-10 to 10 gain

[petesmc]

How many OpAmps would I have to use to make a system that can range from -10 to 10 gain using only one potentiometer to control it? Don't tell me how to do it, just the number of Opamps. Oh, and no transistors.

 

[JimB]

Only one.

JimB

 

[samcheetah]

Only one.

JimB

i would like to know how would you do that

 

[petesmc]

I couldn't work it out for one, but I did for two.

 

[evandude]

oh i bet you could EASILY do it with one if you had one of those neat pots that goes from 10k to -10k ohms :lol:

 

[Nigel Goodwin]

Only one.

JimB

i would like to know how would you do that

Simple, use an opamp in inverting configuration (he never specified it had to be non-inverting!) - use a 10K resistor as the input resistor, and a fixed 1K resistor (to give -10 gain) in series with a 99K variable resistor (to give +10 gain) as the feedback resistor.

To do it with a non-inverting design, give the opamp a gain of 10, and have a volume control feeding it - fit a fixed resistor in the ground connection of the volume control, to set the minimum attenuation to give an overall gain of -10.

 

[bmcculla]

Nigel I think you a bit confused. Gain for an inverting op amp: G = -Rf/Ri so you get G= -1/10 and G= -10/100 so you are going between -.1 and -10 gain not -10 an +10. Same problem with your non-inverting amp - you can't have an attenuator that gives negative gain.

 

[Roff]

I finally found **broken link removed**. I knew I had seen it somewhere. Below is a slightly modified version that uses a pot to vary the gain.

 

[Nigel Goodwin]

Nigel I think you a bit confused. Gain for an inverting op amp: G = -Rf/Ri so you get G= -1/10 and G= -10/100 so you are going between -.1 and -10 gain not -10 an +10. Same problem with your non-inverting amp - you can't have an attenuator that gives negative gain.

I wasn't confused?, but I am now!.

Both suggestions I gave would give adjustable gains from -10 times (attenutation) to +10 times (amplification).

An attenuator (by definition) gives attenuation (negative gain = a loss).

You're possably talking about inverting?, which he never mentioned, and I don't see implied in his question?.

 

[bmcculla]

Thanks Ron H. I was thinking about a modified diff amp but couldn't quite get it.

Nigel: I think we just have a symantics problem - two different ways to talk about gain. The signal processing definition of gain is output over input: Vout/Vin. This definition works out nicely because you just multiply your input voltage by G to get output voltage. Using this definition Attenuators have a G of less than one - for G = .5 a 1V input produces a .5V output. When you have a negative gain there is an inversion - for G = -1 a 1V input gives a -1V output.

I also just noticed that I got the second gain calculation upside down it should be -100/10 => G=-10.

 

[JimB]

Here is what I had in mind:

The attenuator puts varying amouts of signal to the two inputs of the differential amplifier.
When the two inputs are equal, there is no output (gain = 0).
When there is more signal to the non-inverting input the gain is +Ve.
When there is more signal to the inverting input the gain is -Ve.

Set the amplifier input and feedback resistors to give the required gain, which depends on the attenuator losses and the output required.

JimB

 

[Nigel Goodwin]

I think you're all making this far too difficult?.

As I understood the question, he wanted either a gain of 10, or an attenuation of 10, and fully variable between the two options - which is what I provided, either inverting or not, with one opamp.

The schemes for changing from inverting to not-inverting are flawed, because what happens when you get to the balance point - no output at all!.

Perhaps 'petesmc' should clarify the question further - presumably it's a school question?.

 

[eblc1388]

I'm too are confused. What I understand a negative sign for gain means a signal invertion. So gain=-10 means the output is ten times the input but inverted.

Can gain ever be negative? How comes people suddenly find a new meaning for negative gain to mean a loss?

 

[Nigel Goodwin]

Can gain ever be negative? How comes people suddenly find a new meaning for negative gain to mean a loss?

I think you're asking the wrong way round?, it should be 'how come people suddenly find a new meaning for -10 as inverted?' :lol:

It comes from gain as dB, -3dB is a negative change in gain, and +3dB a positive one, not a change in polarity - mostly polarity is totally irrelevent (unless you're talking DC amplifiers of course!).

But, as very often, too little information in the question!.

 

[eblc1388]

But, as very often, too little information in the question!.

The only information missing from petesmc's original post is what happens when we reach gain=1 from gain=+10 by adjusting the potentiometer. Does it increases again but with an inverted output or does it goes pass G=1 towards G=0 which is the attentuation zone by normal understanding.

The schematics posted so far behaves like the latter and goes all the ways to G=0, then change sign and goes back up again to -10.

Does a circuit exists that goes from G=10 to G=1 and then from G=-1 goes to G=-10?

I don't think so because there is an abrupt sign change from +1 to -1 and this means a singularity somewhere.

 

[petesmc]

By Gain of -10 to 10 I meant that inverting.

Hence, Gain = 1, Vin = Vout,
Gain = 0.5 , Vout = Vin / 2

Gain = -1 Vout = -Vin

Gain = 0, Vout = 0

It seems there are two different conventions going on, of which I had no idea that existed. I've always used - to mean inverting.

Looks like some good ciruits have popped up that i should try and remember. Going to go simulate them now and see how they work out.

 

[JimB]

When I first read the original post, I jumped to the conclusion that the gain was to be variable from -10dB to +10dB.

It later occured to me that the gain was to be variable from -10 to +10. Not the same as the first statement.

When using dB, a negative gain is attenuation, a reduction in power.
A positive gain is a gain in power.

A negative gain when just specifying a numerical value infers that the amplifier is inverting, ie when the input goes more positive, the output goes more negative.

JimB

 

[Roff]

Edit:
Never mind - our OP has clarified his question.