LM7815 and LM7915 problem in Multisim.

[Ninerz]

LM7815 and LM7915 not regulating at maximum +15 and -15 volts respectively when supplying to audio amplifier.

Here is some pictures attached.

When my audio amplifier amplifies the input signal to let's say near 15 Vrms, nothing bad happens whatsoever.

But when my audio amplifier amplifies the input signal above 15 Vrms, it should suppose to clip the signal to just 15 Vrms because i supply the audio amp with just +15V and -15Vdc.

However the output reaches beyond 15 Vrms mark and with no distortions.. but when i take a look in the oscilloscope in the LM7815 and LM7915 output voltages, as you can see in the pictures attached, it got weirder.

What's the problem because i thought it should clipped at its maximum 15 Vrms... 
In the real world, will this happen?

Instead of regulating it at maximum +15V and -15 Volts by the LM7815 and LM7915 ICs, it got around like 29+ Vdc with weird ripples...

I will appreciate any suggestions and thoughts regarding this problem. 

[Buechsenschuetz · Accepted solution]

Of course, in the real world,this would not happen. The LM358 is a standard op amp and can't output beyond the rail. Actually, clipping probably would occur at +/- 13...15V since this op amp cannot fully swing to the supply rails. The voltage regulators never go beyond their nominal voltage except when they are broken. 

I think there is some misbehaviour in the simulation. It looks as if the voltage regulators no longer regulate anything. The ripple you see has periods of approx.  10ms. What you see here is the ripple of the rectifier output (100Hz). But I have no idea why the simulation changes the supply voltage and behaves as if the regulators were bypassed. Also, in the real world, the LM358 would not withstand +/-29VDC on the supply pins, I think it's rated to +/-18VDC maximum.

As a sidenote, I would suggest bypassing capacitors on the supply terminals of the op amp. Usually 10uF electrolytic + 100nF ceramic are a good choice, they should be placed as close as possible to the supply pins (with the smaller capacitor closest) and should be connected directly to the GND rail of the op-amp circuit. 

[Ninerz]

Ohhhhh so in reality, it should have clipped the output waveform at around 13-15V. Thanks for confirming it. And regarding about the IC regulators in multisim, I was confused because it acted weird and unusual comparing it to how I thought IC regulators would act normally in actual. Nonetheless, I will do some calculations and some readjusting the values of some resistors to lessen the gain so that and max volume whatsoever, it would be enough or close to clipping/distortions. Kudos to you mate!

[johnsold · Accepted solution]

Many op amp models do not accurately simulate non-linear behavior beyond the power supply voltage limits or datasheet current limits. Either examine the model code carefully or run a series of simple simulations to see how the model behaves under the specific circumstances you need to evaluate.

That caution also applies to three terminal regulator models.

Lynn