I just noticed this on http://www.talkingelectronics.com/p...mplifier/TheTransistorAmplifier-P1.html#Stage

At figure 38. I usually just use .tran 10m or so. The author said this circuit shouldn't work, and I simmed it, and it did work. I tried to figure out why he might say it wouldn't and hit the button to start at 0v external DC, and it failed. I started another known working amp circuit at 0v DC and it turned on after 5ms and started working appropriately.

Should I be always using start at 0V DC? Will some circuits falsely appear to work if it isn't selected?

 

Start at 0V DC is just if you want to see how the circuit starts when power is applied. For example the output on some voltage regulator circuits will overshoot the set value by quite a bit on startup and you want to be able to see that.
If you are not concerned about how the circuit reacts at power on then you don't need to use that option.

 

Okay I think I understand it better. If its only set for .tran 10m, the circuit will oscillate with the signal for up to 20ms but will eventually kill off. With "start at DC 0V", it never gets started in the first place.

Expanding view to 100ms allowed me to see the circuit failing after a time, which "start at DC 0V" showed its failure to operate immediately. Thanks for the help.

I believe the setup is meant to be an amplifier circuit for incoming AC signals, it falls under proper amplifier or improper amplifier circuits.

 

If simulating oscillators, sometimes the 'start with supplies set to 0V' will kick the oscillator in the guts and get it going, while if you do the normal DC solution, and then wait for the oscillator to start, you may die of old age before it does...

This is a non-physical situation, because there is always noise present in the physical oscillator that may not be present in the Spice model...

The circuit of Fig 38 has positive feedback and has two stable states (like a flip flop). The correct way to guide the simulation to either of the stable states is to use the .IC V(node) = X directive. Using the start with supplies set to 0V is the wrong way to do it...

 

That explains why I wasn't getting an oscillator from Gottlieb's 1971 book going, that was so frustrating, lol.

Or possibly the fact that I thought a UJT was a FET. A nice BJT phase shift oscillator is working for me in spice.