Current for the 12x murder experiment.

[Traveller]

So... what's it gonna be, Igor? Meanwhile, we've got mad scientists hitting >750mW @450mA... :wtf:

Btw, why's this topic in off-topic???

 

[IgorT]

I thought I read in the murder thread that one of the diodes was was emitting only spontaneous emissions (LED as we call it)?

Yes, the one set to 453mW initial power went LED at just under 195 hours.
195 hours is a lot, and so is 450mW, so i'm impressed...

Based on the curve of the diodes graphed in the 12X diode thread, and as Hemlock Mike pointed out, there seems to be 2 distinct lines showing up which might indicate 2 different diodes at 2 different efficiencies.

The lower efficiency 12Xs seem to follow a similar curve to the 8Xs you tested.

I have not seen data sheets for either the 8X or 12X diodes (do we have them?) so I would think that it might be possible that some 12X drives are using the same diodes as the 8X drives. However, I do not think this is very likely.

I am 100% sure the 12x's are different from 8x's, because:
1. Their Vf is higher on average (the Vf curves form their own group above that of 8x's)
2. Their efficiency is higher on average.

Some lower efficiency 12x's may follow the same PI slope as a high efficiency 8x, but their Vf will be noticably higher, and when observed as a group, they form a higher efficiency group with a higher average.

Efficiencies may vary as much as to make someone think they are different diodes, but then again, so do the efficiencies of the tested 8x's.

The highest efficiency 12x diode i tested so far came from a Plextor drive, but the sled was completelly identical to the Pioneer sled.


I will be comparing PIV data of three further 12x diodes now, and then i'll know more, but for now i see no reason to believe they are different diodes.


I may have missed some data, but even if another Plextor tested higher than average in efficiency, it's not enough to reach a conclusion.

I have a feeling one of the three Pioneer 12x's will come close to the Plextor. In total i will have the data of six diodes to compare, and i believe we will simply see a group shifted higher in slope efficiency, compared to the 8x's (but i could be wrong - i'll know when i finish testing).

Exerd is testing one 12X to very high powers (over 700mW) and has about 7 minutes on this diode. Already he is noticing RAPID degradation.

Ok, this is very good to know, and it only makes sense. I'm glad others are reporting if and how the power is dropping, it will help with the decision!




Anyway, initially (when i thought i'd pay for it myself) i said i would test it my own way. But i may often be too conservative for most LPF member's wishes (you'll have to forgive me, but my usual work drives me to obsession with reliability).

So after so many people chipping in again, to pay for the new 12x Murder Candidate, i find it impossible to reach a decision by myself. It's just too much responsibility.



I don't know, perhaps we need a new poll (what POWER to test, not current)?

Personally, i have a feeling we shouldn't go over 600mW - that's even if i forget the evil smiley face of the first 12x.

600mW for a 12x would be like 500mW for 8x's. We haven't tested 500mW with 8x's, but the lifetime at ~450mW seems to indicate there would be quite some time left in them there...

And if it's possible to reliably reach 500mW with an 8x, testing only 550mW seems too low.. :undecided:


If it wasn't for the scare i got last time, i'd say let's go for 600mW. But then again, that could just have been a freak accident, and no reason for concern...


But it's impossible to know for sure, so i will need suggestions.


I would also very much like to hear as many degradation reports as we can gather from current 12x owners at powers from 500mW and up. Especially from those at very high powers.

What was the power initially? What is it now? What is the estimated run-time?

 

[StridAst]

My vote is to set it to 550mw. If we push to 600, and the diode dies early, we have data that isn't particularly useful. 550mw is a conservative amount, but what good would another evil smiley face diode be? The first was probably a freak accident, but I am the paranoid type. My vote is 550mw. If it performs with long life, then we could assume 600 is still reasonable with shorter life but not too short.

 

[IgorT]

My vote is to set it to 550mw. If we push to 600, and the diode dies early, we have data that isn't particularly useful. 550mw is a conservative amount, but what good would another evil smiley face diode be? The first was probably a freak accident, but I am the paranoid type. My vote is 550mw. If it performs with long life, then we could assume 600 is still reasonable with shorter life but not too short.

Looks like my last post didn't take. I was having some trouble with my internet connection and couldn't get online for almost three days.

But i was writing that before i finished testing the diodes anyway...


Mostly i was just saying that i would also be happy with a 550mW test, but that i fear i am beeing too conservative about it. I don't want to make those who donated unhappy with the decision. :undecided:



Well the diodes are plotted now, and i have to say i am slightly more concerned about the amount of current i am about to pump into this tiny thing.

Before, i kept saying we should think in terms of power and not current, because the 8x experiments made me believe the power output is the determining factor for diode life (as opposed to the current setting - regardless of the efficiency), now i almost wish i didn't say that, even tho i know it's the optical flux at the die that degrades the diodes, and not the current producing it... :yabbem:


While testing the diodes i set the upper limit on the Diode Analyzer to 400mA. Even that was a little scary to me.

But with this diode i had to plot to 440mA to get to where i wanted...



Here is the second 12x murder candidate:

The slope efficiency is lower than it was with the previous 12x diode.
But i made several different 12x comparison graphs, and the 12x's form a very similarly behaving group as the 8x's did (only at a higher slope efficiency and higher forward voltage on average).


The 8x's have already prooven themselves above and beyond all expectations, even tho the two test diodes were both the lowest efficiency ones of a much larger group, so i don't think i should worry about the efficiency of the diode here...

If 12x's are tougher than 8x's (as they should be) and capable of delivering higher powers, the efficiency shouldn't be a factor here. In comparison, the two tested 8x's could easily have delivered an initial power of 400 and 500mW respectivelly with a decent lifetime.



One interesting thing in this case is, that the Vf of this diode is much lower than that of the others, so when it's output power is plotted against the electrical power going in, the diode comes much closer to the rest of the group (in terms of actual efficiency it's better than it seems, because it requires a lower voltage to reach the same currents)...



In any case, 550mW would require 430mA with this diode. But seeing they are surviving even higher currents for measurable (or should i say testable) amounts of time, i still think it would be a good point to test.




Just in case i went with HIMNL's idea and rebuilt one of the sensors in the torture chamber to allow me to track the power of the diode without having to interrupt the cycling:

It's just a tiny solar panel in the back of the sensor's enclosure, but by measuring the current it produces through a fixed resistance (and multiplying it with a calculated factor) it becomes a simple power meter.

It's not terribly accurate, but it will be reassuring to know that the power is not dropping too fast in between the re-plotting intervals.



So, everything is ready and if we can agree on a decision for the test current i can start the torture....


Needless to say, i will be extremelly vigilant during the first hours, to make sure the diode is off to a good start... Now that i have closelly studied the degradation behavior of two 8x's, i should be able to tell if all is going well after only a short amount of time.

 

[Traveller]

In any case, 550mW would require 430mA with this diode. But seeing they are surviving even higher currents for measurable (or should i say testable) amounts of time, i still think it would be a good point to test... if we can agree on a decision for the test current i can start the torture...

Well the votes have long since been in (I think) but you're still the MC here, Igor, so 430mA has my vote (too) :takeit:

 

[IgorT]

Well the votes have long since been in (I think) but you're still the MC here, Igor, so 430mA has my vote (too)

That's true, but the votes were for the old diode. Some later additions may be for the second attempt, cos i think the poll results changed a little... Not sure... :thinking:


I don't know if this is even the right place to ask this question, i just realized this topic is in Off Topic for some reason, and may not get the attention it needs.

 

[cmak]

thread needs to be moved to BR section, as IgorT just mentioned :lasergun:

otherwise great work guys! :gj:

peace & thanks
-cmak

p.s. love the solar, IgorT