Need Cuircuit of Precise 1 Watt Amplifier

[crissty]

I need to make a Precise 1Watt Audio Amp for test purpose, having a feedback which would controll the output to exactly 2.83V.
give me some suggestions on how to do this by using feedback with comparator OPAMP.

I would be very greatful, for the help extended.

Regards,
Chriss.

[jan.didden]

Just a question: is it not possible for you just to use a good amp and precisely control the input level for the required output level?

If not, you probably are looking for an AGC circuit.

Jan Didden

[crissty]

Actually i can, but i need to have a perfect & assured way that the o/p is precisely 2.83V, without controlling the VC & everytime setting the output.
I have made Sine/Tri/Squ Wave Generator, Pink Noise Gen.

[jan.didden]

Originally posted by crissty
Actually i can, but i need to have a perfect & assured way that the o/p is precisely 2.83V, without controlling the VC & everytime setting the output.
I have made Sine/Tri/Squ Wave Generator, Pink Noise Gen.

One issue you have to face is that sine, square, triangle have all different ratios of peak to RMS value levels for putting 1W in 8 ohms. So the AGC probably needs to be switched also when switching to a diferent wave.

Look at the MAX9756

Jan Didden

[Duo]

If you need to control the output that well, you'll have to set up a limiter/compressor of sorts... Or as janneman says, AGC (Automatic Gain Control).

One way of doing this is to have a detection curcuit at the output of the audio amplifier that is set to 2.83V. The output of this curcuit could control a VCA (Voltage Controlled Amplifier) stage which drives the main amplifier.

This way, you set a limit output voltage in the detection circuit and use that to vary the gain of the input amp so that the output always stays at your set voltage.

(This might be useful if you want to do 1W into speakers with impedance other than 8 ohms.)

16 ohms = 1W @ 4.000Vrms
8 ohms = 1W @ 2.828Vrms
4 ohms = 1W @ 2.000Vrms
2 ohms = 1W @ 1.414Vrms
1 ohm = 1W @ 1.000Vrms

Just take the root of the impedance if you like to find required voltage to get one watt. And if you need the voltage required for a different wattage, multiply watts and ohms and then take the root of that product.

[jackinnj]

if you have some money to spend consider a Thaler SWR200 and an attenuator -- 7.071 VRMS +/-0.05%, 3ppm stability -- Thaler will sell parts over the phone.

elsewise -- (and I have used this) -- use a precision rectifier to sample the output and compare to a reference -- you need a good rail-to-rail opamp for this. the error signal controls the bias of a voltage controlled amplifier -- this is how the very acurate signal generators from HP and Boonton do it. if you need some SSM2018's let me know.

[myhrrhleine]

Most amplifiers are fairly stable wrt voltage out and simply deliver the needed current while maintaining a constant voltage.

All you really need then is a 'volume' control. A linear potentiometer would be easiest to adjust.

If you need 1 watt at each of the waveshapes (~2.83 for sine waves) you might want a switch so that each wave can be set differently/appropriately for 1 watt.

[crissty]

Can i do this Way, first I would make a Regulated +_12V and regulate it with 7812 & 7912.

Then each outputs (i.e. Sine / Tri / Squ & Pink Noise) would have a Pot.

Would that do the job as i have a well regulated PS & a chip amp have +- 12V would have a decent & non distorted 1W Output.

I hope this would ba a good idea, suggest me a good chip amp for +- 12V.

[jan.didden]

Originally posted by crissty
Can i do this Way, first I would make a Regulated +_12V and regulate it with 7812 & 7912.

Then each outputs (i.e. Sine / Tri / Squ & Pink Noise) would have a Pot.

Would that do the job as i have a well regulated PS & a chip amp have +- 12V would have a decent & non distorted 1W Output.

I hope this would ba a good idea, suggest me a good chip amp for +- 12V.

What do you want to test with this? That objective should determine the acceptable linearity and stability. If you want to do listening test for speakers, as an example, your distortion probably is not the first priority.

Jan Didden

[crissty]

Basically intension is SPL & Frequency Response Measurements