Something Weird

216

01:48:28

Okay i designed a simple op amp, like the ones inside IC's and i measured exactly 0% THD+N. I know that in reality that's impossible, but it should also be impossible in simulations too. So far, whatever i designed it had a certain amount of noise. The lowest I've always gotten was 0.01%. Now i measured the noise level, it was 360uV, then i decided to take out the margin of error, by measuring the THD of an ideal source and again found that it was exactly as the previous (360uV). Not a uV more or less. That means exactly 0% THD+N. I can partially get if I've managed to isolate all harmonic distortion, but there are other types of noise, like semiconductor noise, nonlinear distortion, intermodulation distortion and others. As far as I'm concerned that's impossible even in EC. I furthermore measured a ripple rejection of -105dB which also seems improbable. Did Igor made some unannounced changes in the last version, that negates most noise of the devices in EC? 

published 10 years ago

hurz

10 years ago

EC does not have white pink or any form of natural noise. Only numerical noise which is about 120dB below voltages we use for amplifier. What we can measure are nonlinearities and crossover distortion which are suppressed almost perfectly in case you build something with negative feedback and high enough open loop gain. So not a big surprise for a good working opamp to have in simulation a zero THD.

thebugger

10 years ago

Yeah but even EC has some noise. I mean, as we've already stated, there's numerical noise, and probably a sampling error, and semiconductor noise, intermodulation noise, harmonic distortion. I don't think i could've eliminated all these sources. I've never gotten such perfect results without the ideal Op Amp of EC.

flowDAQ

10 years ago

The differential input looks interesting, don't think I have seen it like that.

hurz

10 years ago

Again numerical noise is much below your measurement method. Thermal noise (semiconductor noise you mean?) is not modeled within EC. Harmonics/intermodulation are present cuz they come from nonlinear behaviour. If this nonlinearity (which all semiconductors have) is negated by feedback then there is a close to zero THD. Not a big surprise.

thebugger

10 years ago

But the negative feedback is not complete. I mean it's not 100%. Only around 33% for AC

hurz

10 years ago

Right this only suppresses nonlinearities by 10dB. Crossover is not an issue cuz its a class A amp. I will try later on to measure the THD.

thebugger

10 years ago

Okay, try. I took my reading at 1kHz with 20x oversampling and the notch filter was -90dB i think. Then take out the margin of error, by measuring an ideal source with the same voltage, and let's see what you get

add comment in editor

EveryCircuit is an easy to use, highly interactive circuit simulator and schematic capture tool. Real-time circuit simulation, interactivity, and dynamic visualization make it a must have application for professionals and academia. EveryCircuit user community has collaboratively created the largest searchable library of circuit designs. EveryCircuit app runs online in popular browsers and on mobile phones and tablets, enabling you to capture design ideas and learn electronics on the go.