Op Amps

[UltraRC]

I have just received some op amps in the mail and have watched the eev blog op amp tutorial, I fully understand the non-inverting amplifier configuration (I think), but it simply does not do what it is supposed to do for me. The op amps that I bought are the LM324's. My circuit consists of a 4.5v battery pack connected to a breadboard with 1kohm resistors setup as a voltage divider to make a ground (+2.25v, 0v, -2.25v). I connect the negative voltage to pin 11 and the positive voltage to pin 4. R1 is 1kohm and RF is 1kohm. So shouldn't this mean that if the non-inverting input is fed 0.5v, then the output should be 1v? AV = 1k / 1k + 1. So AV is 2? The problem is when I feed 0.5v into the non-inverting pin, I get 0.65v on the output. Is there something that I am missing? I am using the center tap of the 1k voltage divider as the ground reference. I know there is a limitation to the output, would my output not even be able to reach 1v and it is just being clipped off? I suspect that clipping is the problem but it would be nice to have some confirmation.
Thanks.

[IanB]

I think you are confusing yourself by trying to make a "ground" with a voltage divider. Don't do that. Just connect the negative of your battery to the negative power rail of the op amp and the battery positive to the positive power rail. Then measure all voltages relative to V-.

Also, to help your understanding draw out the circuit schematic and calculate what the voltages should be at each point. Then measure and see if they match.

[UltraRC]

But wouldn't that mean that i wouldn't be able to amplify an AC signal without the bottom being chopped off?

[edy]

I found this book in Google, may be worth checking out. It is over 400 pages though, from MIT, called "Op Amps for Everyone":

http://web.mit.edu/6.101/www/reference/op_amps_everyone.pdf

[rstofer]

But wouldn't that mean that i wouldn't be able to amplify an AC signal without the bottom being chopped off?

Yes it would.  There's no problem with using the 9V battery between +V and -V but you still need to create a virtual ground reference

See page 18 here:
https://people.rit.edu/lffeee/OpAmpSingleSupply.pdf

[Connoiseur]

If you must use a single voltage source for some reason, then add a buffer (opamp in unity gain configuration) to the existing voltage divider. Alternatively you can use a rail splitter IC like TLE2426 to get a ground node.

The reason your output is getting clipped is because current can flow from the output pin to the voltage divider via the feedback resistor and disrupt the potential at the node; since the voltage divider has impedance comparable to the feedback resistor. Had the divider resistance been much lower than the feedback, you wouldn't have noticed any discrepancies.

[IanB]

But wouldn't that mean that i wouldn't be able to amplify an AC signal without the bottom being chopped off?

Maybe. But why try to run before you can walk? Since you are learning from the beginning, keep it simple and work up to the more complex stuff one step at a time.

Since you have not shown a circuit diagram, it is not clear how or where you have tried to include the split voltage supply. I for one cannot be certain what circuit you have built and where you have measured your voltages.

[Brutte]

1kohm resistors setup as a voltage divider to make a ground (+2.25v, 0v, -2.25v)

Ok, so this is called "0V voltage source of 500 ohms output resistance".

I connect the negative voltage to pin 11 and the positive voltage to pin 4.

I always tie GND directly and Vcc via a small resistor. LM324 should not pull more than several mA so tying it your way shortens its operating life
Use 22R resistor for 0.2A short circuit current to tie Vcc instead.

R1 is 1kohm and RF is 1kohm. So shouldn't this mean that if the non-inverting input is fed 0.5v, then the output should be 1v? AV = 1k / 1k + 1. So AV is 2?

No, it is not. You have tied R1=1k to a 0V reference with 500R. So that means a reference voltage is tied to inverting node via 1k5 resistance.

The problem is when I feed 0.5v into the non-inverting pin, I get 0.65v on the output. Is there something that I am missing?

Voltage source always has its output resistance/impedance. Recent discussion about that. It may be 1R, may be 1M but it is always there. If you add 1k to 1R then that is a minor issue but not in your case. Just use R1=500R to get what you want. Or use 2k divider and tie that to input directly, as in summing amp configuration.

I am using the center tap of the 1k voltage divider as the ground reference. I know there is a limitation to the output, would my output not even be able to reach 1v and it is just being clipped off?

The output ranges from Vcc-1.5V (+0.75V in here) down to almost GND (-2.25V in here) with low load. You won't reach 1V on the output.

[danadak]

Something like this, attached ?


Regards, Dana.

[rstofer]

If AC input is required, you absolutely must create a reference voltage of Vcc / 2.  You want the output to be 0V when the input is 0V relative to something.  The questions become 'where to inject it?' and 'relative to what?'.  Vref can be summed at the input.

This document is Chapter 4 of "Op Amps For Everyone" and along about page 10, things get interesting:
http://www.ti.com/lit/an/sloa030a/sloa030a.pdf

Basically, examples are given that solve the equation y = mx+b (a straight line with slope (gain) m and offset b.  They are all single supply circuits.  Without doubt, this is my favorite chapter of the entire book and probably of all books combined.

It is not unusual to create the Vcc/2 with a pair of resistors and a voltage follower.  This keeps the Vref stable and creates a low source impedance without having to use very low value resistors.

[Audioguru]

No schematic was posted so we are all just guessing about how it is connected. Here are a few ways it might be connected:

[singapol]

I have just received some op amps in the mail and have watched the eev blog op amp tutorial, I fully understand the non-inverting amplifier configuration (I think), but it simply does not do what it is supposed to do for me. The op amps that I bought are the LM324's. My circuit consists of a 4.5v battery pack connected to a breadboard with 1kohm resistors setup as a voltage divider to make a ground (+2.25v, 0v, -2.25v). I connect the negative voltage to pin 11 and the positive voltage to pin 4. R1 is 1kohm and RF is 1kohm. So shouldn't this mean that if the non-inverting input is fed 0.5v, then the output should be 1v? AV = 1k / 1k + 1. So AV is 2? The problem is when I feed 0.5v into the non-inverting pin, I get 0.65v on the output. Is there something that I am missing? I am using the center tap of the 1k voltage divider as the ground reference. I know there is a limitation to the output, would my output not even be able to reach 1v and it is just being clipped off? I suspect that clipping is the problem but it would be nice to have some confirmation.
Thanks.

It's how you key in the numbers when using your scientific calculator. 1000 divide 1000+1 =2.
So in fact it should be 1000/1001 =0.99

Correct way is 1000 divide (use bracket) like this (1000 + 1) then result is 0.99 ~ 1 roughly.

So you feed 0.5V and get 0.65V because of opamp dc offset.   

[danadak]

If you have a NI amp, and it has equal Rf and Ri on the inverting
side, then G = 2 (not counting error due to finite Aol).

https://www.google.com/search?q=non+inverting+gain+opamp+image&tbm=isch&imgil=USRc9-6eewqZRM%253A%253B7WxNUtvUOZhDcM%253Bhttp%25253A%25252F%25252Frsrelectronics.com%25252Ft_gain.htm&source=iu&pf=m&fir=USRc9-6eewqZRM%253A%252C7WxNUtvUOZhDcM%252C_&usg=__pJUERIYU-Lh7S83EN0ETftfIHUw%3D


Regards, Dana.

[Brutte]

The op amps that I bought are the LM324's. My circuit consists of a 4.5v battery pack connected to a breadboard

This LM324 is a really cool IC, it is quad, cheap as dirt, includes GND on input and output and goes from 3V up to 40V. It is also quite fast with GBW of 1.5MHz

However it has some limitations for someone who does breadboarding and does not fully understand internals.
LM324 is not rail2rail as it won't reach Vcc on input or output. Also do not bother using it with 1.8V supply.
I'd suggest you should get some rail2rail counterpart of LM324. Like for example Microchip MCP6004. It is still popular, a bit more expensive and not that fast but it would save you some problems. Just do not wire it to Vcc via 0R001 wire - it won't tolerate abuse on its power supply and inputs.

[Audioguru]

This LM324 is a really cool IC, it is quad, cheap as dirt, includes GND on input and output and goes from 3V up to 40V. It is also quite fast with GBW of 1.5MHz

National Semi who invented it says its GBP is 1MHz. The graph of frequency response in its datasheet shows 1MHz with a 30V supply and about 350kHz with a 10V to 15V supply. With a 5V supply they do not say but I guess only 100kHz which is extremely slow.
The graph of its slew rate shows trouble above 5kHz when its suooly is 15V and its output level is high.
It is very noisy. It has awful crossover distortion.