IgorT
0
- Joined
- Oct 24, 2007
- Messages
- 4,177
- Points
- 0
rocketparrotlet said:
If you look at diodes datasheet, they have a certain life time specified at a certain power. And at a certain temperature of course. Heat makes them age faster, yes, but that's a completelly different story - in drives, diodes are powered at a constant power, not constant current.
- At a constant power, heat means the driver has to pump more current through a diode, to reach the same power - it increases the strain (same power MORE current)..
- In pointers, we use constant current, and heat just reduces the power - it reduces the strain (came current LESS power). That's why constant current is normally a safe way of powering diodes.
There is a reverse scenario now appearing, and from reports it may be more common in 4x's, but it was present in some weird PHRs - some of them CLIMB in power with heat! In this case it would look like this:
- At a constant power, heat means the driver actually has to use LESS current if the weird diode increases in efficiency with heat. So the driver will lower the current to keep power the same - the strain will reduce a bit (same power, less current).
- At a constant current, heat will increase the efficiency of the diode and the power will climb - the strain will increase (same current, MORE power).
It's weird, but a diode like that could kill itself under conditions, that are normally accepted to be very safe for diodes (constant current).
Anyway, about the lifetime. Their life is specified in MTTF (Mean Time to Failure). It can mean, that in that time one half of the diodes will die, or that in that time diodes will lose half their power.
It's really hard to say what exactly is going to happen.. At 160mA, they drop in power just a little bit and then die. At higher currents they can just die. At 125mA, they can either die completelly, or they can slowly degrade. Or they can just keep living. With diodes like this, you never really know.
It probably depends on HOW defects appear on the die. If the die is perfect, and one point on it fails and starts causing damage to areas around it, and slowly spreads till it causes COD, the diode will just turn to LED - once the defect becomes big enough, it will take the whole diode down with it. It's like a tiny fracture in the car's windshield, that start's spreading, and when it becomes big enough, it just shatters.
If the die has some imperfections already, and they are evenly spread, the "pressure" will be spread out more evenly as well. In this case the imperfections could grow, but being spread out, they might not cause it to die instantly, but instead the power would gradually drop. As the power drops, so does the pressure. If the ratio is just right, the diode could slowly drop in power, but not turn to LED.
So theoretically, it is possible, that high efficiency diodes would die instananeously, while medium or low efficiency diodes would slowly degrade but not die completelly. Pre-existing imperfections can sometimes make things tougher.
Because of this, i tested diodes from all efficiency ranges i found. If i pushed a low efficiency diode to 190mA where it only did 130mW, it died instantly. At 100mW it may have died gradually.
So it's impossible to say what is going on with your diode without using a power meter. The PHRs are strange diodes, that vary in everything from efficiency, wavelength, beam profile, to how much power they can survive for how long.
It could be just fine. Or it could have lost 10mW by now. Small power drops are impossible to notice without a meter. But i have seen diodes, where the power dropped enough, that the user noticed that it doesn't burn as well anymore.
Even if you don't have a meter, you could make a diode "health checker" using a solar cell, in some kind of a box, shielded from external light. It won't tell you the power, but it will tell you if the power is dropping. Altho, small changes in the reading can come from the wavelength drift, with temperature and power drop with temperature. But they will be reversible. If you see a steady decline, that doesn't reverse itself with temperature, then it's degrading.
In any case, unless we were to set them to 60mW, they will die or degrade in under 5 - 8000 hours. The higher you set one, the faster it will degrade. We don't need them to last 5000 hours. We always overdrive diodes. All the reds are overdriven as well. We risk a little, shorten the life, but gain power.
The real question is only this: How much are YOU willing to risk? How long do YOU want or need your diode to live? Do you want more power, or do you want longer life? That's all it's about.
I set my personal lasers high, but i don't expect them to live very long. If i make a laser for someone else, i want reliability over power. If they want more visibility, i select a higher wavelength diode, rather than setting the current higher. If they want more power, i make them aware of the consequences. If they are OK with sending the laser back to me, when the diode burns out, and if they know that it won't live very long, then i set the laser higher sometims. But i have not gone over 143mA, and while some actually survive this i don't even want to do that anymore, because i now know it's just a matter of time.
At 125mA, the losses are low enough for me to consider them acceptable. But i think this is just on the boundary where it can go either way - either a diode turns to LED or degrades. I have no clue how many are actually degrading. Not everyone has a meter. Theoretically, many could have dropped in power considerably by now, depending of course on how much people use them (i don't limit their use in any way, so it's random).
This is why i am very glad, that now 4x's are becoming available.
Of course now than I said that the thing will break tomarrow. 
