really? I hate to say sorry again... I havent had to isolate anything. I connect the - to - the + to + and the case pin to the - on the Vin - in side. I will try not to give any advice in the future. I just want to run away and make lasers... Why was that bad advice?
What you've just said here ^ is correct. Don't beat yourself up!
BUT - it's only correct if you're using 405, 445, or 635 diodes. Those are
already isolated from the case. IE, they're case neutral. Their case pins aren't attached to any part of the emitter themselves. The case pins are almost the same as if you just soldered a wire to the inside of your host walls.
It's a different story with IR diodes, because they are not isolated from the case. The case pin is actually continuous with the (+) pin. Thus, you would need to isolate them to use a driver without a continuous + or -. (preferably continuous +, but Jib had a creative way for making continuous ground drivers work with case positive diodes too)
Also, I would clarify that what you described when you said:
1)
"If you want a continuous ground just join them"
Isn't actually what you described when you said:
2)
I connect the - to - the + to + and the case pin to the - on the Vin - in side.
In comment 2, all you're doing is taking the battery (-) feed from the tailcap, via the case pin, and inputting it into the driver. Then you're taking the driver's (-) output and sending it to the laser diode emitter. But the (-) in and the (-) out are never actually continuous in this scenario. Regulation is happening in between. They're not a single connected feed the way you've described it. So you're not actually doing thing #1, which is what I was warning against.
This is good. You should never do thing #1 with a non continuous driver
I just got back from physics today with tests on the oscilloscope! Turns out, at most, 50mA ripple when in regulation, 100mA ripple when just getting out of regulation. But the ripple is so quick, I doubt even a diode would notice it - we are talking much faster than 1MHz.
