8X Diode Murder fund

[SHIN]

Any diode murdered?:thinking:
There are only good men, not murderers.:yh:

 

[IgorT]

BTW, did you say before that multiple diodes could be tortured at the same time with your setup?

The setup can torture multiple diodes, yes, but it would need to be expanded to also record the others...

It works like this. The cycler toggles a relay, which simply connects a driver to the output of a switching PSU capable of 3A. It's currently set to output the voltage of a Li-Ion to power switching drivers. The switching PSU is adjustable and can be set as high as Vin - 3V to power linear drivers...

With the 3A output ability, it could cycle multiple drivers side by side. If that's not enough, the relay has another side, which simply acts as a switch. The other side could be used to connect more drivers to an external power supply, if more power was needed. I could run several drivers off the cycler directly, and even more with the use of external power and the other side of the relay...


Problem is in recording all these diodes.. The cycler can cycle multiple lasers. But each laser would need it's own sensor. This means more photo-transistors in similar housings to this one, and more counters.

However if i was to cycle multiple diodes simultaneously, i would use time recorders instead.

The time recorders are tiny MCU's meant to be put in lasers to record the total ON time, so that when a laser dies you take it out and read off how long the diode ACTUALLY lived (because people often exaggerate in their estimates, besides it's not easy to estimate accuratelly)...

Zom-B made these time recorders a while ago and sent me several. I don't have the room to put them in lasers, but i wanted to use them in my cycler.
Otherwise, people who don't want to kill diodes but just USE their lasers could still contribute to the overall knowledge if they put such time recorders in their lasers. But i don't know if Zom-B is still around or making them...



Anyway, at this moment, since it's about expensive diodes, i am using BOTH the counter AND a time recorder wired to the photo transistor, as a backup in case of power failure (even tho the counter officially has a 50h data memory, and i tested it, i am still paranoid).

If i was testing multiple diodes, i would only use time recorders most likelly, altho i have recently found MUCH smaller LCD counters, which are actually quite cheap (and use up MUCH less power than this LED counter from e-bay)...


If i had to expand the setup now, i'd use time recorders tho, since i have them on hand.

So, every diode tested needs it's own driver, as long as the total current draw does not exceed 3A they can be powered by the Cycler directly, otherwise externally (where i could hook up 50 lasers if i wanted to - through the relay), every diode needs it's own sensor, and IF they are tested side-by-side, they can obviously only be tested with the same cycle duration.

That's what i would use to race diodes against each other, like comparing what happens with a high and a low efficiency GGW at the same current - which one would die first.... I could do one after another, but why waste the time (only my old cycler was limited to one diode at a time).


Also, while the test on one diode is running, i could at any given moment simply add another diode, as long as i aim it at it's own sensor connected to it's own time recorder or counter, it would get tested properly and completelly independent of the first one, only the cycle interval would be the same.




Otherwise, Bryce, if you're willing to edit the video and increase the speed and all, no problem.. I'll just put my digicam on a tripod and let it record for a while.. But i have no experience editing and posting videos, not to mention the time... :undecided:




EDIT: Attached is a pic of Zom-B's time recorders and a reader.
Both the reader and the recorders have a microprocesor, the one on the tiny recorders just records the time one of it's pins it's toggled, or in a laser, when the laser is turned ON.

If i had the room to add them, i would put them in my lasers, especially the ones with more expensive diodes, such as 8x's. That way even the lasers made for normal use would provide useful info over time.

The recorder can be taken out and put onto the reader at any time, and the LED display will tell the time in seconds (one digit at a time), then it can be put back into the laser and keep recording the time. It can measure way further than any diode can live even with normal use.

I have four of these recorders and one reader, one of the recorders is now in the "Torture Chamber", serving as a backup, in case the counter's data memory was actually needed and would fail, or any other unexpected problems would occur. Also, to compare the final readings from the counter and the time recorder, to make double sure it's correct...


The recoders have a size just a tiny bit less than the latest flexdrives, i don't know if Zom-B ever finished the project and offered them, but if they were available, anyone who would be willing and able to use them in their lasers would be making a contribution to our knowledge about diodes...

I mostly got them so i could torture multiple diodes at once, if needed...

 

[HIMNL9]

It's inquisition time. Guilty or not, the diodes are gonna confess before they die a slow and painful death! :eg:
May it be as slow as possible!

..... All the other diodes are shaking in their antistatic bags, when they hear the screams of the first one, since they don't know who's next...

..... and i don't care if my neighbours start complaining about the screams! .....

I was not knowing you was so sadistic, IgorT ..... (j/k )


Edit: but judging from the other comments, you're in good company, around here ..... ..... (LOL )

 

[IgorT]

I was not knowing you was so sadistic, IgorT ..... (j/k )

Edit: but judging from the other comments, you're in good company, around here ..... ..... (LOL )

This thread was sadistic from the start.. :evil:

Just look at the topic! Everyone seemed excited about the murder...
Unfortunatelly for the diodes, they have to be tortured first, before they can be allowed to die.


Besides, how else to make them talk, but to torture them? :angel:
I'm all out of truth serum...

 

[Crazy Jay]

Uh... er

Are you testing the limits of diodes in a scientific manner or are you torturing little furry animals?

 

[daguin]

Uh... er

Are you testing the limits of diodes in a scientific manner or are you torturing little furry animals?

A guy's got to keep himself entertained while waiting for the diode to fail, doesn't he? :eg:

Peace,
dave

 

[Crazy Jay]

A guy's got to keep himself entertained while waiting for the diode to fail, doesn't he? :eg:

Peace,
dave

haha very true. I get occasional field mice that come in the house and those tiny little spring traps aren't sensitive enough to "trap" the mouse. Then I tried glue traps and saw this little mouse run back to his hole and suddenly got his tail trapped... then in a matter of seconds his whole body. I recommend them for any pest with legs.

Heck if you put your fingers on it you're going to have a bad time.

For Diodes I suggest sticky electricity

 

[Krutz]

..never heard about glue traps before..
Facts: Glue Traps
it sounds horrible! who on earth would use that?

/threadjack

manuel

 

[Crazy Jay]

..never heard about glue traps before..
Facts: Glue Traps
it sounds horrible! who on earth would use that?

/threadjack

manuel

I have. They aren't inhumane.... to loosen them from it you use typical cooking oil. It allows them to slip out. Of course I would never do that ... I don't want them to return :eg::evil:

 

[FrancoRob]

I fear i have some bad news...

For diodes, that is... :evil:

The Diode Torture Chamber is complete!!!







​

It's fully up and running, doing what it was designed to do - torturing diodes!!!!

It's inquisition time. Guilty or not, the diodes are gonna confess before they die a slow and painful death! :eg:


May it be as slow as possible!

...the Grand Inquisitor....

 

[IgorT]

Bryce requested some pics of the setup in action...

I took some at the very start of the torture, and some more after 5h+ ON time...
These two are the best i could find among them (tried to make sure the beam was visible)....


The first pic was after 7 minutes ON time:

You can see i rearranged things a little bit after fine-tuning with a LED (CREE stuck in the sensor tube - spent most of the day yesterday making sure the accuracy was better than 0.5%). I moved the Cycler/PSU circuit to the front, so it wouldn't be crowded around the laser and sensor. I also added yet another switch and some connectors, to help me with checking the power and re-plotting the diode.

I also used my label printer to mark the switches...


Behind the laser module in the above picture, you can also see the driver powering the diode....

It's the full Buck/Boost version of my BR Boost driver. Unlike the boost only version, this one can power any diode (so it's very suitable for a setup like this), but is much larger (since it requires two coils). This one is actually one of the first i've ever made.

Of course i upgraded it to the latest version before using it here (i changed most of the capacitors), to maximize stability and filtering, so it's output is as clean as with the boost only versions (soft startup and less than 5mV peak-to-peak ripple)...

The output has three separate capacitors, two to prevent spikes and soften the powerup blow to the diode (and absorb most of the switching noise), and a third much smaller value cap to reduce the high frequency switching noise to a minimum. Since the diode is distanced from the driver, connected to it with wires, and the driver is a HF circuit, there is a fourth capacitor directly on the diode.



If you look at the beam, you will notice how it expands before it hits the sensor. I intended to expand it from the start (so it would cover the entire diffusor area), but when i first started this monster up, it immediatelly burned my hand, and as soon as i removed it, smoke started coming out of the sensor.. :crackup:

I am using Jay's 405-G-1 lens here, cos plastic lenses would degrade rapidly in an experiment like this, while i have to check the power every so often and even re-plot the diode - that's the reason the test laser has the 405-G-1 inside..

And as some of you know, 8x's hurt with thin beams, even when the beam is parallel!


I immediatelly expanded the beam, so the laser wouldn't eat through the diffuser...




Here's another pic, after 5 hours and 9 minutes of total ON-time (= 10h 18m into the experiment)...

If you look carefully, you will see that the temperature climbed to 28°C. It was a good decision not to use a laser pointer as the host for this experiment. Again, i don't believe it would overheat, but the temperature it would stabilize at would be quite a bit higher than this...



At the start of the experiment i was very worried, i waited for the first hour (ON time = 2h total time) to be over and then re-plotted the diode, to check for degradation. The diode was fine at that point...

I wanted to repeat it at the two hour mark and use the results to decide what intervals i should re-plot the diode at, but the alarm clock didn't wake me up..... :yabbem:



I woke up an hour ago, stopped the cycler, and the laser is cooled down to the initial temperature. That's why the setup has a thermometer - so i don't confuse efficiency variations due to heat with degradation...


I will re-plot the diode now, and report back...


P.S. I hope i won't have to re-plot the diode too often, that once a day will be enough. That way the diode will go through almost 12h of ON time every day..
Otherwise i prepared a new Excel sheet again one for multiple plots, but this time of one and the same diode, so as to record degradation.

 

[daguin]

op::beer::bday::barf:

Peace,
dave

 

[Crazy Jay]

I wish I had a bluray with a beam like that. I'm lucky to see mine with smoke in the dark

 

[IgorT]

Ok, diode is re-plotted and first signs of degradation are starting to show up at the 5 hour step...

I am hooking it up back to the cycler now, hopefully i will get to re-plot it a few more times, before it expires.
So far no kinks are developing, but the efficiency drop is more noticable than i expected this soon..

The attached pics show three plots. The 0h plot and the 1h plot are overlapping. The 5h plot is showing a drop in all three, most noticable in efficiency itself.


While i am a little worried, keep in mind, this was the lowest efficiency one of all diodes. It actually stood appart from the crowd, it was noticably weaker. And yet it survived 5h of toggling ON and OFF already. 309 average usage cycles...


I think it can and will go on for a while, but how long exactly is impossible to say just yet.
It hasn't developed a kink yet, i guesss that's the good news (altho it's not necessary, that every diode degrades like that), but it doesn't look like it will last for 100h... :undecided:

I hope i'm wrong of course.. Maybe i should wish for it to die. Usually if i make a laser with the specific purpose of killing a diode, it doesn't want to die.. FrancoRob has a 320mW red that was a result of such an unobedient diode.. :crackup:




Anyway, here are three separate plots. P/I, Po/Pe and Efficiency/Pe...

In all three 0h and 1h overlaps. 5h makes a distinctive new line.

Additionally, Ith climbed by 1mA - i was checking them preciselly for this reason, cos i read 405nm diodes degrade by the Ith climbing but that the slope efficiency remains the same.

This last part does not seem to be true however. Because the slope did drop in it's steepness (instead of the entire plot simply moving to the right).


In any case, the experiment continues, and i'm afraid i will have to replot the diode two or even three times a day, if i want to capture what is happening before it goes LED...



Other degradation details (from 0 to 5h):
- Ith went from 34mA to 35mA
- Slope Efficiency went from 1.339mW/mA to 1.318mW/mA
- Po @ 300mA went from 356mW to 351mW

 

[IgorT]

I wish I had a bluray with a beam like that. I'm lucky to see mine with smoke in the dark

CJ: There are no blu-rays with a beam like that. I use smoke when taking beam-shots...

BluRays do make lovelly beams outside at night or even inside if no too powerful light sources are present, but here i had to light up the entire setup, and at the same time capture the beam (since Bryce wanted to see it in action), so i had to use some helpers (AND i always take tens of pictures, to select the one where the beam is most evenly visible for posting)...


Incense sticks, matches, things like that help a lot, if you don't have a smoke machine.

 

[IgorT]

I was speaking of a diFFerent diode

Oh, you were asking because of the F03's...


I can, and probably will hook up one or two of them during the second run of the experiment, side by side with 8x #1...
But it'll take "some" more work, plotting them, making more sensors, finding room to add everything....


Even as it is, this experiment, or rather the preparations for it, already ate three times more time than i had allocated for it...

Plotting all the diodes, preparing the "Torture Chamber" and fine-tuning the cycler circuit was so much work, that all my free time over the last two weeks went into it! I was barelly able to follow anything but the related two threads, i'm behind with emails and PM's, i have a bunch of laser work to do, and even my real work suffered because of this. :undecided:


I wish i could just let it run now till it dies, cos i really need to get some work done, but it seems i will have to keep checking the diodes at least two times a day...

However, at least this last part i can probably make easier... Maybe i should only check two points. Ith and 300mA power. That would speed things up a little...



In any case, i REALLY hope this first 8x diode lasts a while, not just for the obvious reasons, but also to allow me to do all the work this experiment delayed so far!