As part of my feedback from doing the teaching given by my peer, one of the comments was to teach the capabilities of what we are doing in the Physical Computing practical lessons, before actually doing the practice part of the lesson.

I used to do this, and found that it sets expectations higher than what they are able to achieve in the session, also, the sessions are very long (4 hours) and if a lengthy lecture is given before setting up the practical session, we loose energy from the students.

I much prefer to get them building something, perhaps even give some example code (though I much prefer live coding together in class) and have them be hands on with the building blocks, before then expanding, after the practical session, sometimes with an element of surprise. This way, I can see their minds open up and get creative about what they might do next, rather than having them compare their work to others that have gone before.