12x Murder Experiment - Take 2 (With 12x Comparison Plots)

I'm sorry for not posting about this sooner, but it really depressed me. I needed some time to figure out what to do... :undecided:



In any case, i have some BAD news and some "good" news.
That, and some weird observations about the bad news part....



I prepared everything so carefully this time, with all the safeties already in place from the start, and the addition of the optical "LPM" in the 12x sensor, that i was sure i could extract useful data out of the second 12x diode no matter what.

I was so sure, that i freed up an entire day, just so i could focus on the diode alone, monitor it closelly during the early hours of the experiment (by checking the optical LPM every 10 minutes) and do as many re-plots as required (and some more, just to be on the safe side).


I hooked everything up, set an alarm to go off every 10 minutes, and flipped the Cycler switch.

Immediatelly, the 12x counter registered the light, as did the 8x counter, i could see the light, everything seemed to be working fine.

Then i looked at the optical "LPM" and it showed nothing. So i start checking the connections (the "LPM" is completelly independent from everything else in the Torture Chamber, and hooked up to a DMM through alligator clips, so something could have disconnected), but everything seemed fine.

So i took a closer look at the two diodes that just started their first cycle, and i notice that the light is only coming out of the 8x!



It took a few moments for my brain to comprehend what just happened, and by the time it did, i felt like i got kicked in my stomach. I was trying to figure out other possibilities, like the driver dying, but...

The BAD news is, that the 12x is dead! Dead, after only a single plotting up to 440mA and a single power-up at 440mA. :cryyy:



I was slightly paranoid before, due to what happened with the first diode - that was the reason for all the careful preparations... I always check everything at least three times before powering it up, here i was even more careful.

But while i was "slightly" worried before, i got really freaked out when i saw this happen - again!
What the hell is the deal with these 12x's?!?



You may be wondering what could possibly be the good news after this!

Well, it's not really good, it's "good". It depends on the point of view. It's not good for me (or rather my PP account), but it IS very important to me. And it's good for our collective knowledge of these weird diodes, which are slowly starting to gnaw at my nerves....




The "good" news is, that at that point i became completelly obsessed with finding out if 12x's are any better than 8x's at all..
This was the main point for me with the second diode, since the first one got me worried, now it's even more so.

I had one 12x diode left over, so i started testing that one, because i simply HAVE to figure it out!

That's the "good" news - there will be a 12x Murder Experiment, but with a third diode...



The third Murder Candidate is a very high efficiency 12x diode, in absolute efficiency practically identical to the Plextor.


BUT, i am taking a different approach to testing this one. I'll explain more about it tomorrow when i post a report, but it's a test, which starts at a lower power for a certain amount of time and then (if everything is OK) continues at a higher power for the same amount of time, and so on.

It's similar to what DrLava suggested as a possibility, except that it starts higher, in this case at the level of the second 8x Murder Candidate, to compare degradation to 8x's first and make 100% sure the 12x is degrading slower than an 8x at the same power would, before going any further.





The weird news is, that if i was asked to identify the second (now "dead") 12x by it's electro-optical characteristics, i would say it is a low efficiency 8x diode, and not a 12x!

- It's Ith was a whole 7mA higher than with all the other 12x's - 34mA instead of 27mA.
- It's slope efficiency was that of a very low efficiency 8x. Almost too low to make sense for a 12x.
- It's Vf was lower than with other 12x's, again much closer to 8x's which have a lower Vf than 12x's on average.

It's as if i received an 8x sled along with two 12x sleds. Everything about it screamed of an 8x (the Ith was the loudest, all three together are almost definitive).
This had me confused before i started testing, but i thought, how could an 8x diode be in a 12x drive?!? Now i don't know what to think.




But there is even weirder news about it.
It IS dead. Well kinda. But at the same time, i can still bring it to lase! :thinking:

After i noticed that it's not producing light in the Torture Chamber, i immediatelly turned the Torture Chamber OFF, and put the 12x on the Diode Analyzer.

As i slowly increased the current, a normal looking spot appeared next to the spontaneous emissions! At that point i thought "oh thanks god, it's not the diode, the driver must have died for some reason or perhaps the PSU is not powerful enough to power everything!"


But as i increased the current further, the lasing stopped!
If i reduced it, it didn't come back at a specific current. But it would come back if i raised it again, and again it would stop as the current was increased further.

I repeated this a couple of times, then i tried not increasing the current past the point where it started lasing, but it would still go off within moments. After a while i could not bring it to lase anymore, as if a minor increase in temperature stopped it.

If i let it cool off it would repeat all this again. It lases, but only up to a certain current, and if left at the current where it lases, it drops in power rapidly as if minute temperature differences are affecting it dramatically.


So it's dead alright, but like the first one, it became undead, only in a different fashion.



Anyway, this is it for now. Perhaps we should all start praying for the third 12x Murder Candidate.

I need to rename this thread into 12x Murder Experiment Take 3 now. :undecided:

jayrob said:

Just hate to see it die too soon...

Click to expand...

Err... like what's your definition of "soon", Jay...? Say, perhaps you should avoid reading the above post...

Igor, don't beat yourself up, it was a lemon diode ...

"Oh great Lord over laser diodes. Please accept the sacrifice of the first 2 12x diodes and look down kindly upon this next diode." *crosses fingers*

Fantastic work IgorT.

Igor;

The second 12X to perish seems unusual.

Have you looked at the switching supply with a scope ?

Igor - On a separate board is a switching power supply. It is set for approx 3.6V out, like a Li-Ion, in order to power various Buck/Boost drivers. But it can be set higher, to power linear drivers too, the blue multiturn pot adjusts the output voltage, which is then toggled by the relay on the first circuit.

Click to expand...

Are you still using the relay, and does it have reverse schottky diode for fast spike suppression ?

I would try an 8X on the setup for a few hours,
before risking your efficient 12X.

Maybe it was previous electrostatic damage to the diode?

With all the people severely overdriving these 12X diodes, it does seem like it should handle 440ma for ~100 hours or so.

I blew several TO-56 Red diodes in recent testing,
comparing them to the flat Reds.

Momentary loss of contact took out a couple of them.

Testing diode limits will always have some failure mode.

I wish we had someone capable of examining the decanned diodes.

Who was it that took those great pictures of electrostatic discharge failures ?

LarryDFW

Bummer...

But yeah, it must have been a 'lemon'...

That is CRAZY.

Yeah. Something is odd here. My 12x #2 is still putting out over 880mW after 8 mins of runtime.:thinking:

Odd. I lowered the current on my 12X and it is Very happy at 680mW. Tests right on 680 everytime. It has at least 4 hours run time on it now and has endured 2 minute constant on times.

I have 5 12X drives in my posession. I will test 4 of them. Probably sell the 5th diode unless one dies, however I do not see this happening. These diodes are damn strong.

One thing I was wondering about. The graphs in this thread show perfect linear lines for power output with rising input, but as we know in reality, the diodes do not have perfect linear outputs. So, why do these graphs demonstrate such characteristics?

I would say because of the resolution of the equipment we use. Accuracy is one thing, but we can only be so precise with the drivers, and meters that we use for current, voltage, and power.

I thought that blu ray diodes were KNOWN for having a very linear output line. :thinking:

Meatball said:

I thought that blu ray diodes were KNOWN for having a very linear output line. :thinking:

Click to expand...

No, they have dips and kinks. Look at the 12X thread plots.

Exerd said:

No, they have dips and kinks. Look at the 12X thread plots.

Click to expand...

We used to see kinks in all the violet diodes. Then with the 8X, they seemed to disappear. However, as we push them higher and higher, we are discovering them again. We know that if you keep a diode below the first kink, it has a very good chance of living a long time. If you push it past the first kink, the life is shortened. Pass the second kink and life begins to decrease dramatically. Past the third kink all bets are off.

In the 12X plots one of your diodes shows a linear graph up to about 420mA. Conventional wisdom says that diode should be run below 420mA. At least one other shows a slight kink at ~420mA. Unfortunately, diodes are NOT predictable enough to make a good hard rule from just a few examples. Some will show a kink lower and a very few will live a long time well above the third kink.

It took hundreds of PHR diodes before we settled at the 120mA limit for them. Some still die at that current.

So we are in transition and still learning about the capabilities of these diodes. The experiment is NOT done. You are pushing the limits and finding the kinks. For most hobbyists, who desire to be a bit more conservative, they will never see a kink except in our graphs.

Peace,
dave

Pout vs. Iin is linear in ideal laser diodes. Any non-linearity is from a non-ideality in the diode, and is something that engineers are trying to avoid.

If the line is decreasing in slope, ie rolling over in power, with a CW drive then that's from heat. It is much more pronounced in red diodes, because the materials that red diodes are made of are much more temperature-dependent. Violet diodes show a lot less rollover with increasing current because they're less temperature-dependent. If you run diodes pulsed, which mitigates the effects of heat, that rollover won't happen, and the line will be linear.

Generally, "kinks" are some sort of a mode hop in the laser diode. With better-made diodes, better diode design, this mode-hopping can be avoided. This kink is a big problem for diodes in optical drives, because it makes the current-power response not only non-linear, but even discontinuous! You can't predict your power output and control it if it jumps suddenly, so diodes are always run below the kinks in optical drives. People on here have seen that lower life is sometimes observed above kinks, but the diode makers don't even get that far because the diode is of no use if it's kinking at too low of a power, making the response non-linear. So with good engineering, good design, the kinking can be avoided, and that's what we've seen with better diodes getting to faster write speeds.

@IgorT:

Mine are just hypothesis, so take them for the few they can worth ..... but these hypothesis came from the observation of a lot of sled dismantled.

6X and 8X sleds have usually a discrete complexity in optic paths and elements, and i guess 12X are almost the same (also if,til now, i still have not dismantled any 12X) ..... and i've noticed that, usually, with all the shutters, diaphragms, diffraction gratings, wave filters and lenses in the optic paths, say that 50% of the light produced from the LD is carried til the disk, is a very optimistic consideration (more probably, 35 or 40% ..... in these optic assemblies, don't count the transferred power, that what count more is the perfect shape and focusing of the beam, and we're talking about some 1/100 of millimeters alignments) .....

So i was wondering ..... there's the concrete possibility that, just making a better / more efficent optical path for the beam in the sled, that a 12X speed can be reached still using an 8X diode ? .....

HIMNL9 said:

@IgorT:

Mine are just hypothesis, so take them for the few they can worth ..... but these hypothesis came from the observation of a lot of sled dismantled.

6X and 8X sleds have usually a discrete complexity in optic paths and elements, and i guess 12X are almost the same (also if,til now, i still have not dismantled any 12X) ..... and i've noticed that, usually, with all the shutters, diaphragms, diffraction gratings, wave filters and lenses in the optic paths, say that 50% of the light produced from the LD is carried til the disk, is a very optimistic consideration (more probably, 35 or 40% ..... in these optic assemblies, don't count the transferred power, that what count more is the perfect shape and focusing of the beam, and we're talking about some 1/100 of millimeters alignments) .....

So i was wondering ..... there's the concrete possibility that, just making a better / more efficent optical path for the beam in the sled, that a 12X speed can be reached still using an 8X diode ? .....

Click to expand...

The 8X and 12X sleds are "visually" identical except in one difference. The 12X sled is missing a beam shaping optic right in front of the diode. You can compare the sleds in these links
http://laserpointerforums.com/f51/pioneer-8x-blu-ray-bdr-203-a-36790.html
http://laserpointerforums.com/f51/pioneer-12x-blu-ray-bdr-205-a-44607.html

8X (BDR-203)


12X (BDR-205)


The extracted diodes are also visually identical

The disk carrying mechanisms are different between the drives as well. Someone posted a link (I can't find it now) about a better vibration dampening system being developed right after the 8X came out (before the 12X)

Peace,
dave