help needed for opamps

[deepakchikane]

Dear all,

can anybody tell me selection of opamps resistors..??

on wot basis the resistors are been decided from ohms to megaohms..

i am looking for the answers as i coudnt get the idea to select a proper resistors & capacitors...

plz plz plz

 

[KerimF]

Opamps could be used in too many configurations. Perhaps you like to start with the simplest one that you heard of.

 

[deepakchikane]

but can u give me one opamp circuits by sating every resistors inportance wrt to its value...

 

[LvW]

but can u give me one opamp circuits by sating every resistors inportance wrt to its value...

Here are some general rules for selecting resistor values:
1.) Do not use values which are too low (some ohms). In this case, the resistance of connection lines could produce unwanted errors.
2.) Do not use values which are too large. In this case, parasitic effects (unwanted capacitances) could influence the behaviour.
3.) Resistor values should be large in comparison with the output resistance of the opamp (at least factor 50...100)
4.) Resistor values should be small in comparison with the input resistance of the opamp.

As a result the preferred range is between 500 ohms and 1 Megohm.
Of course, this is just a very general rule of thumb. In some cases one can or must deviate from this rule. The actual selection depends also on the opamp type (Fet input or BJT input).

 

[tgootee]

Also remember that larger resistance values produce more noise. So if low noise is important, use the lowest resistor values that won't cause problems.

 

[goldsmith]

Also remember that larger resistance values produce more noise. So if low noise is important, use the lowest resistor values that won't cause problems.

Hi Hi tgootee
Of course it depends on the BW of the input signal which i don't think in him/her case it is important .

Regards
Goldsmith

can anybody tell me selection of opamps resistors..??

on wot basis the resistors are been decided from ohms to megaohms..

i am looking for the answers as i coudnt get the idea to select a proper resistors & capacitors...

Hi deepakchikane
Certainly but may i ask you what kind of circuit you're referring to ? amplifier ? integrator ? differentiator ? .... ?
It is really hard to answer your question but "LvW" gave you some *** criteria for general mode .
Best Wishes
Goldsmith

 

[tgootee]

Actually, the resistor NOISE only depends on the resistance. Resistors produce their own noise, whether or not they are even powered.

 

[zoulzubazz]

mate op amps have numerous applications and feedback resistors have a big picture to play in most op-amp circuit. here is a youtube video explaining the very basics of op amp circuit. hope it helps.

https://www.youtube.com/watch?v=6ok-ruHlopE

 

[steveelliott]

How about simple guidance?
Feedback resistor for most opamps should be 10k-100k range to produce up to 1ma or so output current. If you think there is something different about your opamp or circuit, look at the manufacturer's data sheets in the application notes/examples section of the datasheet.

There are a few "exotic" opamps that have very peculiar impedance requirements, but you'll probably never use them until you've been building opamp circuits for several years. (Of course, there's always an exception - the first opamp circuit *I* built professionally used vacuum tubes and had a 4000V output - it was used for steering the 800meV proton beam in the Los Alamos Meson Physics Facility accelerator - its feedback resistor was 50Mohms).

 

[kam1787]

Resistors produce their own noise, whether or not they are even powered.

I put several resistors on the table, and listened very carefully but could not hear anything.

 

[tgootee]

I put several resistors on the table, and listened very carefully but could not hear anything.

Half of your resistors must have had negative values, so that the noise cancelled-out.

https://en.wikipedia.org/wiki/Johnson–Nyquist_noise

One reason to care about the noise power that is generated by a high-value resistor is that it represents the limit for the smallest voltage we can resolve across the resistor, in a certain bandwidth.

See the seventh slide, at:

**broken link removed**

For a 10k resistor, the noise voltage is 13 μV (rms), for a 1 MHz bandwidth. And that's only the thermal noise, with or without any current flowing.

A more-obvious reason: In a high-gain audio amplifier application, high-value resistors can result in audible noise at the output.

For a low-noise application, I usually try to use only resistors that are well-under 10k Ohms, when possible, where it matters.

The thermal noise of resistors is an often-used criterion for choosing resistor values for op amp circuits.

 

[goldsmith]

Actually, the resistor NOISE only depends on the resistance. Resistors produce their own noise, whether or not they are even powered.

Hi again
Sorry but i'm disagree with you .
Vn= sqrt(4*k*t*bw*R ) where k is boltzman coefficient which is 1.38*10^-23 so it depends on some of the parameters .

I put several resistors on the table, and listened very carefully but could not hear anything.

Are you kidding us ? the noise is called Johnson–Nyquist noise ! it's amplitude is so limited and you can not hear that ! ok ?

Feedback resistor for most opamps should be 10k-100k range to produce up to 1ma or so output current

Completely wrong ! feedback resistor can select the gain and some of the other things . who told 1 ma is our destination ? do you know what is purpose of feedback resistor in an inverting amplifier with op amp ? or do you know what is the duty of a resistor in feedback path in a damped integrator ?



Best wishes to all !
Goldsmith

 

[tgootee]

Dear all,

can anybody tell me selection of opamps resistors..??

on wot basis the resistors are been decided from ohms to megaohms..

i am looking for the answers as i coudnt get the idea to select a proper resistors & capacitors...

plz plz plz

There are also inputs' bias currents and offset voltages to worry about, with many op amps (Check for differences between op amps with FET inputs and BJT inputs). And the resistor values you use can determine how those might affect your circuit's behaviors. In some types of circuits, it can be important. For example, maybe sometimes you would want the impedances seen by the two inputs to be matched, to minimize the effects (at the output) of the input bias currents. Those are sometimes reasons, for example, to use a resistor in the feedback loop of a unity-gain op amp buffer, or in series with an input that might usually be tied directly to ground.

This CLASSIC application note will explain a lot about how the resistor values are chosen:

https://www.ti.com/lit/an/snoa621c/snoa621c.pdf


And this one is an essential old classic, too:

**broken link removed**


Here is one of Walt Jung's (Analog Devices) books, on line for free ("Op Amp Applications Handbook"):

https://www.analog.com/library/analogDialogue/archives/39-05/op_amp_applications_handbook.html


Here is another great free book ("Op Amps for Everyone"):

**broken link removed**


Those books can answer all of the questions you have asked, here, and far more.

Cheers,

Tom

- - - Updated - - -

Besides the resistor values, one extremely important thing that you will need to do, to get your op amp circuits to work correctly (and sometimes just to get them to work at all), is to ALWAYS use bypass and decoupling capacitors, at the power supply pins.

You should always connect a physically-small capacitor very close to each power pin, usually one from each power pin to ground. These are bypass caps and are needed for stability, i.e. so the circuit won't oscillate or "ring" (oscillation that dies out). The bypass capacitors need to connect VERY close to the power pins; within 2 mm or less is good. Their connections and lead lengths should be as short as possible, to minimize stray inductance, so they will be able to be effective for high frequencies. If in doubt, just use an X7R ceramic cap of 0.1 uF or 0.01 uF. (The reason these bypass caps are needed is that there is usually a hidden positive feedback loop through the power rail, and the cap will short the HF to ground so the circuit won't oscillate at high frequency.)

You should also always connect a larger cap from each power supply pin to ground. This is for decoupling. It acts as a small, local "point-of-load" current supply, for fast transient currents that are needed by the power pin. If these transient currents had to try to come through the power rail conductor, all the way from the main power supply, they would induce voltage spikes across the parasitic self-inductance of the conductor, possibly polluting everything else connected to the power supply, and might also not be delivered to the device's power pin with accurate amplitude or timing. if you're just building an op amp circuit with no really-demanding requirements, you can usually just use a 10 uF or larger electrolytic, for the decoupling cap for each power pin. It doesn't need to be as close to the power pin as the bypass cap. But it should be as close as possible.

Later, you can learn to easily calculate the required values of the decoupling capacitors, and the maximum lengths that their power pin connections should be allowed to be.

.

 

[deepakchikane]

Hi Hi tgootee
Of course it depends on the BW of the input signal which i don't think in him/her case it is important .

Regards
Goldsmith




Hi deepakchikane
Certainly but may i ask you what kind of circuit you're referring to ? amplifier ? integrator ? differentiator ? .... ?
It is really hard to answer your question but "LvW" gave you some *** criteria for general mode .
Best Wishes
Goldsmith

Hi GoldSmith..

Basically i am designing a zero crossing ckt for current..

I AM DESIGNING ONE DIFFERENTIAL AMPLIFIER WHOSE GAIN WILL BE 100 (I/P=SMALL milivolts AC shunt signal)
THE 100 * GAIN AC SIGNAL WILL BE CAMPARE WITH THE VREF+0.6V & VREF-0.6V (VREF=ZERO CROSSING OF AC SIGNAL)...

THE ZERO CROSSING OF AC SIGNAL I AM GIVING FOR THE MOC 3052 WHICH WILL DRIVES SATIESFACTORY MY CAPACITIVE LOADS TYPES

 

[goldsmith]

Hi GoldSmith..

Basically i am designing a zero crossing ckt for current..

Hi deepakchikane
Well , but may you show us a schematic which you're referring to ? ( i hope you know there are many of the circuits with aim of zero cross detection and for judgement regarding them we need to see what you're trying to refer . isn't it ?
Best Wishes
Goldsmith

 

[steveelliott]

Hi again
Sorry but i'm disagree with you .
Vn= sqrt(4*k*t*bw*R ) where k is boltzman coefficient which is 1.38*10^-23 so it depends on some of the parameters .



Are you kidding us ? the noise is called Johnson–Nyquist noise ! it's amplitude is so limited and you can not hear that ! ok ?

Completely wrong ! feedback resistor can select the gain and some of the other things . who told 1 ma is our destination ? do you know what is purpose of feedback resistor in an inverting amplifier with op amp ? or do you know what is the duty of a resistor in feedback path in a damped integrator ?



Best wishes to all !
Goldsmith

1 ma is within the typical current capability of opamps. resistor values in the previously suggested 500 ohm range will require 20mA of current with 10V output - that's too much for most common opamps. "do you know what is purpose of feedback resistor in an inverting amplifier with op amp ?" Its purpose is to provide a voltage signal back to the inverting input of the linear differential operational amplifier which is subtracted from the voltage at the non-inverting input of the linear differential operational amplifier and the resulting difference voltage is multiplied by the linear differential operational amplifier's open loop gain - yea I think I do. "What is the duty of a resistor in feedback path in a damped integrator ? " Again using the same previously named 500 ohm resistor, at the corner frequency of a "damped integrator" the magnitude of the impedance is only 707 ohm. Depending on the input signal and the function of the "damped integrator" in the overall circuit (phase compensation, PI loop, etc), "high" output voltages could be required, beyond the current capability of the opamp. I dare you to say that when you are designing a common x10 inverting amplifier, the first thing that comes to mind is Ri = 2.2Meg and Rf = 22Meg or 10ohms/100ohms - instead it will be 10k/100k or 1k/10k or 4.7k/47k and you know it.

 

[kam1787]

Half of your resistors must have had negative values, so that the noise cancelled-out.

That explains it. so if we matched positve and negative resistors there would be zero noise?

If I apply pulsating voltage across a 66 mega-ohm resistor would I get R&Bs?

 

[yuvan]

the noise from the resistors are all uncorrelated; only possible thing to do will be reduction of noise amplitude by proper resistor values or by proper filtering in our frequency of interest. Reduction is possible & not elimination

 

[FvM]

That explains it. so if we matched positve and negative resistors there would be zero noise?

Hopefully everyone understands the irony. Some might still wonder where to get a negative resistor.

 

[tgootee]

Quote Originally Posted by kam1787 View Post

I put several resistors on the table, and listened very carefully but could not hear anything.

Originally Posted by tgootee View Post

Half of your resistors must have had negative values, so that the noise cancelled-out

That explains it. so if we matched positve and negative resistors there would be zero noise?

If I apply pulsating voltage across a 66 mega-ohm resistor would I get R&Bs?

Hopefully everyone understands the irony. Some might still wonder where to get a negative resistor.

You can simply flip them the other way around, so they have the opposite polarity.





(Yes, it's all just a running joke. Apologies to the OP.)

- - - Updated - - -

- - - Updated - - -

1 ma is within the typical current capability of opamps. resistor values in the previously suggested 500 ohm range will require 20mA of current with 10V output - that's too much for most common opamps. "do you know what is purpose of feedback resistor in an inverting amplifier with op amp ?" Its purpose is to provide a voltage signal back to the inverting input of the linear differential operational amplifier which is subtracted from the voltage at the non-inverting input of the linear differential operational amplifier and the resulting difference voltage is multiplied by the linear differential operational amplifier's open loop gain - yea I think I do. "What is the duty of a resistor in feedback path in a damped integrator ? " Again using the same previously named 500 ohm resistor, at the corner frequency of a "damped integrator" the magnitude of the impedance is only 707 ohm. Depending on the input signal and the function of the "damped integrator" in the overall circuit (phase compensation, PI loop, etc), "high" output voltages could be required, beyond the current capability of the opamp. I dare you to say that when you are designing a common x10 inverting amplifier, the first thing that comes to mind is Ri = 2.2Meg and Rf = 22Meg or 10ohms/100ohms - instead it will be 10k/100k or 1k/10k or 4.7k/47k and you know it.

1 mA?!

That is certainly "within" the typical opamp current capability; WELL within.

I went to the Analog Devices website's op amp selection table, which only displayed the first 100 devices, apparently in no particular order. I exported it to Excel and averaged the Iout drive capability.

The average output current capability is 53.3 mA, for those 100 op amps.

An image of Excel sheet is attached.