How did my diode die?

[nocturnalBadger]

I have an M140 (A type) as part of a homemade CNC build. It is now shining dimly, which I’ve read is a sign that its lasing days are over. I’d like to detail my whole setup and the events leading up to its failure in an effort to find out how to not kill any more diodes.

Here’s my setup:

• M140 type A diode from DTR (copper module, standard glass lens)
• Heatsink is a simple aluminum mount from Aixiz. The driver also has an overheat protection freature with a thermal sensor that I affixed to the diode heatsink
• Driver: Simpledrive I from X-Wossee. I got it because I wanted adjustable current output and PWM input (also it was recommended on DTR’s CNC page). I’ve been very happy with it overall.
• Test load: Homemade string of 6 1N5408 diodes (0.8v drop from each totaling a voltage drop of 4.8v to compare to an expected drop from the M140 of 4.7v)
• Input power supply: an old laptop power supply: 2A @ 12V
• PWM source for driver activation: 4.8v from arduino

Here’s what happened:

I initially set the driver’s gain current using my test load to 0.8A. After about a day of happy lasing, I realized I was demanding a lot of power dissipation from the driver (24W in, 3.4W out), so I again used the test load to set the driver to 1.45A. Still no problem: it worked for several hours after this. During this time, I was tweaking the software for the CNC machine so I was switching the driver power on and off as I test gcode commands. I don’t believe this was was a rough duty cycle (probably 30s on to 1min off); the heatsink didn’t feel warm at all. At some point, without any incident that I can recall, I switched on the diode to see it shining very dimly.

I’m baffled by how sudden this was; I can’t think of any event that would have caused it to fail. If you can provide any insight into ways a diode can die and what I can do to make that not happen in the future, I would really appreciate it. If you want more pictures or info, I’ll be happy to provide it.

Here is an album with pictures of my setup as well as before and after photos of the laser:
Click Here!

 

[Cyparagon]

I realized I was demanding a lot of power dissipation from the driver (24W in, 3.4W out)

That's not true. The power supply may have been rated for 2A, but you clearly said you were drawing 0.8A from it. This would mean 9.6W in.

Did the driver have adequate heat sinking? If so, it sounds like you're doing everything properly. The diode and/or driver just failed. Just as light bulbs can spontaneously burn out after a few days, It happens some times.

 

[nocturnalBadger]

Yes, I think the heatsink on the driver is sufficient (you can see it on the left of the last picture I showed).

I guess I misunderstood how the driver works. I thought that it would take the input power (2A @ 12V) and dissipate the excess energy as heat to reach the target current (0.8 and later 1.5A). Or maybe I'm just misusing the term 'gain'; I meant that I adjusted the pot labeled 'gain' until the output read 0.8A.

The driver btw is working after the failure. It's still outputting a constant 1.45A to the test load.

If it is a random failure, that's quite disappointing since it was only working a day. I can't afford to pay $2.50 an hour to have a laser!

 

[RedCowboy]

I would upgrade to the NDB7875, they are tough and have anti static built in, it's basically the same size emitter/spot but more power, better heat transfer, and more efficient.

I run mine at 2.4 amps and get about 3 watts out of a G2, if you set yours at 1.45 it should last a very long time.

I had a brand new Osram 1.6w go dim/led after a very short time, for no apparent reason it just quit.

7875's are the most robust diodes I have seen and I beat on them something awful, the price is not much more than the 140.

https://sites.google.com/site/dtrlpf/home/diodes/9mm-445nm

 

[paul1598419]

You called it an M140, but said it was an "A" type. These are commonly called an A140. They are usually an older version of the M140 and can't take the current the M140 can. Many of the "A"s have hours on them already, though I can't say that is the case with yours.

 

[nocturnalBadger]

Yeah, I'll admit I got the A just because it was cheaper. I expected the only downside would be less output but I guess just being an older model could mean it's more prone to failure, idk.

 

[RedCowboy]

OH, you ran an A-130 at 1.45a...Well there's your problem.

Get a new 7875 from DTR and you will have no problems and a lot more power, other sellers sometimes sell well used and mislabeled parts, DTR's diodes are brand new and as advertised.

 

[paul1598419]

Yeah, you really aren't saving money when you use an A140 instead of the M140. They aren't worth the money, IMO.

 

[nocturnalBadger]

Ok, just to clarify, this is the diode. It was advertised by DTR as an "A-Type M140 Blue Laser Diode". On the page is says "The A-Type M140's average about 1.8W @ 1.8A using a G-1/2 lens."

 

[Isaac Clarke]

this post is becoming confusing...
-"anything that can go wrong, will go wrong".
-you suspected there's something wrong with the diode and driver... so that's the root of the problem.... fastest and easiest solution IS TO BUY A NEW Module NDB7875 that come with the Adjustable 2.4 Amp X-drive combo..... and stick to the specified input voltage and current...

 

[paul1598419]

You are an LPF member. It costs DTR more to sell on eBay than on his home site. You can have an M140 laser diode for $37.00 at this site: http://sites.google.com/site/dtrlpf/home. If you read the whole listing of the page you'd have seen the difference between the two diodes.

 

[Benm]

I think it is/was what is commonly referred to as A140, i.e. the first generation of such diodes harvested from crapio projectors. They can handle less output power (and hence input current) compared to the more recent diodes in similar projectors.

Laser diodes usually die very quickly often with little or no warning signs. We are often pushing diodes well beyond their rated maximums, so you can expect them to fail at some point.

Though these diodes are very similar, you can tell them apart mostly by looking at:

- number of bond wires between die and pins: should be 3 each for the A140, 4 each for the M140

- The M140 usually has a circle next to the 'QR code' printed next to the pins

Note that A140 and M140 are NOT part numbers for the laser diodes, they simply refer to the projector they are extracted from. In theory you could find a projector labeled A140 where some or all of the diodes inside are of the M140 type, or less likely, the reverse. In the projector the diodes are not driven THAT hard so they would not fail so easily.

 

[RedCowboy]

You have a lot more latitude with the 7875, it's emitter is the same size as the M-140 but it can take more power and put out more power, it's a little more efficient, it has better heat transfer as it's a 9mm diode and your feed rate can be higher, also your distance is less critical, mostly you will want to not set your work piece on fire, but a stream of air can help with that.

It doesn't benefit me either way, just sharing my experience, the 7875 is the boss and worth a couple extra bucks, at 1.45 amps it will likely last for thousands of hours as it's a 20,000 hour diode at 1.6 or 1.8 amps I think.

Here's a quick clip playing with a 1.35mm thick popsicle stick.

 

[Pman]

Good advise here. For what you are doing I would pay a bit more like Red suggests so you have much more leeway.
I've built a ton of lasers and I can truly say that there's a reason they are called DIEodes My bet is everyone here who builds a lot will agree that you just never know when a diode is going to go on you. Everytime I am about to turn on a laser I wonder if today will be the day the diode blows although most of my issues have admittedly been from pushing them too hard but that's part of the fun. In your case you want a nice even stable unit so you are far better off driving a more powerful diode with less output to greatly reduce the possibility of it happening again.

 

[Benm]

They go 'poof' on all of us once in a while, that just seems to be a fact of life, and is to be expected when pushing these things above absolute maximum ratings.

Then again technology developes, and you could just replace that A140 diode with an M140 and leave everything else as-is, expecting a much longer service life for the replacement, as well as a bit more output power for the same current.

Using a 9mm diode will have some advantages, but the big disadvantage is that you need to get a new module to contain it. So i'd go for the M140 drop-in replacement in this case.

 

[RedCowboy]

If his mount http://www.amazon.com/gp/product/B00GGX4MDQ?psc=1&redirect=true&ref_=oh_aui_detailpage_o07_s00 holds a 12mm module and he runs a 7875 at 1.45a it should fit his set up and last a long time, he would actually be driving it under spec.

He can't press a 9mm into a 5.6mm module socket, but a full copper 12mm module that fits a 9mm diode is just what 5 dollars?

Why not just get a 7875 already pressed into a full copper 12mm module from DTR?

--------------------------------------------------------------------------------------------
p.s. How do you do the hypertext link deal where you have "Heatsink" typed that goes to the link?
It's up top in the tools, right?