Just a thought, I discovered recently that the TIP125 very quickly gets alarmingly out of control where current is concerned with only 1mA base current..., it has a MINIMUM DC gain of 1000, so does the 122.............
Not sure the resistance you have going into the base of the TIP122, but perhaps the "uncontrollable" problem is related to this (inventing your circuit conditions, 30V/30Kohms = 1mA; if you have much greater resistance into the base e.g. 110K, then no idea of why that happens).
Disconnected from a circuit, I measured the resistance of the TIP125 and that of a 2N2222A from collector to emitter out of curiosity, and there is an alarming difference, my TIP125 had so little resistance in comparison to the other much lower gain BJT, which is not a Darlington, that when powered up it quickly reminded me of something about to become a short-circuit. The Darlingtons have been excluded from my transistor tester forever, unless I build a Darlington tester.
Maybe that's a good current regulating circuit you are copying, I don't know, but could be worth copying another with a diferent design - there are plenty in adjustable regulator datasheets, and unless you really need it, use a power BJT instead of the Darlington.