrap heat-shrink around them and seal it up.

Reply to
bitrex

I think the few commercially-available OTAs suitable for building a quick-and-dirty VCA only have a very limited linear voltage input range; a couple tens of millivolts. So noise could be a factor?

Reply to
bitrex

What's the signal on the sine wave? If you're just trying to recover phase or frequency information then skip the AGC altogether. Run the thing into a good fast comparator and what comes out the other end will be nicely squared up. Then if you must have a sine wave you can filter the snot out of it.

--
Tim Wescott 
Wescott Design Services 
 Click to see the full signature
Reply to
Tim Wescott

malize a 50Hz sine wave of varying amplitude (50mV to 5V), to an output of the same sine wave but with an amplitude of 4.8V. (I'm looking at a max of

1kHz)

ector used in it's feedback circuit, the peak detector would feed an op-amp with a 4.8V reference that represents the highest peak output I would like and that op-amp would adjust the voltage controlled amplifier until that p eak voltage is reached.

get it to work.

oach make sense?

Coo, they aren't giving the VCA810 away. For less than that you can buy an analog multiplier. Scroll down here and they have an AGC circuit.

formatting link
f

How are you resetting the peak detector?

George H.

Reply to
George Herold

malize a 50Hz sine wave of varying amplitude (50mV to 5V), to an output of the same sine wave but with an amplitude of 4.8V. (I'm looking at a max of

1kHz)

ector used in it's feedback circuit, the peak detector would feed an op-amp with a 4.8V reference that represents the highest peak output I would like and that op-amp would adjust the voltage controlled amplifier until that p eak voltage is reached.

get it to work.

oach make sense?

Thanks for all the responses

- I'll probably pass on the OTA solution

- I'm guessing I would use the opto's resistance to change the gain... coul dn't I also use a FET in the linear region in a similar way?

- full wave rectifier > LPF > PID, sounds like an interesting approach I'll look into something like this

- multiplier circuit might work too

The circuit would have to handle random signal wave forms as well as sine w aves.

Reply to
Fibo

ormalize a 50Hz sine wave of varying amplitude (50mV to 5V), to an output o f the same sine wave but with an amplitude of 4.8V. (I'm looking at a max o f 1kHz)

etector used in it's feedback circuit, the peak detector would feed an op-a mp with a 4.8V reference that represents the highest peak output I would li ke and that op-amp would adjust the voltage controlled amplifier until that peak voltage is reached.

to get it to work.

proach make sense?

pdf

I hadn't gotten that far, but maybe just a resistor across the cap... or I might not need to reset it for my application, only when the circuit first turns on... the VCA810 is pricey but this is a one-off so it's tolerable

Reply to
Fibo

A CdS photoresistor is going to be a heck of a lot linear than a FET in the linear region -- you'll get to the same place with a lot less fuss and muss. Of course if it's a one off then there's a lot of fuss and muss to be avoided just by buying a VGA or a good multiplier.

What sort of response speed do you need? You need to filter out the "bumps" from a 50Hz sine wave, which is going to significantly limit your response speed. There are ways to speed this up (detect zero-crossings and "filter" by integrating and dumping is one way), but they're error prone unless you have the right signal.

And, of course, if you're starting with an amplitude-modulated signal, then the worst thing in the world would be cycle-by-cycle AGC...

If the signal needs to handle random waveforms then make sure you know _what_ you want to be constant -- the peak to RMS to average absolute value ratios are all fixed for a sine wave, but are very much not so for random waves.

--
Tim Wescott 
Wescott Design Services 
 Click to see the full signature
Reply to
Tim Wescott

ormalize a 50Hz sine wave of varying amplitude (50mV to 5V), to an output o f the same sine wave but with an amplitude of 4.8V. (I'm looking at a max o f 1kHz)

etector used in it's feedback circuit, the peak detector would feed an op-a mp with a 4.8V reference that represents the highest peak output I would li ke and that op-amp would adjust the voltage controlled amplifier until that peak voltage is reached.

to get it to work.

proach make sense?

uldn't I also use a FET in the linear region in a similar way?

ll look into something like this

waves.

ah, you can't just stick it into an oscillator then.

There are so many options. Are vogad ICs still available?

NT

Reply to
tabbypurr

50Hz is within audio range so LM1036 and similar "audio processors" could do the variable volume (aka gain) function cost effectively.

piglet

Reply to
piglet

How about a comparator followed by a LP filter? That would be about 25 cents.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
 Click to see the full signature
Reply to
Phil Hobbs

Yes it does. But a better circuit might be the ancient NE or SE 571

At these low frequencies the easiest way to build a hard peak-limiter

Reply to
David Eather

** In audio signal processing, that system is know as a "peak limiter".

The audio is first full wave rectified and used to charge a capacitor via a diode. The charging time constant is made much shorter than the discharging one - for example 2mS and 200ms - known as the "attack" and "release" times.

The cap voltage is used to control system gain, using a JFET attenuator or an app specific IC.

The system rests at high gain with no input, dropping to low gain a few mS after a strong signal appears, returning to high gain in about 600mS if the signal drops to the minimum.

There are errors in the scheme, a signal that rises faster than the attack time will overshoot and maybe clip while during the release time a small signal will be made smaller than the desired output.

Making the attack as fast as possible reduces the first error while making the release shorter causes significant waveform distortion at low frequencies.

With audio signals, if the attack is made longer than about 25mS, the same system expands the dynamic range of the input signal rather than compressing it.

The inherent errors are usually tolerable when applied to a voice or music signal, but may not be so tolerable in your application with its 40dB range requirement which exaggerates errors.

.... Phil

Reply to
Phil Allison

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.