high offset voltage in 741 voltage follower.

i was trying to do a simple voltage follower but I am seeing a very high offset voltage, about 1.8-2v. I am using a 12v-gnd supply. I used a 10k feedback resistor and grounded the non-inverting end.

Even when i apply a signal to +ve terminal that varies from say 0-5v, I get steady 2v output till my input also reaches 2v. after that the output follows the inputs.

how do i fix these issues.

The 741 can not operate with a single supply. It can go below about 2V above the negative supply rail (as shown in its data sheet), as you've observed. So you either need to add a negative supply or use a single-supply opamp such as an LM324. If you also want the output to go to 12V then you need a rail-rail type opamp.

What is the highest frequency of your signal?

No single supply buffer can reproduce voltages down to zero, but many OPs can make it down to a few 10 millivolts. 741 isn't designed for single supply operation at all, review the data sheet for input common mode range to understand the issue.

doors666 said:

I used a 10k feedback resistor and grounded the non-inverting end.

Click to expand...

In addition to the supply problems - you didn´t mention the second resistor between signal input and the inverting terminal. Without such a resistor there is no feedback at all.

http://www.ti.com/lit/an/sboa092a/sboa092a.pdf
http://cie-wc.edu/1436VoltageFollower02-07-2012.pdf
http://www.talkingelectronics.com/projects/OP-AMP/OP-AMP-1.html

thanx for the help guys, i tried with a -ve rail and it works fine, I get an offset of 44mv, which is ok.
I will try to answer all the questions here.

What is the highest frequency of your signal?

Right now I am just trying to test the concept, but eventually it will go to about 1mhz, square wave upto 1.25v.

No single supply buffer can reproduce voltages down to zero, but many OPs can make it down to a few 10 millivolts.

Could you please suggest which ones go down to tens of mv and can also handle 1mhz signal.

In addition to the supply problems - you didn´t mention the second resistor between signal input and the inverting terminal. Without such a resistor there is no feedback at all.

I am using a 1K resistor for that.

I also have a simple lm317 based supply that had a minimum starting point of 1.2V, which is as per the datasheet. I tried the circuit mentioned at the link below to get down to 0v. It doesnt seem to work for me. I used a tip31c as the transistor. I am using a 12v dc supply for input instead of transformer and a BR. I get some 11 point something volts at the output. No impact of changing the pot value. Is the circuit good in principle?
https://www.bristolwatch.com/ele/lm317.htm

doors666 said:

I also have a simple lm317 based supply that had a minimum starting point of 1.2V, which is as per the datasheet. I tried the circuit mentioned at the link below to get down to 0v. It doesnt seem to work for me. I used a tip31c as the transistor. I am using a 12v dc supply for input instead of transformer and a BR. I get some 11 point something volts at the output. No impact of changing the pot value. Is the circuit good in principle?
https://www.bristolwatch.com/ele/lm317.htm

Click to expand...

Your referenced circuit uses an emitter-follower current boost which degrades the output voltage regulation since the feedback voltage is taken from the base of the transistor not the emitter. To boost the current and not degrade the regulation you could use a circuit such as this.

I don't see any mention in your link about how to get the output down to 0V. For that you need a negative supply in addition to the positive supply. If you want a regulator that can go down to 0V with a single supply you could use a LT3083, a 3A Linear Regulator.

The LM317 requires a minimum worst-case voltage across it under load of about 3V so the maximum output with a 12V supply is about 9V.

I want to try the first circuit in the link you posted above. I have 0.5 ohm 5w and 1 ohm 2w for the current sense resistor. lm317 is on a decent sized heatsink. Is 1k a good value for the base resistor r2. I currently dont have a pnp bjt, I should be able to use a p channel mosfet instead without any modifications, right?

If I use a dual supply, will varying the pot give me -ve voltages also at the output?