I'm pretty sure I tried doing that in the beginning. Is it possible that the set load current will vary with applied voltage by tying the collector to IN+ ? That's what I seem to remember.
I'm pretty sure I tried doing that in the beginning. Is it possible that the set load current will vary with applied voltage by tying the collector to IN+ ? That's what I seem to remember.
No. The emitter currents from T1 and T2 flow through the current sense resistor (R19). You want the collector currents of T1 and T2 to flow through the load, so the only error is T1 base current. This assumes you are applying a known voltage on the control input (Vctrl), and want to be able to calculate the output current as Vctrl/R19. I realize you are not doing this.
Nevertheless, it should still solve your high current problem caused by removing the load.
So it's been two years and I just realized I never reported back. The test load has been in use at both home and mostly work without a problem. At this point I haven't needed to build the larger one although I plan to when I have time. Here's a couple shots of it.