Something that's always helped me in doing SFX using PSGs/trackers is to look at the waveform of a real example of the sound. If you want to make a good footstep sound, open up a real recording of a footstep and look at what makes up that waveform.
As you analyze this you're looking for a few things:
How does the amplitude (volume) change over time?
How does the pitch change over time?
Can you identify a general shape of the waveform?
Are there a lot of resonant frequencies?
Amplitude and pitch are pretty easy to edit and automate in almost any given tracker so you can simply recreate those elements of the sound in your tracker. Unfortunately, most real world sounds are not simple waveforms so you'll have to get creative about how to emulate overtones and resonant frequencies.
Fast arpeggios and pulse width modulation can sometimes be effective to get closer to the original sound and for sounds that appear extremely resonant or seemingly random you can use a noise channel to get some of the sharp sound.
Eventually through this sort of exploration you'll get a pretty good sense of what a sound is made up of without even having to open up a graphical representation of it and soon enough you'll have a whole slew of tricks for mimicking all sorts of different sounds.