Maker Pro
Maker Pro

Op-amp help please!!!

dislocations

Feb 21, 2012
25
Joined
Feb 21, 2012
Messages
25
I'd appreciate any help! Trying to breadboard an LM741 filter but don't understand the concept of two different power supplies.
 

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
Think of two batteries. The + of one is connected to the - of the other. That junction is then the ground. The unused + terminal is the positive supply, and the unused negative terminal is the negative supply. Two 9V batteries make a fine dual supply for op amps.

Bob
 
Last edited:

CDRIVE

Hauling 10' pipe on a Trek Shift3
May 8, 2012
4,960
Joined
May 8, 2012
Messages
4,960
There are two basic OpAmp categories, 'single ended' and 'double ended'. We use these terms to describe the power supply requirements for these two categories, which can be better described as unipolar and bipolar. A single ended OpAmp can only output a voltage from zero to some positive voltage. A 741 is a double ended OpAmp. Since it has both positive and negative supply voltages it's capable of outputting + and - voltages with respect to circuit ground.
 

Laplace

Apr 4, 2010
1,252
Joined
Apr 4, 2010
Messages
1,252
First, realize that all voltage-input operational amplifiers are inherently dual-supply - they expect to have a positive supply rail, a negative supply rail, and a ground halfway between the positive and negative rails. This is true even for op amps advertised as working with a single power supply. The most common method to obtain a positive rail, negative rail, and ground is with a positive power supply and a negative power supply with their grounds tied together. So how can an op amp work with a single supply? You just fake it out with a virtual ground. When using a single supply, the positive power becomes the positive rail while the power ground becomes the negative rail. A resistor voltage divider provides a virtual ground voltage halfway between the positive and negative rails. Since the input bias current of the '+' op amp terminal (the one usually grounded) is about 100 nA, it is reasonable for the divider resistors to be in the range of 10K. This works fine for the op amp input which is a high impedance load, but it will not work for high current op amp outputs where the signal is ground referenced, i.e., when the output is at the virtual ground voltage, then the output current is zero. For that you need a dual supply setup where the ground is a low impedance point.
 

Harald Kapp

Moderator
Moderator
Nov 17, 2011
13,700
Joined
Nov 17, 2011
Messages
13,700
@Laplace:

Not exactly right. I fully agree if you want to handle positive AND negative signals. A single supply OpAmp will need a virtual ground (usually, but not necessarily near Vcc/2) for this to work. A positive signal is represented by a voltage above virtual ground, a negative signal by a voltage below virtual ground.But all voltages are positive with respect to the power supply ground.

But there are applications that deal with positive signals only. Then you do not need a virtual ground. A rail-to-rail OpAmp does the trick in many of these applications.

Harald
 

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
First, realize that all voltage-input operational amplifiers are inherently dual-supply - they expect to have a positive supply rail, a negative supply rail, and a ground halfway between the positive and negative rails. This is true even for op amps advertised as working with a single power supply. The most common method to obtain a positive rail, negative rail, and ground is with a positive power supply and a negative power supply with their grounds tied together. So how can an op amp work with a single supply? You just fake it out with a virtual ground. When using a single supply, the positive power becomes the positive rail while the power ground becomes the negative rail. A resistor voltage divider provides a virtual ground voltage halfway between the positive and negative rails. Since the input bias current of the '+' op amp terminal (the one usually grounded) is about 100 nA, it is reasonable for the divider resistors to be in the range of 10K. This works fine for the op amp input which is a high impedance load, but it will not work for high current op amp outputs where the signal is ground referenced, i.e., when the output is at the virtual ground voltage, then the output current is zero. For that you need a dual supply setup where the ground is a low impedance point.
I like to think of the other way around. All opamps work off a single supply. If they did not, why don't they have 3 power connections? The difference between dual supply and single supply opamps is in how close to V- the inputs and output can go. Older one like the 741 can only get within about 2V of each rail. So it is pretty useless with a single supply IF the output needs to be near ground. On the other hand, if an application will always have the inputs and and output in a range of say 2V to V+-2V, the 741 will work as a single supply opamp.

Bob
 

Laplace

Apr 4, 2010
1,252
Joined
Apr 4, 2010
Messages
1,252
@Harald: Note that for the single-supply case I never claimed the negative rail had negative voltage with respect to power ground. In fact I said the negative rail was power ground.

@BobK: "All opamps work off a single supply." This is true, but not all op amp circuits will work correctly off a single supply, that is, unless you devise a low impedance virtual ground which you get by default with a dual power supply.
 

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
@BobK: "All opamps work off a single supply." This is true, but not all op amp circuits will work correctly off a single supply, that is, unless you devise a low impedance virtual ground which you get by default with a dual power supply.
Okay, then, tell my what is different about single supply op amps that allows them to work without a low-impedance virtual ground?

Bob
 

dislocations

Feb 21, 2012
25
Joined
Feb 21, 2012
Messages
25
Many many thanks again for you're knowledge. It's wonderful that you are able to share it with me in your own time!!!!
The reason I ask about op-amps in the first place is cos I'm building this:

http://www.electro-music.com/forum/viewtopic.php?highlight=lm741&t=3489
It's the hand drawn schematic if you scroll down the page.

Apparently it's a Sallen-Key Filter????

Ok, I've got this on the breadboard and it doesn't work till:
1. The + of the battery was connected to ground
2. The - of the battery was connected to pin 4

Not totally sure how this happened but it is definatley filtering the inputted sound DRAMATICALLY and not just adjusting the volume. It's exactly the kind of filter I've been looking for for some time. However when the dual pot goes down past a certain point the sound does die right down????
 

Laplace

Apr 4, 2010
1,252
Joined
Apr 4, 2010
Messages
1,252
@BobK: The point is that all op amps can be faked out with a virtual ground but if the op amp is in a circuit where high current will be expected to flow through the virtual ground, then the virtual ground must be a a low-impedance point. One way to make it a low impedance point is for it to be the power ground in a dual supply (but then we don't call it a virtual ground anymore).
 

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
@Laplace

Okay, please describe to me why this circuit does not work. You will note that there is no vrirtual ground in the circuit, and it is using a "dual supply" op amp. Also explain why it would start working if I put two 15V supplies instead of 1 30V supply.

The specification:

The circuit must buffer an input signal in the range of 10 to 20V. I.e. the output voltage should equal the input voltage.

Bob
 

Attachments

  • 741.JPG
    741.JPG
    20.3 KB · Views: 155

Laplace

Apr 4, 2010
1,252
Joined
Apr 4, 2010
Messages
1,252
@BobK: "Okay, please describe to me why this circuit does not work." Why do you believe that a voltage follower would not work?
 

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
Because you told me op amps only work with a single supply when you provide a low impedance virtual ground, and this circuit does not do that. (I don't think it would not work, I am disputing your previous statement)

Bob
 

dislocations

Feb 21, 2012
25
Joined
Feb 21, 2012
Messages
25
Just re-thought my idea...........the dual supply was a pain and achieving excellent results with an LM386. Basically feeding the output back to the input with an RC network. May be not the best way of doing the job but it works! I'll revisit the op-amp method at a later date. Many thanks again all!!!!!!!!!
 

CDRIVE

Hauling 10' pipe on a Trek Shift3
May 8, 2012
4,960
Joined
May 8, 2012
Messages
4,960
Just re-thought my idea...........<snip>....... I'll revisit the op-amp method at a later date. Many thanks again all!!!!!!!!!

I realize this is your topic but to call the fight when we might have had a TKO is a major bummer. Please, just a standing 8 count. Discussions like this are stimulating, educational and can be exciting; especially when you're just an observer. Let Bob & Laplace continue.

DING! :D
 

KrisBlueNZ

Sadly passed away in 2015
Nov 28, 2011
8,393
Joined
Nov 28, 2011
Messages
8,393
Re the argument between BobK and Laplace. Perhaps you can both agree with this description.

Apart from some very special-purpose devices, op-amps have two power pins: positive and negative. They have two input pins (inverting and non-inverting) and one output, and they produce a voltage at the output that is (ideally) equal to the difference between the input pin voltages, multiplied by the voltage gain (but the op-amp cannot drive its output voltage outside an upper and lower limit, which are determined by the supply voltages).

Op-amps can be used with several voltage conventions. Three common schemes are:
(a) split supply: the circuit has a 0V (ground) rail, a positive supply rail, and a negative supply rail, often provided by two separate batteries; the op-amp is powered from the positive and negative rails, and signals are referenced to the 0V rail;
(b) derived ground rail (or "virtual ground"): a single voltage source is connected across the op-amp's positive and negative supply pins, and a voltage divider, buffered voltage divider, or similar circuit provides a derived "ground" rail at some in-between voltage, often half way between the positive and negative supplies; signals are referenced to this derived ground rail;
(c) single supply circuits without a derived ground rail, where the negative supply rail (which is connected to the negative supply pin of the op-amp) is considered to be the reference or 0V rail.

In cases (a) and (b), the voltages seen at the op-amp's input and output pins are measured relative to the ground rail, and usually don't go very far above or below it (i.e. not far positive or negative relative to it). If these voltages go too far from the ground rail, the signal will be clipped, because of the limited positive and negative voltages that can be accepted and/or generated by the op-amp, which are limited by the positive and negative supply rail voltages. In cases (a) and (b), neither the inputs nor the output of the op-amp will normally see voltages approaching the positive and negative supply rails. All op-amps will work in these situations, including the 741, whose input and output voltages can't be closer than about 2V to the positive or negative supply voltage rails for correct operation.

In case (c), the voltages applied to the op-amp's inputs, and taken from the op-amp's output, may need to be equal to the ground rail, which in this case is the negative supply pin of the op-amp. For these applications, a "single-supply" op-amp is needed; these have the special feature that "the input common-mode voltage range includes the negative supply rail", which means that the op-amp will still correctly amplify the voltage difference between its inputs even when one or both inputs is at the ground voltage (the op-amp's negative supply voltage), or slightly below it (more negative than the negative supply rail). Single-supply op-amps are typically specified to operate with their inputs up to 0.3~0.6V more negative than the negative supply rail.
Common op-amps that are single-supply compatible are the LM324 (quad) and LM358 (dual). The 741 is NOT single-supply compatible.
Single-supply op-amps are also usually specified with outputs that go down to the negative supply, though this normally only applies when the output is sourcing current; if you try to "pull" such an output towards the positive supply with a load resistance, it will not usually be able to pull itself fully to the negative supply.

In some designs, a significant load may be presented to the op-amp. A headphone amplifier with a DC-coupled output is a good example. The output needs to be able to swing positive and negative relative to the ground rail, and there is also a significant load current (the current through the headphone transducers) that flows into the ground rail when it returns. In this case, a true split supply (case (a) above) is normally used, because a derived ground source typically cannot source or sink these significant currents while still providing a clean and stable earth reference.

These same considerations apply to comparators, but most comparators are single-supply-compatible because they are often used in circuits that don't have a ground rail between the supply rails (these arrangements, cases (a) and (b), are more common in audio circuits, and comparators can be used in all kinds of circuits).
 

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
Yes, I agree with everything you said. My point has always been the that only difference between a "dual supply" and "single supply" op amp is how close the inputs and outputs get get to V-, and I see nothing in your post that disputes that.

Bob
 
Top