Blue laser 74W NUBM35 Portabl?

[RedCowboy]

I like your magnetic lens attachment, I mounted different lenses to cover plates that fit my enclosure so it takes me 4 screws to change out my 1st lens, but your magnets are faster.

 

[Borislav@87]

These are phone magnets if anyone is interested. They are strong enough to hold the lens firmly. It's quite a distance from the diode, so they won't harm it in any way.

 

[Borislav@87]

So after about 10 days of use, one diode died. I haven't changed anything on the laser. It operates at 56.7V and 4A current. I do not exceed 50 degrees in use. Space suddenly as I was using it the laser went off (BMS protection) Then I turned it on and saw that one diode died. I don't know where the reason is. These NUBM35 arrays are probably very buggy. I then checked the amperage to see if it had changed, but the meter shows 4A again as if that diode was working and nothing had changed.

 

[RedCowboy]

Time will tell if any more give out.

Except for the one dead diode, both my nubm35's are still working just fine.

 

[Borislav@87]

It's unpleasant, but I hope it's just this one. I don't believe the problem is that they work on 4A. As I said, I don't exceed 50-55 degrees. Apparently these NUBM35 arrays are not that successful and are very easily damaged.

 

[RedCowboy]

Not necessarily, they are close together, so waste heat build up could be more of an issue, they may be more easily damaged by overdriving than the single nubm 0E,0A,08,44 by themselves. Also sometimes a diode will simply fail early, 20K hours is the average time to failure but there's also a certain number that do fail early, time will tell if it's a lone issue or if something else is at work.

 

[Gryphon]

Looks like not all LM338K's are created equal. I ordered some off of amazon to play around driving a string of seven A140 diodes from a 9S lipo set and it worked!......for about 5 secs. I thought I might have popped one of the diodes but I soldered in another LM338K and got another 5secs of light before it too burned out. I was using one 1ohm 2W resistor to keep the current in the 1.2A range with the regulator setup as a current regulator.

After popping a second regulator I decided to decan the first one and see what was inside.

Looks like the bond wire gave up the ghost, I have no doubt the other one failed in a similar manner. Anyone know what the inside on a "good" LM338K looks like? I imagine it wouldn't be as inept as the ones I've got.

 

[RedCowboy]

I used the caveman de-canning method ( squeezed the lid with channel locks until it burst near the base ), looks a little different inside, but not much.

 

[Gryphon]

I'd say its better than the ones I've got, there are more bond wires and they look a good deal thicker too, and the chip has got a double pedestal probably for better heat dispersion. I ordered some National Semiconductor brand LM338K's from ebay as replacements. I initially tried to get some Texas instrument ones from DigiKey but they are selling them by the tray for $83. I'm not sure how many come on a tray but it will certainly be more than I need.

 

[Gryphon]

The new LM338K's came in and they are performing much better, driving all 7 diodes now with no problem at all.

Are you using any capacitors for startup spike protection? I know modern 445's are pretty robust I'd figure I would ask.

 

[RedCowboy]

Nope, no caps at all and no troubles so far.

 

[Borislav@87]

I decided to improve the cooling of the laser using a Peltier cell. I will glue the Peltier to the end of the aluminum heatsink. On the warm part I will stick the same finned radiator again. Thus, the fan will be cooled by the Peltier, and its cold part will help to cool the large radiator. I will connect the Peltier to the converter that the fan is connected to because it needs 12V and 3-4 amps

 

[WizardG]

I decided to improve the cooling of the laser using a Peltier cell. I will glue the Peltier to the end of the aluminum heatsink. On the warm part I will stick the same finned radiator again. Thus, the fan will be cooled by the Peltier, and its cold part will help to cool the large radiator. I will connect the Peltier to the converter that the fan is connected to because it needs 12V and 3-4 amps

And how many watts is that peltier cooler rated for?

 

[Borislav@87]

You can view them here. I will install a weak 12704, since I also ordered 12709.

https://a.aliexpress.com/_mtcff7w

Description :

Get ice cold in minutes or heat to boiling by simply reversing the polarity, used for numerous applications from CPU coolers to alternate power sources, or even for your own
custom car drink warmer/cooler.
Since they consit primarily of semiconductor material sandwiched between ceramic plates and have no moving parts
These devices must be used in conjunction with a heat sink to avoid burned
Each device is full inspected and tested

 

[WizardG]

You can view them here. I will install a weak 12704, since I also ordered 12709.

https://a.aliexpress.com/_mtcff7w

Description :

Get ice cold in minutes or heat to boiling by simply reversing the polarity, used for numerous applications from CPU coolers to alternate power sources, or even for your own
custom car drink warmer/cooler.
Since they consit primarily of semiconductor material sandwiched between ceramic plates and have no moving parts
These devices must be used in conjunction with a heat sink to avoid burned
Each device is full inspected and tested

All right. With the 12709 rated for 86 watts that should definitely help a little. But that 86 watt figure presumes the 'hot' side of the thing is held at a constant 25C. YMMV

 

[Borislav@87]

I can not tell you. I am yet to experiment. I will mainly use it to remove the collected temperature from the aluminum heatsink, because in my case it is quite large and takes more time to cool down. That will help for sure