was Re:Ir-remote Now is: Op Amp discussion

[audioguru2]

Hotwaterwizard,
Our project's receiver has an opamp without an input bias voltage. The receiver will probably be a lot more sensitive if pin3 is biased to half-supply with a couple of 2.2M resistors.

 

[nomad13]

if u want to expand your ir application why not try a microcontroller decoded ir since you can have as many as 255 devices to control. i have used 16f84 and use a sony remote as my controller. i control my toy car using the volume up & down and the channel up & down. i appreciate it if some 1 could give me other coding techniques for different infrared remote. thanks

 

[MP1]

Hotwaterwizard,
Our project's receiver has an opamp without an input bias voltage. The receiver will probably be a lot more sensitive if pin3 is biased to half-supply with a couple of 2.2M resistors.

An op-amp does not require an input bias voltage. I see no problem with the circuit "as is".

MP

 

[audioguru2]

MP,
Of course opamps need an input bias voltage.
Without it, the floating input is pulled to the positive rail by the uncharged input capacitor during power-up, then the input is pulled to the negative rail by the base bias current of the opamp's NPN input transistor. Either condition results in the opamp performing as a rectifier, requiring many volts of input signal to bring the input to within its linear voltage range. With a proper input bias voltage, the opamp is a very sensitive and linear amplifier.

I errored earlier when I recommended 2 resistors to divide the supply. Our project actually has bipolar supplies, requiring only 1 resistor to ground for the opamp's input bias voltage, which does the same thing.

 

[MP1]

Op amps have an input impedance of approximately 2 Meg ohms to ground. The input of an op-amp is not floating. This is an application where you want a high input impedance. I have not tried this circuit and I cannot tell you if it functions as is without a little bench work. But if you are condemning the circuit for this reason, it is the wrong reason. Also note the function of this circuit. It is not a linear type of circuit.

MP

 

[audioguru2]

MP,
Of course the input of an opamp is floating. It is waiting for a designer to connect it to a voltage reference, drawing an extremely low current.
If you are not going to properly voltage-bias the opamp by simply adding ONLY ONE resistor, then it cannot provide its gain of 101, as determined by its feedback, and allow the project to have a high sensitivity. That is not condemming the circuit, but only making a technical recommendation. If you are not allowing the opamp to operate as intended, then it might as well be eliminated.
The photo-transistor is capable of driving the very-sensitive, but obsolete, filter IC by itself.

 

[MP1]

audioguru, the term "floating input" refers to nothing connected. Any 741 or similar type op-amp will have an input impedance of approximately 2 MegOhms on an input when the other input pin is connected to ground. This would tell most that the input is not floating. You also have a resistance value to work with. You have a strange impression of "how an op-amp is intended" to be used. There is no specified intended use. There are a wide variety of uses. The formula for gain of an op-amp does not factor in a resistor with a bias voltage. If it were a necessary part of the circuit, the data sheet would indicate this.
i have attached some examples of circuits without the bias you claim is a must:

View attachment 35438

 

[MP1]

...and here:

View attachment 35439

 

[MP1]

The last post was part of the fron end of a well known Fender Bass amp. The one before that is a well known guitat effect. Here I have added another well tested and favored guitar effect.
Although you will see different configurations, you will note from the examples that all op amp uses do not need a bias voltage.

MP

View attachment 35440

 

[audioguru2]

MP,
Look again. ALL of the opamps in your schematics have their non-inverting input biased to ground:
1) A direct connection to ground.
2) DC coupling to a preceding stage that has its input biased to ground.
3) A resistor to ground.

The latter is what I recommended in the IR project. A similar opamp application is C!a in your last schematic, which also shows the non-inverting input biased to ground through a resistor.

The input impedance of an opamp is caused by its input transistor's base current that flows through your bias resistor. A FET-input opamp has an extremely high input impedance but also needs its input voltage defined, not floating.

The gain of an opamp is determined by the feedback ratio that is applied to its inverting input, which is biased by the feedback divider from its output. The output, in turn, simply follows the voltage that is applied to its non-inverting input. Therefore this input MUST be biased to a defined voltage.

If you think that opamps have built-in input biasing, look at the naked base of the input transistor of a 741 opamp on page 4 here:
http://www.national.com/ds/LM/LM741.pdf

 

[MP1]

audioguru, I gave you a mixed variety. Some do and some do not. Look at all the op-amps, not just some of them. One simple use that I have not shown is a simple buffer/follower.
I know that you will not be convinced, even with the elaborate explanation that I gave you. It's rather funny, because I have built many applications using op-amps without your bias resistors and they are functioning in electronics all over the world.
No, you do not have built in bias in an op-amp, but an op-amp is not like a big transistor. You do however have about 2 meg on the inputs and when you connect your device, you are completing the circuit.
There are many applications for op-amps and your use is only a small limited area. But if it works for you, go ahead and stay with it.
Just don't tell everyone it can't be done.

MP

 

[audioguru2]

Some do and some do not. Look at all the op-amps, not just some of them. One simple use that I have not shown is a simple buffer/follower.

MP,
Look even again. Not a single opamp in your schematics has a floating input, like our original IR project. Of course, in many cases (inverting amplifiers) the non-inverting input doesn't use a resistor for bias, but is simply connected directly to ground for its voltage reference. The same applies for DC-coupled non-inverting opamps, a resistor is not necessary since the preceeding stage provides the voltage reference.

Show me a floating input on a buffer/follower opamp or any other opamp circuit.
Try it. Make a 741 circuit with a small (10nF) input capacitor but no bias resistor, and watch its output sink to near the negative supply, and stay there, just like our project.
No, an opamp CANNOT have a floating input to work properly.

 

[MP1]

None of them are like the IR project. But several are not using a set of resistors biasing the input of a pin with 2 resistors between V+ and GND as you have described is necessary. Also note that none of them are passing a digital 40 KHZ signal. Biasing an op-amp determines the zero crossing of the wave, not the amplitude.

I am going to quit on this thread now as it is only exhausting and a waste of my time. It is not my responsibility to educate you or prove anything to you when you are not right. Since you have become a member of this community, I have spent more time with your wasteful arguments than posting things that I would like to share with the group.

MP

 

[audioguru2]

Would any other members like to join this discussion about opamp biasing?

 

[ante1]

audioguru,

This will be a long answer. I sometimes get the feeling that you don

 

[audioguru2]

Ante,
It always starts with me making a suggestion to make a faulty circuit better. Here I wanted the first "helper" to join me in a discussion about improvements. Is he miserable because I replied to him and made further recommendations along with his?
Then I am wrongly told that my fix is wrong by someone else who stepped in.
I am not backstabbing, I am simply standing-up for the way that circuits should be, then much of my proof gets deleted.
I am the one who has been backstabbed, since you and the rest of our community did not join-in.
Isn't this an adult forum for discussions?

 

[audioguru2]

Our project's receiver has an opamp without an input bias voltage. The receiver will probably be a lot more sensitive if pin3 is biased to half-supply with a couple of 2.2M resistors, but since it has bi-polar supplies, a single 1M resistor to ground.

To All,
I have discovered a very interesting University Tutorial document describing the history, details and applications of opamps. Notice that 2/3 the way down, beside figure 10, the author states, "As a rule of thumb, an opamp circuit will not work AT ALL (his accent) unless both inputs have direct-current return path to ground of a similar reference."

Isn't that what I said?

The tutorial is here:
http://www.uoguelph.ca/~antoon/gadgets/741/741.html

 

[ante1]

audioguru,

This is just one example

 

[wired1]

Would any other members like to join this discussion about opamp biasing?

I do not see where there is a discussion. This is a cut and dry subject. Adding a bias resistor assumes that all op amps are powered from a single supply. This is not the case in today