I spoke to Lazeerer though and he pointed out that I had been using the wrong pin setting for Juffran's Test Load. He said the red setting, I was using, was only for 650 - 670 nm diodes, NOT the 635's! The 635's have a lower voltage drop and so pin 2 must be used! I tried this and got 1.4 A! :tinfoil: It also means that my first attempts to solder the chip had probably been successful, so many hours wasted there...:toilet:
He also said that there are 2 possibilities; 1 the driver had a current spike, 2 I could have made a continuous path to negative through the heatsinked driver. Unfortunately I can't test either of these, as I don't have a scope and I had already disassembled the pill by the time I spoke to him.
Anyway, If I manage to reconnect the diode bond wires I will start a new thread and explain how to do it.:thanks: again for everyone's input!