100209-1310 EST
I am not clear on your question.
You could put a small control transformer in a box so its terminals are not directly accessible. Maybe 500 to 1000 VA, small enough that it has moderate internal impedance relative to the loads you might use. Thus, I favor 500 VA. These are usually available as two primaries and two secondaries. Wire it so that the two secondaries form a center tapped 120-0-120 output. Most likely the inputs would be in parallel from 240. These transformers are usually wound about 2 to 1.
250 W lamps can provide substantial load and therefore secondary voltage change. The center tap can connect to your main EGC for safety. You can setup a system with loose connections, or specific resistance values for simulation of different problems. GFCIs should be part of the test, maybe some old ones without new features. You can simulate bad neutral connections, miswired circuits, bad hot connections, leakage resistance problems with the GFCI, voltage drop problems, phantom voltage on a three-way, have a buck transformer to create a low voltage on one phase of the supply transformer, get some old bad backstab receptacles to create their problems, have a a slightly bad transient voltage limiter that will trip the GFCI, have a KIll-A-Watt EZ meter for experiments, a 1/4 HP capacitor run induction motor (no load), a compensating capacitor for the motor, and may more possibilities. Pick what is appropriate for your student base. Do not try to fool them. Create real world problems to be found or observed.
Setup two 15 W bulbs to use to evaluate neutral problems. I can detect a 1.5 V change at 120 V with one of these bulbs.
.