Failure analysis help for Sharp 700mW 638nm

[Philipnzw]

Hey folks, i just recently LED-ed a sharp 700mW 638nm diode.

Its quite odd to me since it has been functioning well for 1 week, without any issues. I ran it at 650mA using an AMC7135 based LED driver.
It just died all of a sudden.

After a bit of checking, it seems like there could be 2 reasons.

Reason 1: I did not sit the diode far enough into the module. Since i did not have my vice with me when i constructed this laser, i just used a hammer and diode push tool to gently tap the diode into the module. I think i could have been too gentle, to the point where the front end of the diode base did not touch the module.

Evidence to support reason 1: Low waste heat. Through the 1 week of its life, i never experienced much waste heat from the laser diode at all. It didn't even heat up much after running for 3 minutes. This of course could be the fact that there was inadequate thermal contact between the diode and the module.

Reason 2: The cheapo LED driver killed the diode.

Evidence to support reason 2: There are no capacitors on the driver. This means that any ripple in the current will be sent to the poor little diode, damaging it in the process.

At the same running current, does an LED-ed diode still output the same amount of waste heat as a non-damaged diode?
Is the sharp 700mW 638nm more sensitive and delicate to current spikes compared to the mits 500?

Folks, whats your opinion on this? Personally, i think it could be a mixture of both reasons. However, i would also like to listen to all of your experiences. It sucks to kill diodes and i have already killed a mits500 previously (in my infinite wisdom, i pulled a rookie mistake and messed up the polarities) and i would like to prevent future diode deaths.

 

[Snecho]

Is this the canned or open-can one? I can't remember at the moment.

 

[Philipnzw]

Is this the canned or open-can one? I can't remember at the moment.

Its a canned diode.

 

[Snecho]

Its a canned diode.

Ah I see. I was thinking of the Mitts decanned diode, those are hyper sensitive.

 

[RedCowboy]

I will put mt bet on :
Reason 2: The cheapo LED driver killed the diode.

 

[Philipnzw]

Ah I see. I was thinking of the Mitts decanned diode. Those are hyper sensitive.

Ah i see, that's a bummer, there goes the set of 3 RGB sharp diode builds that i wanted to do. I'm thinking i should just switch back to the mits500.

I will put mt bet on :
Reason 2: The cheapo LED driver killed the diode.

Will a mits500 fair better against such a driver? i have saw many posts here about running the mits500 with an amc7135 based driver.
I'm starting to think that i should just suck it up and get a proper driver instead, though it has been quite fun messing around with cheap chinese drivers and trying to make things work.

 

[RedCowboy]

AMC7135 is a linear isn't it ? I wouldn't think it would kill your LD, I suppose you could add a cap across the output.........ok here's a data sheet.> HERE

I found an older thread here > HERE

 

[Giannis_TDM]

Ah i see, that's a bummer, there goes the set of 3 RGB sharp diode builds that i wanted to do. I'm thinking i should just switch back to the mits500.



Will a mits500 fair better against such a driver? i have saw many posts here about running the mits500 with an amc7135 based driver.
I'm starting to think that i should just suck it up and get a proper driver instead, though it has been quite fun messing around with cheap chinese drivers and trying to make things work.

I bet driver issue, I have the same diode and it doesn't produce much waste heat either, let me remind you that the red diodes are way more efficient because of their high WL. Don't use cheap drivers or at least test them to make sure they regulate properly. Personally I am using an ACS500SE with my red.

 

[Philipnzw]

Thanks for the replies folks. Turns out, my mits500 is still alive despite me wiring it in reverse! I think i will attempt to add a capacitor across the output like others once did. Lets see if that does the trick and if the laser will survive more than a week.

Edit: I did just that, it worked and was a couple hours of fun at night. I went to bed and woke up with a mits500 giving me almost LED-ed like brightness. All of this was even with the added capacitor across the diode. At this point imma just save up for a proper driver, f**k this shit.

 

[Snecho]

Ah i see, that's a bummer, there goes the set of 3 RGB sharp diode builds that i wanted to do. I'm thinking i should just switch back to the mits500.

I don't understand. I was just saying I thought you were using an open can diode, which are super sensitive.

 

[Philipnzw]

I don't understand. I was just saying I thought you were using an open can diode, which are super sensitive.

Wait...aren't the mits500 known as one of the more durable diodes?

 

[Snecho]

Wait...aren't the mits500 known as one of the more durable diodes?

I can't recall at the moment. My point was just that you get dust in an open can diode, you're donzo

 

[aaronnoraa]

I can't recall at the moment. My point was just that you get dust in an open can diode, you're donzo

Maybe in the already doomed ~450nm "open can" diodes but for whatever reason the Mits diodes seem to be ok running inside projectors in dusty, foggy and otherwise non ideal operating environments and seem to hold up. I'm sure that a particle of dust directly on the output junction wouldn't be great but for whatever reason that doesn't seem to happen often in practice.

 

[hakzaw1]

great thread--TY all for the info.

+rep when sys allows.