M@g 3D/Cree Q5/alkaline batteries/voltage sag

Ok, bear with me because I get punch drunk on this stuff. I've read about a dozen threads on this site where people say to run a resistor on 3 D M@g build's that have a Seoul/Cree emitter (3.x Vf) or otherwise the LED will over current and fail.

I wanted to do an experiment on voltage sag, so I modified the switch on this new 3 D M@g build by cutting off the tower so that the power lead could be soldered directly to the switch, DD mode. I installed it in the light. I put some lightly used Duracell batteries in it, and had 4.45 volts at the top terminal of the switch. Good so far. I pulled the switch back out, soldered the LED power lead straight to it, put it all together, and fired it up DD for a few seconds to see what the voltage at the emitter was, and it was 3.45 volts.

How is that possible? It should not have lost that much voltage, should it?

It's bright, but not all that bright. I turned it back on and it's been running for about two hours now. Everything is ok.

I'm going to pull this emitter off and put a Seoul on it tomorrow night. This Cree doesn't have a very good beam pattern, as I expected from reading others experiences. I have 5 Q5's and only 2 Seoul P4's so I preferred to risk a Cree for this test. It's ok though. It didn't see the voltage I thought it would.

So why would I want to run a resistor on this build as long as it's only going to run on alkalines?

Yes, it's normal for alkaline voltages to drop that much under 1+ amp load like you are probably feeding that Q5. I've done a couple builds direct driving a Q5 off 3D alkaline cells. Using brand new unused cells for a worst case scenario, I found that a 0.10 ohm resistor gave me 1.15A to the LED. Just a hair over the rated current but tolerable. That's why alkaline cells are poorly suited to even moderately high current uses, the voltage drop is terrible. But, in the right application, you can use it to your advantage.