High Level Heatsink Design

[thebouljello]

Hi everyone, I haven't been on here much lately. I hope everyone is doing well.

I am currently in the process of making a new laser. It will be high powered, and I will most likely choose the NUBM44 diode for it.

So my main question is, how could I run the diode continuously at a high power (I'm not specifying the power yet because I didn't choose the operating point yet) without damaging it? Would a big enough heatsink work, or would the rate at which the diode releases heat eventually outpace the rate at which the heatsink dissipates heat, regardless of the size of the heatsink?

If there is a way to run the diode continuously with proper heatsinking, what is the best high level approach to take for choosing/designing a heatsink. I have played with some heatsink calculators that are readily available online, but I wanted to get some input on this.

Thanks in advance.

 

[Alaskan]

You would need a fair amount of mass to run that diode continuous, better to use copper and have deep heat sink fins and lots of them, better to have a fan blowing air across them, better to have high velocity air than slow.

I'm not going try to guess on how large or how much air because I haven't worked with that diode in continuous operation, you can however search the net and find systems which do use them that way for cutting and see what they did. Perhaps someone here can help, RedCowboy has a lot of experience with those diodes.

 

[CurtisOliver]

If you want to run that diode continuously in a portable host then heat management will be crucial. A hefty copper heatsink with fins is definitely a way forward. I would look into how other high heat devices are capable of running continuously. CPU’s are a good example. They typically have full contact via thermal paste with a conductive material connected by a heatpipe to a heatsink. What this does is takes the heat from one area to another where it is safe for it to dissipate at a slower rate. Heat pipes are very fast at transferring heat from one place to another. You could move the heat quickly away from the module to an area where you can mount a heatsink away from the diode and driver.

 

[Lifetime17]

Hi,
Well you have all the answers to your question by very knowledgeable members here. More mass better sinking ..
Rich

fullsizeoutput_a14.jpeg

• Lifetime17
• Jan 19, 2019

 

[thebouljello]

Thanks for the input everyone. For the interest of power consumption, I would rather have a larger heatsink without a fan than a smaller heatsink with a fan. Would it be possible to use a larger heatsink without a fan? How about water cooling or using other coolants?

I am also open to the possibility of using PWM to decrease duty cycle and hence, decrease heat emitted from the diode. However, I am not sure how much PWM would decrease the brightness of the laser.

 

[trephanation]

Large with grooves like Lifetime17 posted would be ideal. Ideally it would be made of copper. Get a thicker flashlight host to be able to hold larger batteries and I think you'll be able to make it work.

 

[Lifetime17]

Hi,
Ok coppers the best for sinking , but most expensive also a large Aluminum sink will do quite well for the NUBM44 diodes. I have done many with Aluminum and aluminum with a copper core pressed in. All In all you should keep the duty cycles down respectively ,
we run these higher power diodes way out of what they were originally intended for keep that in mind .
By basic rule of thumb is 30sec on , 30 sec off so the diode doesn't get a licking .
Rich

 

[paul1598419]

You might want to consider a lab host like those used for engraving. They are built to have an unlimited duty cycle and can be found on eBay for less than you might think.

 

[thebouljello]

You might want to consider a lab host like those used for engraving. They are built to have an unlimited duty cycle and can be found on eBay for less than you might think.

What kind of lab host do you mean? I just did a quick ebay search, and there are lots of lab lasers available.

 

[thebouljello]

Hi,
Ok coppers the best for sinking , but most expensive also a large Aluminum sink will do quite well for the NUBM44 diodes. I have done many with Aluminum and aluminum with a copper core pressed in. All In all you should keep the duty cycles down respectively ,
we run these higher power diodes way out of what they were originally intended for keep that in mind .
By basic rule of thumb is 30sec on , 30 sec off so the diode doesn't get a licking .
Rich

It would probably have to be able to dissipate 13W of heat. What kind of heatsinks do you have that could do that without fans?

 

[thebouljello]

Thanks for the replies everyone. Has anyone ever considered using water or liquid coolants to cool their lasers?

 

[paul1598419]

Yes, there are many of them. These are a box design with a fan at the back end and which one you get will depend on the output power and the can size of your diode.

 

[thebouljello]

Found the heatsinks on ebay. I may just go with a smaller heatsink but pulse the diode.

 

[thebouljello]

Fan forced air across the fins of your aluminum heat sink is a simple way to go.........at 4.5a a nubm44 will generate about 15w of waste heat and just a static sink will get too hot if running continuously unless it was unreasonably huge............or you could use a PC type cooler like this: https://www.ebay.com/itm/CPU-Cooler...595665?hash=item5d9bd83251:g:2j0AAOSwh-JeJ~2Z

Due to power consumption and space issues, I will not be able to use a fan. However, I will opt for a lower power output if I can not run the diode continuously at 4.5A. I could probably do 3A to give a power output of about 5.4W. My priority is to have an effective heatsink with no fan, even if it means compromising power output.

 

[RedCowboy]

If this is to be portable as in hand held size then you won't be running continuously on batteries anyway but just how long in minutes do you need to run at once ?

I have used pretty big heat sinks to run for 5 minutes @ 4.5a but @ 4.5 it's putting out about 15w of waste heat and the warmth is palpable after 5 min. and that's quite a large sink and using 26650 cells.

----edit----

At 3.0a you will only get 4w output and 9w of waste heat.
At 4.5a you will get 7w output and 15w of waste heat.

That's CW with a G2 lens. Pulsed you can no doubt run longer.

https://imagizer-cv.imageshack.us/v2/xq90/923/3B5tYd.png

 

[thebouljello]

If this is to be portable as in hand held size then you won't be running continuously on batteries anyway but just how long in minutes do you need to run at once ?

I have used pretty big heat sinks to run for 5 minutes @ 4.5a but @ 4.5 it's putting out about 15w of waste heat and the warmth is palpable after 5 min. and that's quite a large sink and using 26650 cells.

----edit----

At 3.0a you will only get 4w output and 9w of waste heat.
At 4.5a you will get 7w output and 15w of waste heat.

That's CW with a G2 lens. Pulsed you can no doubt run longer.

https://imagizer-cv.imageshack.us/v2/xq90/923/3B5tYd.png

I am not quite sure how many minutes I will be running it. I want to achieve continuous running because the length of the cooldown time at the end of the cycle is not very defined, so I do not know how long I will actually have to wait. So, I would rather be able to run it continuously than have to wait for an undefined cooldown time.

I am currently studying engineering and am willing to calculate how many Watts a heatsink could dissipate and ensure that the wattage of the waste heat does not exceed the wattage of the heatsink.