QUCS is easily installed by following the steps on their home page. For ubuntu it would be something like this
sudo apt-add-repository ppa:qucs/qucs
sudo apt-get update
sudo apt-get install qucs
and for windows or mac users, at the similar level of easinicityness.
The rest of the story is a drag-and-drop exercise and a bit of playing around to find all the components needed to start the simulation. However, for the TSEI01 course I prepared an example file as illustrated in the figure below.
It simulates an inverting amplifier using an operational amplifier, and a couple of passives to create a first-order pole, a low-pass filter.
Three different simulations are added: DC (operating point), AC (frequency response), and transient (behavior in the time domain). There is a slight problematic situation for the simulation when applying a 0-V DC input and hence I added a small offset through voltage source V3. The ratio between the two resistors connected to the operational amplifier is 100/50 = 2, and we can see both in the frequency plot and the time-domain plot that indeed 2 is the gain. (In the lower-right transient response we do see the effect of the offset too, hence the slight shift up and down of the two curves). Further on, we see in the transient response that the inversion is there: the phase of the sinusoid is shifted 180 degrees.
From the frequency plot we can see that the bandwidth is in the order of 200 kHz. That does not seem to be all correct compared to the expected pole, right? The expected cut-off frequency given this configuration should be something in the order of 1 / RC ~ GHz range. So what is happening? Well, in this case we use a BJT model of the operational amplifier, a model of the uA741 OP. This means that the bandwidth of the opamp itself is limiting since the opamp is far from being ideal. In fact, the bandwidth is approximately 1 MHz, and due to the feedback we see beta MHz (feedback factor) instead as a system bandwidth. Replace the capacitance with a 1-uF instead and you will see that the pole now shifts downwards to expected values.
Qucs is quite simple to go from idea to simulated results. But requires a bit of tweaking to get going if the simulator gets into convergence issues.
- The QUCS project can be downloaded here.