Question about a relaxation oscillator using two NPN transistors

[eletmoraes]

Oscillator done in Orcad 16.6 as in figure 1


Waveform obtained in Orcad as the picture 1


And the same circuit using Proteus (image 3 )






The doubt is because they do not get the same waveform as the Proteus?

 

[esp1]

hi,
The cross coupling capacitors are shown as 'polarised', the right side cap is reversed.?
Perhaps Orcad calculates based on a reversed biased cap, whereas Proteus does not.?

E

 

[FvM]

Perhaps Orcad calculates based on a reversed biased cap, whereas Proteus does not.?

I don't believe that polarized capacitors are modelled specifically at all.

Trivial explanation, 1 ms simulation time in Orcad is just too short to see oscillations.

In addition, you might stumble upon a standard oscillator simulation problem. By working of the initial transient solution the oscillator can power up in a labile (or stabile, depending on the DC loop gain sign) equilibrium with no oscillation. The circuit needs either a kick-start pulse or an artificial inbalance by an .IC statement for a node voltage.

 

[Audioguru]

The base-emitter voltage of each transistor becomes reverse-biased when the capacitor drives the base to a negative voltage. The reverse voltage is too high for the transistors (the maximum allowed reverse voltage is 6V) and causes the base-emitter to have avalanche breakdown. The supply voltage should be reduced to 7.5V or less to avoid it or diodes can be added to prevent it.

The Orcad scope trace shows only part of the slanted voltage of each pulse because the scope is tracing much too fast.

Eric,
Good for you to spot the reversed cap polarity.

 

[eletmoraes]

Analyzing the best circuit I had ridden realized that some components had loaded the wrong way. When mounting again as the image 4 got waveform in the same way it was assembled in Multisim 13.

 

[Audioguru]

If you build the circuit it will look like both LEDs are continuously lit at the same time because it is switching so fast.