op amp questions

Thread Starter

patrick

Joined May 12, 2004
3
Hi, I'm confused about the voltage gain of an op amp. I thought the voltage gain of a closed loop configuration was flat in frequency responce until the new fc (given by the open loop curve), am I right?
so what's the meaning of the difference between open loop and closed loop gains? why should I care about it?

Thanks
Patrick
 

Dave

Joined Nov 17, 2003
6,969
Open loop gain refers to the amplifier gain when there is no feedback in the amplifer circuit. Op-amps cicuits with no feedback tend to be rather unpredictable with varying gain values depending on a number of external factors (for example temperature), and thus are very seldom used in practical circuits.

Closed loop gain refers to the amplifier gain where there is feedback in the circuit, whether it be inverting or non-inverting feedback. The objective of providing feedback is to stabilise the performance of the amplifier circuit so that it is practically useful when designing an electronic system. The closed loop gain, noted Gv is defined by the resistor values that comprise the feedback network.

For an inverting amplifer:

Gv = -R2/R1

And for the Non-inverting amplier:

Gv = (R1+R2) / R1

Obviously these are very basic (yet fundamental) equations, and are to be considered in the context of the amplifier circuit which you are designing. You may find the following link here on All About Circuits useful on this topic:

Operational Amplifers

You will find the first few sections should cover what I have explained above in more detail with diagrams. Hope that helps a little :)
 

mozikluv

Joined Jan 22, 2004
1,435
hi patrick,

you said in your post "why should you care about open/closed loop gain" of an op-amp. my reply is you should because you wont be able to do any good with your op-amp if you do not consider this parameter. maybe you are referring to the inverting/non-inverting voltage gain formula. you apply it and your device now functions, right? that formula already consider the loop gain of your amp.

to make your op-amp operate at its best there are other parameters to consider like input offset voltage, input bias current, frequency response and many more.

it must be remembered that the dominating error in a low source resistance is the amp offset voltage, while for high source resistance the dominating error term is the offset current. so to minimize the offset voltage error, we must have an pffset voltage at the amp output equal to the closed loop gain multiplied by the offset voltage at the amp input.

what i have cited plays an important role in the gain-frequency trait of the op-amp and its feedback network which by the way prevents oscillation.

we must also bear in mind that the phase shift in the circuit must never exceed 180 degrees for any particular frequency whenever the gain and the feedback network is greater than unity.

as what dave has said check the articles about op-amps in this forum. maybe my reply here is already duplicating what is presented in the op amp column.
 

Thread Starter

patrick

Joined May 12, 2004
3
thanks for your help, but this is my problem:

I'm reading the book Op Amp in Audio by Jung, and it says that if I want to have 60dB of gain "with an accuracy of 1%" (???) at 10KHz, there must be 40dB of feedback, so the op amp must have 100dB at 10KHz in the open loop curve, or in other words, 1000MHz of gain-bandwidth product GBW.

He defines the difference between open and close curves (in dB) like the amount of feedback, which takes sence if we think about the buffer configuration (maximum feedback).

But why should I care about the amount of feedback? why should I need a op amp with 1000MHz of GBW ( 100dB x 10KHz ) to have a gain of 60dB @ 10KHz? I thought if I had 10Mhz of GBW I could have this gain at 10KHz ( 60dB = 1000 ; 1000 x 10KHz = 10MHz )

Patrick
 

mozikluv

Joined Jan 22, 2004
1,435
hi

you are probably wondering why such a very high gain, that's because that very high gain is good only for low frequencies. with the internal compensation capacitor of the op-amp, the response to any frequency above 10Hz will be rolled off. meaning, as frequency increases there would less feedback to correct the internal distortion. furthermore, the output impedance increases with increasing frequency. if you use an oscilo. you will notice that the op-amp will be hard up trying to maintain an output starting from about 17khz up to 20khz and above 20khz there is no feedback to talk about. so even at 10khz the gain has drop significantly and there is little feedback available.

now do you see the point why the 60dB gain? but in real life that gain will be reduce by several factors.
 

Thread Starter

patrick

Joined May 12, 2004
3
hi mozikluv, thanks but I still don't understand... you said

"even at 10khz the gain has drop significantly and there is little feedback available"

why this happened, suppose I use a 16MHz GBW op amp? what's the relation between amount of feedback and gain at X frequency? I understand that the external resistors fix the amount of gain....

Jung talks about 60dB of gain in a mic preamp application... also, what factors would reduce this gain in real life?

Patrick
 

mozikluv

Joined Jan 22, 2004
1,435
hi patrick,

usually when we start talking about gains and very good accuracy, we are then talking about precision circuits. but when out circuit is for audio purposes, precision need not be a dominating factor.

i will not talk here of technicalities, but rather on down to earth knowledge. if you want better sound quality we use a wideband op-amp for the simple reason that this device has more elbow room for frequencies beyond the audio range. we must also remember that there is no VFB op amp that has an open loop bandwidth that goes over 1khz. because of that this kind of device will have what you call frequency dependent decrease in the feedback. as a result there will be an increase in distortion and output impedance will also increase. but this does not mean that we should not use this device, as i have said if you want better sound quality use of wideband op-amp is the choice. again we must remember that sound is usually made via mechnical or biological ways and they dont possess fast transients.

for microphone pre-amps a gain of 10 to 20 would suffice and for other pre-amps a gain of 1 to 22 would suffice.
 

beenthere

Joined Apr 20, 2004
15,819
Hi Patrick,

More on the loss of gain with frequency. While the DC gain of the op amp is set by the feedback resistor(s), there is an internal capacitor in the op amp that is in parallel with the feedback resistor. Remember that Xc decreases with frequency.

This gives any op amp circuit with significant gain the quality of a low pass filter. At the frequency where the internal cap starts to have its Xc equal the feedback resistor, gain is down 3dB. It will fall at 3dB/octave, and can even go <1 depending on the external resistors.

Frankly, I'd be worried about getting that much gain out of one device. I've been building amps with gains in the 40 dB range using two gain stages with good results. Of couse, I don't have to pass above 1 kHz....
 

mozikluv

Joined Jan 22, 2004
1,435
hi,

thanx to "beenthere" for pointing out the internal cap of an op-amp, i forgot to mention that in my response. this internal cap has a direct relationship to the op-amp gain which beenthere cited which we call capacitive reactance. as i have said before the higher the frequency the lesser the feedback.

in an inverting circuit, the input impedance is equal to the input resistor and the closed loop bandwidth is equal to unity-gain frequency divided by the closed loop gain plus 1.

for a non-inverting circuit, the input impedance is equal to the differential impedance x the loop gain. the closed loop gain is equal to the ratio of the total of the ground resistor plus the feedback resistor to the ground resistor. the source resitance should be equal to the parallel resistance of the ground resistor and feedback resistor.
 

mozikluv

Joined Jan 22, 2004
1,435
hi,

hereare some factors that affect efficiency of an op-amp;
1. tolerance of parts used
2. operating temperature
3. stray capacitance
4. ground lay-out
5. magnetic fields or harmonics generating from transformer or rectifiers
6. thermal noise
7 rf interference

these are some of the few.
 
Top