Im working on a project (still in the 3d modeling phase) where I will have a 25x30mm diameter heatsink, however it will not have any airflow (the entire thing, heat radiating tines and all, will be enclosed in a plastic cylinder not allowing for great heat transfer. As a protection, both for the diode, and not letting my enclosure melt, I was thinking of ways I could prevent it from over heating. The obvious solution was a thermal cut off switch rated for the diodes safe temp and inserted into the copper/aluminum/brass diode module so that it pressed against the walls of it and connect the leads between the power source and the driver board.
LPF Donation via Stripe | LPF Donation - Other Methods
Links below open in new window
ArcticMyst Security by Avery
Anyone here used thermal switches (cutoffs) in their builds?
- Thread starter snek
- Start date
- Joined
- Sep 20, 2013
- Messages
- 17,437
- Points
- 113
Why not add a fan at the back end of the enclosure and leave the front end open to allow air to flow over the fins you claim to have on the copper heat sink? This will work much better than trying to ascertain the diode's temperature by adding a thermistor on the outside of the heat sink. If it is going to be a portable unit like a handheld, it would be better not to enclose the heat sink.
10fenny
0
- Joined
- Sep 10, 2014
- Messages
- 1,054
- Points
- 83
You got room for a tec unit?
Benm
0
- Joined
- Aug 16, 2007
- Messages
- 7,896
- Points
- 113
There are various ways of going about this. If your laser has any sorts of microcontroller control you could use any temperature sensor and then throttle it back or shutting down when approaching maximums.
Simple thermal fuses -may- work, but if you have the type the just goes open circuit at a sudden point and goes back closed circuit when it cools down you could destroy your laser diode: The driver will by trying to push more and more voltage since it cannot reach the constant current set, and when the thermal switch kicks back it the diode will briefly get the full brunt.
Simple thermal fuses -may- work, but if you have the type the just goes open circuit at a sudden point and goes back closed circuit when it cools down you could destroy your laser diode: The driver will by trying to push more and more voltage since it cannot reach the constant current set, and when the thermal switch kicks back it the diode will briefly get the full brunt.
- Joined
- Mar 5, 2015
- Messages
- 1,336
- Points
- 113
Ohhhhh thats cool I will have to look into that
I might have room for a tec but I definitely don't have room for another heatsink to cool the hot side of the tec. Also that would use up more power. I don't plan on running my laser for more than 1 minute at a time... I dont really need any thermal cooling systems however I am trying to make this more or less idiot proof
.
- Joined
- Sep 20, 2013
- Messages
- 17,437
- Points
- 113
What diode are you planning on using and what power are you looking to get? You are going to have a heat sink that is ~1 inch by ~1.2 inches. That is not much heat sink. If it is to be used with anything over 500 mW, you need to make it copper. But, copper is soft and doesn't take fins well at all.
- Joined
- Jul 10, 2015
- Messages
- 9,907
- Points
- 113
As you are building a limited duty cycle device using as large of a heat sink as possible affords you more latitude, a small heat sink can get real hot real fast.
I run the 44 and other nubm diodes in a large heat sink until the sink feels noticeably warm and they last very well.
I run the 44 and other nubm diodes in a large heat sink until the sink feels noticeably warm and they last very well.
JimK
0
- Joined
- Dec 1, 2012
- Messages
- 60
- Points
- 18
This has been used in flashlights for a while, the PCBs over at mtnelectronics have thermal custom cutoffs with open source driver software, but youll probably have to find a way to port over to laser drivers.
Benm
0
- Joined
- Aug 16, 2007
- Messages
- 7,896
- Points
- 113
If you use a microcontroller to run things it gets a lot easier. Many flashlights actually have a microcontroller on board to do things like variable brightness levels, sos/blink modes and all that. This often is the 'unmarked 8 pin chip'.
Obviously there are 'easier' methods to achieve such functions, but given space constrains and the very low cost of small microcontrollers you may as well just put a low end PIC-something in there to do all of it.
With laser diode drivers it can be a bit different: some people want linear control instead of PWM and purpose built circuits do a better job there.
If it's not microprocessor controlled to begin with implementing thermal sensing and throttling or shutdown is more difficult to add - if you have the uC on there it only takes 1 spare input pin and a temperature sensor.
Also, with flashlights it can be a desired feature to have it run in 'turbo' mode briefly, where it operates at a power level the cooling cannot sustain for long, but can absorb due to thermal capacity for several seconds. With lasers you'd mostly want them cooled so you can operate them at full power indefinitely.
Obviously there are 'easier' methods to achieve such functions, but given space constrains and the very low cost of small microcontrollers you may as well just put a low end PIC-something in there to do all of it.
With laser diode drivers it can be a bit different: some people want linear control instead of PWM and purpose built circuits do a better job there.
If it's not microprocessor controlled to begin with implementing thermal sensing and throttling or shutdown is more difficult to add - if you have the uC on there it only takes 1 spare input pin and a temperature sensor.
Also, with flashlights it can be a desired feature to have it run in 'turbo' mode briefly, where it operates at a power level the cooling cannot sustain for long, but can absorb due to thermal capacity for several seconds. With lasers you'd mostly want them cooled so you can operate them at full power indefinitely.
Benm
0
- Joined
- Aug 16, 2007
- Messages
- 7,896
- Points
- 113
That's true, many people operate diodes even over their absolute maximum rated currents, for long periods of time. Doing that does kill a diode once in a while long before the expected lifetime, but then again, that can still take a long time. If you operate a laser diodes under conditions that reduce it's life by a factor of 10, but it was designed to last 10.000 hours (like a dvd or bluray writer) you'd still have to use it an hour a day for 3 years straight - and very few people do that with pointers.
This something odd that people just tend to do with laser diodes: ignore the manufacturers maximum rating and run higher and higher currents to them, often to the point where pushing even more current does not produce any more output power.
Such a reduction of power is usually the result of the laser just getting so hot it's efficiency drops faster than running higher current can compensate for in light output.
Improved heatsinking and thermal protection may help at that point, but everyone should realize when you surpass the absolute maximum rating on any electronic component it could fail right away as well.
This something odd that people just tend to do with laser diodes: ignore the manufacturers maximum rating and run higher and higher currents to them, often to the point where pushing even more current does not produce any more output power.
Such a reduction of power is usually the result of the laser just getting so hot it's efficiency drops faster than running higher current can compensate for in light output.
Improved heatsinking and thermal protection may help at that point, but everyone should realize when you surpass the absolute maximum rating on any electronic component it could fail right away as well.
