Op-amp circuit help

[axro]

I'm just playing around with a TL081 to try and learn how to use them. I'll trying to use it as a simple comparator and I can't figure out why this is happening.

I'm using a 6V PSU. The Inverting input of the opamp connected to the middle point of 2 10K resistors. So the inverting input is getting 3V and all times.

I have the non-inverting input connect to the wiper of a 100K pot.

VCC+ is connected to 6V and VCC- is connected to Ground.

I have the output connected to a Green LED with a resistor in series.

Whats happening is, the only time the LED is off is when the wiper is in the middle. If I turn the knob to either far end(6V or 0v) the LED is lit. Why is this happening?

 

[audioguru]

The minimum supply for a TL081 is 7V. Your supply voltage is a little too low.

When the wiper of the pot is within a few volts from the negative supply (ground) then the problem of Opamp Phase Inversion causes the output to go high. The output is supposed to go high when the non-inverting input (the wiper of the pot) is higher than the inverting input.

 

[nigelwright7557]

Try using a rail to rail op amp

 

[axro]

Where did you find the 7v on the datasheet? All I ever see is a max of - + 18 Which I would assume means up to 36V if your VCC- is ground.

Is there a way to cure that phase inversion if that is the cause? Or are you thinking the low supply voltage is the cause?

 

[audioguru]

Where did you find the 7v on the datasheet?

Texas Instruments sales sheet for their TL081 is here:
Operational Amplifier (Op Amp) - Standard Linear Amplifier - TL081 - TI.com

Is there a way to cure that phase inversion if that is the cause? Or are you thinking the low supply voltage is the cause?

No it can't be fixed and has nothing to do with the supply voltage. The inputs do it when they get near the negative supply voltage (the common mode voltage limit). Newer opamps do not have the problem.

 

[axro]

I thought the TL081 was a new op amp? What is a newer general purpose op-amp that I could use that doesn't experience phase inversion?

 

[audioguru]

The OPA2141 or OPA134 opamps do not have the phase inversion problem.

 

[axro]

Wow. Those are WAY more expensive then the TL081.

Would this be an effective way of curing the phase inversion problem? Put another resistor on the Ground side of the Pot. This way, I can still get 6V Max but I could adjust my min to be 2V(or w/e). Then I won't be so close to VCC-.

Is there a set number that phase inversion happens at? How many volts from VCC- do you have to be for it to occur?

 

[audioguru]

Each TL081 opamp has a different input negative common-mode voltage limit.
The datasheet shows 4V max or 3V typically above the negative supply when the supply is plus and minus 15V.

There are thousands of opamps without the phase inversion problem. Most do not have Jfet inputs.

 

[axro]

So my idea won't really work then, because it inverts and such a high voltage?

Also what are some cheaper Opamps that don't have that problem? Should I stay away from all Jfet opamps?

Would an LM358 work? It's a dual amp and I don't really need that, but I could just not use the second one. Is there a single amp version of this?

 

[audioguru]

I have used TL071, TL072 and TL074 audio opamps for many years in audio circuits with no problems. They are a TL08x selected for low noise. I used them usually with 12V or plus and minus 6V supplies.

The LM833 dual audio opamp is also good and does not have the phase inversion problem. Its minimum supply is 10V.

I have used the MC33171, MC33172 and MC33174 opamps to replace LM358 and LM324 opamps because they have many of the same spec's (minimum supply of only 3V and the same low supply current) but have much better performance. They also do not have the phase inversion problem.

The MC34071, MC34072 and MC34074 opamps also have a minimum supply of only 3V but have "normal" supply current and wider bandwidth. They also do not have the phase inversion problem.

 

[axro]

Alright, thanks for the options.

What is input common-mode. I see it alot of places but I can't find a definition.

 

[audioguru]

The input common-mode voltage range for an opamp is the range of voltages that the inputs will work. For the TL08x opamps it is 3V or 4V above the negative supply pin but goes up to the positive supply pin.

The LM358, LM324, MC3317x, MC3407x and TLE214x opamps have an input common-mode voltage range down to the negative supply pin which is 0V in many circuits.

Most Cmos opamps have an input common-mode voltage range from the negative supply pin to the positive supply pin (rail-to-rail).

 

[axro]

Slightly unrelated. JFETS are voltage devices and BJTs are current devices. How can a BJT work in an opamp when the inputs of a opamp are voltage sensative not current sensative.

 

[audioguru]

Slightly unrelated. JFETS are voltage devices and BJTs are current devices. How can a BJT work in an opamp when the inputs of a opamp are voltage sensative not current sensative.

Ohm's Law says that when you apply a current in a resistor then a voltage is produced. The input current of opamps is listed in their datasheet.

 

[axro]

Do you ever have to put resistors inline with an input? If you connected in input directly to VCC would it short or burn out or anything.

 

[audioguru]

Do you ever have to put resistors inline with an input? If you connected in input directly to VCC would it short or burn out or anything.

The emitters of the input transistors of an opamp are not connected directly to ground or the negative supply. They are in a very complicated circuit so that the input transistors have an input common-mode voltage range of at least from 1.5V above the negative supply to at least 1.5V below the positive supply. The input resistance is not just a few thousand ohms of a transistor but is hundreds of thousand ohms or millions of ohms of the circuit.

 

[axro]

So from what I have seen, is this right

JFET Op-amps can't go down to VCC- but can go up to VCC+ on the inputs

BJT Opamps can't go up to VCC+ but can go down to Vcc-

CMOS Op amps are Rail to rail and can go up to vcc+ and down to vcc-



as much as it may sound like it, this is not a homework question

 

[audioguru]

So from what I have seen, is this right

JFET Op-amps can't go down to VCC- but can go up to VCC+ on the inputs

No.
Texas Instruments made TL09x opamps many years ago with N-channel input Jfets. Their input common-mode voltage included the negative power supply pin voltage. I got some.
Most opamps with Jfet inputs use p-channel Jfets.

BJT Opamps can't go up to VCC+ but can go down to Vcc-

No.
The LM358, LM324, MC3317x, MC3407x and TLE214x opamps have BJT inputs that work at the negative supply pin voltage.

CMOS Op amps are Rail to rail and can go up to vcc+ and down to vcc-

Maybe. Most do.

 

[axro]

No.
The LM358, LM324, MC3317x, MC3407x and TLE214x opamps have BJT inputs that work at the negative supply pin voltage.

Isn't that what I said? The inputs can go down to VCC- and not up to VCC+?


Also what happens when you are beyond the common-made input range? I know on the TL081 when you go too low there is the phase invesion. But what happens on a LM358 when you go above the VCC+ - 1.5V. Does it just not notice any change in voltage or does something actually happen.