- Joined
- Aug 25, 2007
- Messages
- 2,007
- Points
- 63
That's not the correct way of thinking about efficiency for laser diodes. The differential, or "slope", efficiency of a laser diode is proportional to the slope of the current vs. power output curve. The differential/slope efficiency is the overall efficiency of the diode. The slope of his chart is basically invariant across the whole range given (R^2 >.999 for a linear fit, so it's a very good line, and doesn't curve visibly), so the efficiency isn't changing. Plot the data, and you can see why there's a point where the mA is equal to the mW with these particular diodes: the slope is greater than 1, but the line is offset because you have to reach a certain current before lasing begins, ie threshold. You can move the "mA = mW" point by changing the efficiency (ie the slope) or changing the threshold current, and either one would change that "mA = mW" point. So, think of efficiency as the slope of the region, and threshold as another important factor in diode operation.
With laser diodes, if anything, the chart gets "less steep" as you increase current, as they get less efficient as you increase current. This is especially evident with CW operation due to heat: if these lasers could survive the optical output, the slope would keep decreasing, and could actually roll over and the slope would become negative, but the diodes die from the optical powers before that happens. With these diodes, you really can't see any change in slope on the ranges given, the optical flux seems to kill them before other effects (eg heat) make the efficiency drop.
According to your second paragraph, you say that the slope on the graph is increasing at a decreasing rate. This makes sense no doubt, because of the heat. So naturally with an increase in heat, we should also look for an increase in electrical resistance across the diode. Someone ought to plot us a current vs. resistance curve, and according to what physics says about heat and resistance's direct relationship to each other, than we should expect to see a concave up curve on such a graph. With just a bit of Ohm's law, should we not be able to figure that at 350mA, this diode's electrical resistance has been tuned up to 16 ohms? That seems like a lot for a diode....
