I suspect I am making a probe connection mistake, but couldn't figure it out.

The circuit works fine in LTSpice but I don't get anything in a real experiment.

What might be the problem?

enter image description here

  • \$\begingroup\$ What parts you used? How did you built it? Have you checked for errors? What errors did you check? \$\endgroup\$
    – Justme
    Jun 13, 2022 at 20:23
  • 1
    \$\begingroup\$ Your R2 looks connected to ground but the bottom wire from C2 does not appear connected to that ground. Any place there's 3 or more wires intersecting, you should see a dot. I don't see that dot. \$\endgroup\$
    – horta
    Jun 13, 2022 at 20:29
  • 2
    \$\begingroup\$ Akagami, you have enough gain but you don't have a way of constraining the gain (diodes or an incandescent lamp?) to just what's needed for oscillation. You might read this page which shows up right away on google for me. Go to the bottom of the page and read the bullet points. I gather you are looking to make a 1 kHz oscillator? (Also, if you are having trouble with an LM741 in simulation you might first try using a good quality rail to rail opamp like the LT1800, get that working, then go back to the LM741.) \$\endgroup\$
    – jonk
    Jun 13, 2022 at 20:52
  • 1
    \$\begingroup\$ What precision resistors are you using? Since you are using a 741, it suggests that you are working on the cheap. If those are 10% resistors, then it's entirely possible that the op amp gain is too low. Start by swapping R1 and R2. If that doesn't work, try increasing R4 to 25k. \$\endgroup\$ Jun 13, 2022 at 21:05
  • \$\begingroup\$ Akagami, you have not responded to anyone, yet. There are some questions/critiques that have been mentioned. Could you respond? Also, do you need or want +/- 15 V rails? Or a single supply example? Do you plan to build one? Or just simulate it? (Considerations are a little different, depending.) \$\endgroup\$
    – jonk
    Jun 13, 2022 at 23:45

3 Answers 3


As @horta suggested, use wires when needed and "see" the "dot" connection.

You should see something like this (distorted.)

If oscillation does not start (R3 < = 21k,) make this resistor a little greater as also @WhatRoughBeast suggested in his comment.

enter image description here

If you want something more "sinusoidal", you should use diodes as a gain "limiter".

Here is an example.

enter image description here


I may as well add a practical circuit using cheap parts.

The LM324 provides roughly four LM741's in a single package. (Nice, if you consider having an LM741 nice.) At least the "knock-off" ICs are cheap (around 5 cents each, if you look around a bit and are willing to buy 100 of them -- and you are willing to risk the fact that their performance won't be anything like a real LM741.) And they can be used with a single rail, too, if you construct a "virtual ground" for them. (Shown in the schematic below.) Since there are a bunch (four) opamps in an IC package, this is also cheap and easy to do. This can save you the difficulties related to creating a bipolar supply. (Note: Although I still do have some LM324's from many decades ago, I haven't used one in about the same period of time -- decades.)


simulate this circuit – Schematic created using CircuitLab

That uses half an LM324 IC. The potentiometer is needed so that you can dial in oscillation. If \$R_7\$ is too small, you will get oscillation but with distortion. If \$R_7\$ is too large, you won't get any oscillation and the output will just settle on some DC value. Having the potentiometer in the circuit means you can build these over and over and make adjustments for the vagaries of parts.

In the above, I am thinking about \$V_{_\text{CC}}\ge 9\:\text{V}\$, or so. That supply rail simulates as:

enter image description here

Here's the FFT:

freq: 1/(t2-t1)=1050.42

Harmonic    Frequency    Fourier    Normalized   Phase      Normalized
 Number       [Hz]      Component    Component  [degree]    Phase [deg]
    1       1.050e+03   3.090e+00   1.000e+00    -107.87°       0.00°
    2       2.101e+03   7.771e-04   2.515e-04     176.48°     284.36°
    3       3.151e+03   3.885e-02   1.257e-02     -12.03°      95.84°
    4       4.202e+03   2.015e-04   6.522e-05    -119.87°     -12.00°
    5       5.252e+03   4.612e-03   1.492e-03     -24.53°      83.34°
    6       6.303e+03   1.153e-04   3.732e-05    -167.34°     -59.47°
    7       7.353e+03   1.004e-04   3.248e-05      57.14°     165.02°
    8       8.403e+03   1.107e-04   3.583e-05      70.47°     178.34°
    9       9.454e+03   2.440e-04   7.896e-05     114.42°     222.29°
Total Harmonic Distortion: 1.266211%(1.266473%)

If you have a bipolar supply, obviously there is no longer a need for the left-hand section that develops \$V_{_\text{GND}}\$.

If you don't want to include zeners, you can use a multi-feedback amplifier design to get to a similar place -- again with two opamp sections. But I suppose that isn't of interest, so I'll hold short on that topic.

  • \$\begingroup\$ The LM324 and LM358 opamps are noisy and produce crossover distortion. A DC load resistor can be added to convert the class-B output to class-A with heating and no crossover distortion. \$\endgroup\$
    – Audioguru
    Jun 14, 2022 at 16:19
  • \$\begingroup\$ @Audioguru Thanks. The scope of the above was to demonstrate (1) the use of zeners (as often found being applied) and (2) how a single supply rail might be applied and (3) why a potentiometer may be of interest. I will add a simulation result, though. \$\endgroup\$
    – jonk
    Jun 14, 2022 at 16:51
  • 1
    \$\begingroup\$ +1 The rather common 4558 is a dual LM741 (without the nasty crossover distortion of the LM358/LM324). Quoting TI: "The RC4558 device is a dual general-purpose operational amplifier, with each half electrically similar to the μA741, except that offset null capability is not provided". IIRC, the JRC version was well thought of in some circles for some (audio-related) reason. \$\endgroup\$ Jun 14, 2022 at 18:56
  • \$\begingroup\$ @SpehroPefhany Thanks. I don't recall using a 4558. But I'll read up on it and compare its datasheet with the two I mentioned. Appreciated! \$\endgroup\$
    – jonk
    Jun 14, 2022 at 19:00
  • \$\begingroup\$ The RC45568 was first used 48 years ago in 1974 before audio opamps were invented. The NE5532 was invented 5 years later. \$\endgroup\$
    – Audioguru
    Jun 15, 2022 at 15:20

This is a practical circuit. Previously built and tested.

The J-FET provides the automatic gain control.

R8 & R3 linearise the J-FET

Wien Oscillator.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

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