Blue laser 74W NUBM35 Portabl?

[Borislav@87]

I have one question. I expect to receive the laser array tomorrow. I looked at other threads on the forum. In many places it is mentioned that it is best to connect each row of diodes in parallel. In my case, if I connect each line in parallel to one driver, will it be better?

 

[RedCowboy]

It's better to drive both rows in series or each row with it's own driver, but as long as their all matched diodes then you can likely get away with putting the two rows in parallel, that means if you want to drive each diode @ 3.5A you will need to set your driver for 7A but the row voltage will be the same for both rows in parallel as for 1 row, about 31.5V ( 7 x 4.5V )

So what's happening is each row should be drawing 3.5A of the 7A....... however if one row has a lower resistance then it could draw more of the 7A and the other row less.

I have a couple of the nubm35's and one of then arrived with a dead diode and that row is a little weaker as well, so I would be careful about putting the rows in parallel, definitely not more than 3.5A each 7A for both and you are taking a risk if you do.

 

[Borislav@87]

Thanks, I got it. I will connect them in series. This driver that I use from 1800w will easily handle all 14 diodes. But I still wonder what amperage to set 3.5A or 4A. I don't know how durable these arrays are. I guess 4A current won't be a problem for him? You had previously shared that they withstand up to 5A without problems, or at least that's how I remember it

 

[RedCowboy]

The individual chips can take 5A I expect and if they were each on a 9mm backing plate like NUBM0A, NUBM0F, NUBM08 then you can press into a copper module and the heat build up would be slow, because there is a heat bottleneck at each diodes substrate stack the output slowly falls when driving above 3.5A

When driving @ 4.5A you will see good output, but without active cooling the power slowly ticks down as heat builds up, now @ 3.5A they will output stable as long as the heat sink holds up or with fins and a fan, but as close as the chips are together in these arrays, you will need a TEC pad and a finned heat sink with fan forced air underneath it to draw away the heat if you wanted to push to 5A for more than .... ???? 20 seconds, 30 seconds, ?? seconds......... I was running 1 row @ 4.5A for a while but only for 15-30 seconds at a time, I wouldn't run it @ 5A, you're just making a lot of waste heat in a small spot, maybe 4A and use some thermal compound between the array and your aluminum or copper heat sink.

p.s. Also your heat sink should be flat and smooth and screw the module down tight with a thin layer of thermal compound between the module and heat sink, you want to move that waste heat away or you will melt the gold bond wires and indium mountings and/or burn out your diodes.

 

[Borislav@87]

I don't want to play 5A. I plan to set it to 4A and use it that way. The array will be firmly pressed against the radiator. I have put a second fan to help and an additional aluminum radiator. I think cooling will do the trick. And just in case, there is a thermoswitch at 70C. So it can't overheat. If by chance it gets very hot and exceeds 70C, the circuit will break. I'll post tomorrow when I do.

 

[RedCowboy]

Thermal paste/grease is inexpensive and it makes a difference, it only takes a small amount, remember you are over-driving 14 laser diodes in a small space so the better you move heat, the more you can get away with pushing the lasers.

 

[Borislav@87]

Trouble again! I picked up the laser array today and wired everything up. It turned out that only 6 laser diodes out of 14 were working. Already at the beginning I saw that the number of beams was not the same. I don't know why I came across a defective item again. I am very angry because I wait a whole month and finally I get a defective array.

 

[farbe2]

Its more likely that you shitty driver is killing the arrays and/or your thermal management is complete garbage.

Sorry but mounting a 150$ 75W laser array with (what looks like) a wood screw is certainly a good way to kill it.

Also: you know about ESD right?

This thing is putting out more than 150W of waste heat at 4A drive current.
Even in this bad picture, you can see a gap between the module and your WAY WAY to small heatsink. (does not matter that you got two fans on it, its. too small)
Attache the array to a sufficiently sized heatsink with all mounting holes and use a proper thermal grease.

The placement of the thermistor also does not make sense. It wouldn't detect a overheating array as its placed on the coolest part of the heatsink with the worst thermal conductor known to man: playdough
Calculate the thermal resistance from your heatsink (or measure it), make sure its sufficiently sized, than place the temp sensor on the hottest part with proper thermal interface material.

The driver is most likely kicking out spikes that reach up to the sun. Which you wouldn't be able to detect without a proper oscilloscope.

Use something proper or use basic electric knowledge to engineer something. Like using a voltage regulated supply with a resistor in series and a cap over the array. That gives high losses but at least keeps any spikes away from the array.

I also hope that your complete project does not catch fire. The wires you used are not fused and way to thin for the current that your battery can put out. Especially the small wiring that goes to your voltage display.

All the solder joints on your battery pack look very cold, even the joints on your BMS look cold.
This will cause waste heat and voltage drop and could lead to a cell temperature run away.
Please at least isolate the BMS board from your cells. The shrink wrap is very sensitiv and easy to damage which could cause a direct short -> fire.

 

[Borislav@87]

The laser array is pressed hard enough to the radiator and has thermal paste. I haven't run 4A as seen in the video, it barely burns cardboard. When I first tested it, they didn't work. It's certainly not the problem with the radiator or whatever, because it hasn't had time to warm up. If it was the driver it would kill all 14 diodes. I don't know why he's stuck like that either. If it was my fault, I wouldn't be so angry.

 

[farbe2]

If it was the driver it would kill all 14 diodes.

Depending on the nature of the spike, only a few lds die.
Just because all are in series and therefore get the same current does not mean that the die at the same current.

Even running at 2A will produce enough waste heat (around 100W) to heat the array up very very fast.

What did you do in therms of ESD protection?

 

[RedCowboy]

Farbe I haven't scoped these drivers yet, but they do have caps on board and I haven't had any trouble with mine yet, yes I know that's not a proper test and is no guarantee, that said techhood in China has sent me defective stuff before and I test my laser diodes and arrays with a quality bench top power supply when they 1st come in, so I know of the 2 nubm35 arrays I bought from techhood that one had a dead diode and a weak row when I got it, the other is just fine and has held up well.

---------

Boris I can see a gap under your array and that's a gob of compound, you only need to use a small amount spread out in a thin layer and you need to make firm contact with your heat sink.

p.s. Don't test the array without a heat sink when you 1st get it, there is very little material to absorb a lot of waste heat.

 

[Borislav@87]

Farbe I haven't scoped these drivers yet, but they do have caps on board and I haven't had any trouble with mine yet, yes I know that's not a proper test and is no guarantee, that said techhood in China has sent me defective stuff before and I test my laser diodes and arrays with a quality bench top power supply when they 1st come in, so I know of the 2 nubm35 arrays I bought from techhood that one had a dead diode and a weak row when I got it, the other is just fine and has held up well.

---------

Boris I can see a gap under your array and that's a gob of compound, you only need to use a small amount spread out in a thin layer and you need to make firm contact with your heat sink.

p.s. Don't test the array without a heat sink when you 1st get it, there is very little material to absorb a lot of waste heat.

I'm sure I've come across a defective item! The first one had broken pins and this one had non-working diodes. Massiva has a very tight fit. There is no gap, it's thermal paste and it's lying to you. At the moment, 6 diodes work at 4A without problems. The temperature is about 50C for 60 seconds. During the initial test, the array worked at 1.2A and did not heat up at all and could not have been damaged by the temperature. You have also come across a defective item, but with a smaller defect. I do not know what to do. Should I buy a new one or not risk it? Also I see there is a NUBM38 I guess it will be a new model. It is also very likely that these arrays have been running for a long time and are not new

 

[RedCowboy]

You need 4 screws, if you don't want to drill the lip of the array or use small screws you can pinch all 4 edges, but all 4 edges need to be pressed tight with a thin layer of compound.
This is ugly but I made good contact with a thin layer of compound.

 

[Borislav@87]

Here are which diodes are live

 

[RedCowboy]

Did you buy your array from techhood ?

 

[Optixbeam]

Did you buy your array from techhood ?

I bought the array (nubm35) from techhood and was also planning on doing a portable build but I didnt get to set it up yet…

This thread kinda has me nervous about techhood but I trust him