Losses in NE555 oscillator and Sallen-Key filter

I use a Sallen-Key bandpass filter in a project at university to isolate the 1kHz sinusoidal part of a square signal produced by a NE555 oscillator. Here is my circuit:

When I simulate it on PSpice, I get this:

But when I check the voltage with an oscilloscope on the real circuit, this is what I get (5V per square):

How can I explain the big differences in amplitude of the signal between simulation and reality ? Is that only losses in the NE555 and non-ideality of the opamp in the filter ?

• Can you post the two schematics? The one that you simulated, and the one that you built. Commented May 17, 2012 at 15:05
• In fact, both are wrong. If your filter has G=1 @ 1 kHz, and your square wave is 5 Vpp, then you should see an output sinusoid of 5*4/pi=6.37 Vpp in both cases. Commented May 17, 2012 at 15:22
• Are you using the same circuit as in your previous question? Commented May 17, 2012 at 16:09
• With G=2, you should read/measure 12.7 Vpp. // If you compute the Fourier series decomposition of a 50% duty ratio, square wave that goes from -1 to +1 V (2 Vpp), the component at the fundamental frequency has amplitude 4/pi Vp, so 8/pi Vpp. (8/pi)/2=4/pi, so that's the factor. See section "Examining the square wave" in en.wikipedia.org/wiki/Square_wave Also, the first figure here en.wikipedia.org/wiki/Fourier_series Notice that the sinusoid amplitude is larger than the square wave amplitude. Commented May 17, 2012 at 19:58
• Can you alter the frequency from the 555 (up and down) and see if you can find the frequency where the amplitude is maximum? This problem may be due to component tolerances and a high Q circuit. Commented May 18, 2012 at 8:12