I placed the motor in the right place, worked for 1/100s then stopped. I had forgotten pin13 to ground without resistor, from testing. Killed pin13 on my microcontroller.
This sentence is really a boggle.
There are multiple problems with your ckt. First, the original circuit is an emitter follower config,
and you'll not get much more voltage on the motor than Vbase1 - 0.7 - 0.7 - 0.7 = approx 3V.
Secondly, as elsewhere noted, a 3 transistor Darlington probably has too much gain for life in
this universe. Leakage currents alone in Q1 will be multiplied greatly and trigger the motor. You
absolutely need R2 to ground to hold the ckt off with no input.
Thirdly, if you fried pin 13, you connected the motor up wrong, and it's not at all clear from your
"boggle" statement above what you actually did.
It would also help to know how much current the motor draws when running flat out.
I'm curious though...what's the explanation? Why would everything work with the motor between collector and ground when the arduino was not involved? Feeding the base of Q1 with power from my breadboard worked just fine. Hmm...
Did you mean "emitter and ground" here?