PHR-803T batch testing.. Windowless 803Ts?

[pullbangdead]

pullbangdead I tried to rep you "+2"but it backfired and now you're repped "".

LOL, that keeps happening to me for some reason, Krutz tried to to +1 and it ended up -1 as well, that's the deleted one that is on the list. Not a big deal, c0ld can clean it up at some point. Thanks!

 

[IgorT]

Well, the first diode, which i saw exhibit the mode hopping did so at a certain power, not current.

When i reduced the current, the spot became normal, but then it cooled off, the power increased slightly, and it distorted again. I had to set it even lower, for it to stop.

But just now i set it to 185mA.. When i put it together, it was doing 165mW, dropping to 155..

Just now i turned it on again, and noticed, that the power was >177mW and the spot round. But then the spot deforms and the power drops..

I think it must be just above a power related "kink/mode hop", and as it warms up and the power drops, it falls into the kink area. I think i should raise the current slightly, to keep it out of this area...


It should work. Your graphs seem to suggest so too.

 

[rocketparrotlet]

Oh my goodness...I just can't believe this. You guys are repeatedly pushing these things no problem to 250mW?! A month or so ago, when I started researching Blu-ray here, I thought that 120mA was going to be about as high as I could get these and have them live for a long time.

Would anybody be willing to push a diode like this for me until you get a high current, or at least determine approximate maximum current? I will be making my build with a jayrob heatsink, so I think if I used a reasonable duty cycle, I could sustain a high current safely. Please correct me if I am wrong. (Note: I am not sure I want this done. I just want an idea of whether or not someone would be willing to do this for me for a reasonable price; I don't have a power meter.)

-Mark

 

[pullbangdead]

Oh my goodness...I just can't believe this.  You guys are repeatedly pushing these things no problem to 250mW?!  A month or so ago, when I started researching Blu-ray here, I thought that 120mA was going to be about as high as I could get these and have them live for a long time.  

Would anybody be willing to push a diode like this for me until you get a high current, or at least determine approximate maximum current?  I will be making my build with a jayrob heatsink, so I think if I used a reasonable duty cycle, I could sustain a high current safely.  Please correct me if I am wrong. (Note: I am not sure I want this done.  I just want an idea of whether or not someone would be willing to do this for me for a reasonable price; I don't have a power meter.)

-Mark

That part I emphasized is the other important part not really known yet. Seeing these kinks and mode shifts is giving us a lot more information, and some are living at really high currents, but they may well be degrading invisibly before our eyes at this point. How long is long enough for you? It way well be that 120mW you'll never hit its lifetime. You can look in datasheets for the CW rating where the diode should never die. At these levels, the higher you go, the less time it will live, that's just how it is. So how long is long enough? I'll probably never get to 100 hours of use on a single diode. Maybe it's worth it to some people to have the maximum amount of power even if the diode only lives for 5 or 10 hours of use, since they're less than $20 apiece now. So we'll see how things keep going, and then maybe figure out some better way to know where to set diodes.

 

[Zom-B]

I would be willing. As long as it contributes to the collective knowlkedge of blu-rays ;D. I'm, in Europe though.

 

[Zom-B]

[quote author=rocketparrotlet link=1214268330/80#98 date=1223687846]Oh my goodness...I just can't believe this. You guys are repeatedly pushing these things no problem to 250mW?! A month or so ago, when I started researching Blu-ray here, I thought that 120mA was going to be about as high as I could get these and have them live for a long time.

Would anybody be willing to push a diode like this for me until you get a high current, or at least determine approximate maximum current? I will be making my build with a jayrob heatsink, so I think if I used a reasonable duty cycle, I could sustain a high current safely. Please correct me if I am wrong. (Note: I am not sure I want this done. I just want an idea of whether or not someone would be willing to do this for me for a reasonable price; I don't have a power meter.)

-Mark

That part I emphasized is the other important part not really known yet. Seeing these kinks and mode shifts is giving us a lot more information, and some are living at really high currents, but they may well be degrading invisibly before our eyes at this point. How long is long enough for you? It way well be that 120mW you'll never hit its lifetime. You can look in datasheets for the CW rating where the diode should never die. At these levels, the higher you go, the less time it will live, that's just how it is. So how long is long enough? I'll probably never get to 100 hours of use on a single diode. Maybe it's worth it to some people to have the maximum amount of power even if the diode only lives for 5 or 10 hours of use, since they're less than $20 apiece now. So we'll see how things keep going, and then maybe figure out some better way to know where to set diodes.[/quote]
Maybe it's time for me to mass produce my time recorder .

 

[Zom-B]

BTW, before I go to bed, I forgot to mention some very important key variable in the tests I did:

The room ambient temperature was 20.4°C (69°F) at the time of the measurements, and the difference of working during the day and night is about 1.0% (linearity of the sensor is disputable).

 

[rocketparrotlet]

Approximately how long do you think a diode would last at 250mW, and what would the duty cycles be for it? Would a jayrob heatsink be sufficient cooling?

-Mark

 

[Zom-B]

We'll see. Not many diodes have reached 250mW so there is not much to compare with. Maybe daguin knows more.

 

[IgorT]

Approximately how long do you think a diode would last at 250mW, and what would the duty cycles be for it?  Would a jayrob heatsink be sufficient cooling?

-Mark

If you use Jayrob's heatsink, it will keep the diode under 30°C for an hour straight. But the diode might not live that long.

I don't think you can get that kind of power out of a PHR with any kind of reliability.. Many will jst pop on powerup.

 

[IgorT]

Zom-B:

I took my last mode-hopping freak from the old batch yestrday (the one i mentioned in the previous post), and wanted to set it to 185mA. But i fried the last of my boost drivers by accident (a wire disconnected and shorted something - smoke appeared), it was late and i didn't want to make another, so i slapped in a flexdrive, set to 185mA instead...


Unfortunatelly, the diode was again inside another mode-hopping kink. It started with a nice beam, but as it warmed up and the power dropped slightly, the beam split, the spot became deformed and the power dropped massivelly.

But as the battery voltage dropped, the spot suddenly became perfectly round again, and the power was 194mW holding steady.. Unfortunatelly the current is only that high shortly before the battery shuts down.

This means it takes just a little more current, to get it out of that kink. Like 4-6mA.. On an empty battery it does 194mW and on a full one, where the current is only a couple mA lower, it deforms and drops to 153mW or so.


I need to take it appart and test what current the flexdrive does on an empty battery, and set the diode there permanently.


Hmm, i have an idea.. Indeed it works. I just powered it from a 3V Li-Ion instead of a Li-Ion, and now it's suddenly doing 199mW with a perfect TEM00 spot! Need to figure out the current now and then make a new driver. Unfortunatelly this one is stuck in the heatsink rather hard..  


EDIT: Damn! The AixiZ module was stuck so hard, that i broke the positive leg of the diode by accident..  :'(
Now i need to insulate the heatsink from the host.. But the heatsink was specifically designed to have PERFECT contact to host..

EDIT 2: Success! Another PHR blu ray, where i have to set the range higher.. It's "only" 206mW, but that means like 240mW raw output! Need to measure with an AR coated glass lens!

 

[Zom-B]

Are the stars out of alignment or something? Today is not a good day to test diodes. I've tested two today and three died.

At first I wanted to push the 54mW in my sig, but after desoldering and removing blobs of solder with braid, it was COD, probably overheated. It was in an Aixiz which in turn was in a heat sink too when it died.

This is one I haven't shown before. Originally I thought this was the strongest, but this morning it was on top of the list of weakest diodes I will push. It died without reason and without warning.

This is the one I promised to push further. I redid the whole graph, because temperature differences were making it impossible to continue half-way (see ghost graph line). After the fourth kink (yes I was surprised too) it hinted that it may even be a 5-tier diode, but I stopped as I was happy with it's power.

To further investigate the effect of temperature, I let it cool down for 20 minutes, and then applied 250mA at once. It was 264mW and dropping. I lowered the current until the output power raised (in the middle of the third kink), and that happened at 237mA. It shifted left again, and temperature dependence is proven. I let it on at that current and forgot about it and after 2 minutes I read 000.0mW on the meter. Did I turn it down? no, it's at 237mA. Is it not aimed at the sensor for some reason? I looked at the laser and it was flickering dimly! I looked at the voltmeter and it was rapidly changing between 1V and 3V. It was internally shorted and COD at the same time! I was able to break the internal short by hooking it to a constant voltage power supply with lots of amps, and while increasing to 5V, the current dropped (the short was burnt away) but I did not see any lasing going on and no short flashes. I hooked it to the transistor curve tracer, the V-mA curve looks normal, so the die is normal again. It must be simply COD now.

This means that my 300mW laser might not even last at 250mW.

 

[Zom-B]

EDIT: Damn! The AixiZ module was stuck so hard, that i broke the positive leg of the diode by accident.. :'(
Now i need to insulate the heatsink from the host.. But the heatsink was specifically designed to have PERFECT contact to host..

Use insulating thermal compound, like Arctic Alumina. It's something like $10 for 10 grams and it works like a charm (it's a permanent bond though). I once made a mistake with this compound by gluing an aixiz in a leadlight, expecting to get power from the case, but then it dawned on me that the glue is made to be a perfect isolator and I had to improvise with a spring wire.

You could also send me the bare diode. I've got experience with soldering little wires to broken off pins.

 

[IgorT]

I already fixed it.. See above.. It's doing 206mW.. (Correction.. It's doing >210mW! It hasn't even cooled off yet from me holding it while putting it together! EDIT: It reached 211mW after fully cooling down.)

I have used arctic alumina for a similar purpose before, but it was to drive reds off an AMC... However i could not possibly use it here! I don't want to glue the heatsink into the host.

Luckily, the head is fully anodised inside, and i only had to insulate it from the body, which normally presses it hard into the head, with direct aluminum to aluminum contact, to make the host a heatsink too. That was the only part i had to insulate. A resistance measurement later it was working again. Unfortunatelly this now means, that the heatsinking is not as perfect as i would want. The heatsink is still big, but has indirect thermal pathways into the body, which is more massive still.



In any case, it seems i did not raise the current enough! On a full Li-Ion the current from the FlexDrive still drops into the kinky range..
It's only reliably out of it from a 3V CR123 battery, where the current is about 6mA higher. When the power is just above a kink such a small current change causes a huge power change..

I need to make some drivers...

 

[IgorT]

Zom-B, what hardware do you use to make those graphs?

Does the efficiency calculation take the Vf into account? Did you see any abnormal Vf behavior, or does it always steadily increase with the current?

 

[IgorT]

BTW, i just measured the power of that diode with a 405nm AR coated glass lens (the bad one - v1). It's doing 246mW!

Due to the bad transmittance of the glass in the v1 custom lens, this means somewhere above 256mW raw output... Imagine how much the ones you measured at 300mW after plastics are actually doing!

I wonder how long it would last with the reduced strain on the die....