colin55, here is an example of what we called in the R&D across the world High/Low side drivers.
https://www.electro-tech-online.com/custompdfs/2009/02/p395.pdf
True, Derived from the law of conservation of energy that upon switching OFF, the energy of the coil will distribute to the diode, the internal coil resistance of the coil and the energy lost due to the magnetic properties of the the relay coil (includes energy lost through magnetic field).
About the sudden drop, if your design is drawing current about 50mA when the switch is off, and then, eventually the switch is turned which has a draw of 200mA. Therefore, there is a change of current from 50mA to 250mA at certain instance as seen in your power supply side. At your switch side for the relay (if your using a semicon switch), it has a leakage current of 1uA (probably, depends on specs) when it is off, and draws 200mA during on.
Let say the transition happened for 1us. So, we can use the V=L*di/dt.
di = change of current = 200mA - 1uA
dt = change of time = 1us - 0us
L = inductance of the coil
so, you can know the voltage at 1us. But if you're interested what is the voltage at <1us, we need to know what is the current at <1us. So we have V as a function of t, V(t). If you plot this into a graph, you will get the actual spike.
The last paragraph is presented in a technical way. It may be clear to others who has the experience in dealing with transient suppression or EMC stuffs. This is more on the automotive electronics engineering.
One thing, if you already know about high side and low side switch, the negative and positive transients is produced due to the flow of current and the position of the load with respect to the supply.