Properties of lower vs higher efficiency diodes.

[Dusty_Lenses]

Got a question for you folks. I can't find any information on it but I do recall reading somewhere that the lower efficiency blue diodes tend to have higher wavelengths. Now in relation to the nubm 08, is this true and what properties do the higher efficiency diodes present, if any difference?


Or did I dream the whole thing?

 

[RedCowboy]

Generally speaking the earlier phosphor pumping blues were 445nm, the 455nm came later..... they also became more efficient later.
That said there is no hard and fast rule.

 

[Dusty_Lenses]

Generally speaking the earlier phosphor pumping blues were 445nm, the 455nm came later..... they also became more efficient later.
That said there is no hard and fast rule.

I was referring to a scenario where say a block of diodes when tested individually at a given current the lowest efficiency diode in that batch would present a higher wavelength than the others in that same batch.

I'm going to have another look with the search, or perhaps I'll try it out on Google as the search here on the forum tends to omit terms that are deemed too short / too common.

 

[Dusty_Lenses]

Found it through Google search.

"This is in agreement with what DTR and other members said earlier that the less efficient LDs usually have higher wavelengths."

L

Thread 'Test of diodes in NUBM08 block one by one'

May 5, 2018

From many tests with single laser diodes discussed here we know that the LDs with G-ball of even the same type can slightly (or not) differ in power, wavelength and quality of collimation by attached G-ball.

Since 2 years ago when the 8LD blocks have appeared on ebay I wanted to know if the LDs inside one block would be similar or not by all or some of these properties. But as far as I know no one has checked this and posted on LPF before.

So recently I have received a block of NUBM08 from a Chinese colleague who bought in taobao (kind of Chinese ebay) for pretty cheaper than prices in...

• Light superglue
• Replies: 24
• Forum: Reviews

• 

So this is the 08 block that I was testing. Picked this up on eBay and was a block of clean unsoldered pins.

I selected the one which outputs the lowest and mounted it up. Im going to run it side by side with the one that outputs the highest at the same current and see what differences I can spot in the beams.

 

[Darktron]

So many things can affect a LPM reading. And extracting diodes from a block/bank is a process that can break the diodes easily. Not many people have the patience and tools to extract diodes properly without overheating them or dropping them to hard.

I test my diodes at minimum threshold 500mA or so and read LPM without added lens if is gball then read off from that and if is flat window read off from that. Test all 6-10 diodes. Whichever gives highest LPM at minimum lasing threshold then thats a good or sometimes a gold sample off the batch. The rest usully aren't too far off. Something I do need is an accurate spectrometer.

 

[Dusty_Lenses]

So many things can affect a LPM reading. And extracting diodes from a block/bank is a process that can break the diodes easily. Not many people have the patience and tools to extract diodes properly without overheating them or dropping them to hard.

I test my diodes at minimum threshold 500mA or so and read LPM without added lens if is gball then read off from that and if is flat window read off from that. Test all 6-10 diodes. Whichever gives highest LPM at minimum lasing threshold then thats a good or sometimes a gold sample off the batch. The rest usully aren't too far off. Something I do need is an accurate spectrometer.

What does that tell you about that one diode that reads the highest at threshold?

I have not tried this approach. Typically I go by which diode it is and if there's a data sheet available I set my current limit to its maximum rated current and mark down what output they all have at that current testing one at a time and allowing the block to cool back down to room temperature in between each test.

Regarding extraction, I test them all and mark down the outputs, then following a gentle extraction using a rework station I mount them up and test them again individually with no additional optics.

My biggest fear with these soldered blocks is damaging them with heat. So far I have not seen any losses post extraction.

 

[Darktron]

So far I have not seen any losses post extraction.

Then you're doing it right.

What does that tell you about that one diode that reads the highest at threshold?

Tells me that this diode can potentially hit a higher output than the rest. Essentially more efficient in the batch.

 

[Dusty_Lenses]

Then you're doing it right.

Tells me that this diode can potentially hit a higher output than the rest. Essentially more efficient in the batch.

Not picking at your methods or anything, but why not just run it at its rated current? Then there's no question what power it can reach. You also get to see if there's any anomalies in the wavelength when it shifts, if it shifts etc.

I noticed with some diodes when you run them at threshold and they start to output up until about an amp the blue is much deeper and as you approach it's rated maximum current input if the diode is going to shift wavelength that's where you see it.

I also need to build a spectrometer, it's quite difficult to determine wavelength and differences because they can be so subtle even when testing the entire block.

One thing is for certain these newer diodes from nichia are a lot more fun to test than the diodes of yesterday

 

[Darktron]

Not picking at your methods or anything, but why not just run it at its rated current? Then there's no question what power it can reach. You also get to see if there's any anomalies in the wavelength when it shifts, if it shifts etc.

I noticed with some diodes when you run them at threshold and they start to output up until about an amp the blue is much deeper and as you approach it's rated maximum current input if the diode is going to shift wavelength that's where you see it.

I also need to build a spectrometer, it's quite difficult to determine wavelength and differences because they can be so subtle even when testing the entire block.

One thing is for certain these newer diodes from nichia are a lot more fun to test than the diodes of yesterday

When you test so many diodes running them at full rated power means you'd need to heatsink them and wait for temperatures to stabilize while a threshold just bypass most of that . But yeah id get a selected few 4 off a block and test those further after i know they are the better of the bank.