1
\$\begingroup\$

I've been struggling on build a noninverting amplifier with TL972IP (datasheet).

Here's what I've made:

enter image description here

Orange is for VCC+, Yellow is for VCC-, Green is for GND. Blue is for a signal input. Resistors are for feedback.

As it is hard to see, here's a diagram:

schematic

simulate this circuit – Schematic created using CircuitLab

With the configuration above, I've got this:

enter image description here

Yellow is the input and Blue is the output. As you see, input is attenuated (expected: gain of 2) and nagatively biased.

Where can I look into and see what's going wrong with this circuit?

\$\endgroup\$
  • \$\begingroup\$ Are you sure everything's hooked up correctly? Pin 8 on the DIP goes to one of the breadboard rails, but there's no indication in the photo if that's connected to anything. \$\endgroup\$ – Nick Johnson Jul 10 '15 at 12:34
  • \$\begingroup\$ Ah, yes. That's because I use the breadboard too wide and sparse. If I take a full view, a part and connections around op-amp won't be seen. \$\endgroup\$ – Jeon Jul 10 '15 at 12:35
  • \$\begingroup\$ And, I checked every wire is connected properly as far as I can. But, there maybe some which I've missed. \$\endgroup\$ – Jeon Jul 10 '15 at 12:38
  • \$\begingroup\$ In your opinion, @NickJohnson, isn't there no errors and problems on diagram and circuit? (If I connected them correctly.) Thus, you suspect there's human errors (e.g. misconnection) or defects on resistors, op-amp and other equipments? \$\endgroup\$ – Jeon Jul 10 '15 at 12:39
  • 1
    \$\begingroup\$ Measure the actual DC voltages (with a voltmeter and input shorted to ground) on pins 1, 2, 3, 4 and 8 wrt ground (and tell us what they are). \$\endgroup\$ – Spehro Pefhany Jul 10 '15 at 13:26
6
\$\begingroup\$

It might be the probe attenuation. Some probes have a switch on them that let you select an attenuation of 2x, 5x, 10x. Other probes are simply built to attenuate and don't have a select switch, but simply have a, say, 10x written on them.

To account for this, oscilloscopes have an option for each channel to specify what kind of probe is connected to it. For example, if you're using a 10x probe you would select the 10x option on that channel, so that the oscilloscope can display the right signal.

My guess is that you either are using a 10x probe without the proper option set on the oscilloscope, or vice versa you're using a regular probe with the 10x option activated on the oscilloscope.

If you multiply the output channel vertical scale by 10, you get 1V, and with that scale the output would be exactly twice the input (since the input vertical scale is 500mV).

\$\endgroup\$
  • \$\begingroup\$ Well spotted ... a x10 setting on the probe on Ch2 would explain a lot. \$\endgroup\$ – brhans Jul 10 '15 at 13:52
  • \$\begingroup\$ Yeah, it was my fault. probe is set to attenuate by x10. Thanks. \$\endgroup\$ – Jeon Jul 11 '15 at 11:46
0
\$\begingroup\$

For the attenuated output, check if you are using any attenuation on the probe as Leo said.

The bias in the output is due to the offset current (the input is made with BJTs). To compensate for that, decouple your voltage source with a capacitor and add a resistance of 235k (470k//470k) to ground before de positive input of your opamp. I guess that will do. I that is not enough, you could try decoupling R1. To learn the explanation for this, google "opamp offset current".

You can see here what I mean. http://www.filedropper.com/bias The second option is what I sugest, although the third one will probably

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.