TEC - cooled 5mW->75mW pointer 'Mod', Awesome Results!!!!

dr_lava

Enlightened
Joined
Feb 8, 2006
Messages
412
I thought I'd share a little project I've been putting together today with some found parts around my room. I have wanted a continuous-on laser of high power for various purposes, and didn't want to spend a few hunderd (or thousand) for it.. plus, I wanted to have some fun.

I took the laser out of a old version luckyduck 110 (their new lasers are no longer the same quality edit 2/2007), modded it to continuous current mode, and did the various 'stages' to the optics for optimal output and thermal conduction from the diode and crystal. Then stuck a peltier (thermoelectric cooler) on a heatsink and made a aluminum bracket for the laser. The backet is held down with nylon screws to the heatsink so heat isn't transferred through them from the hot to cold side.

I then connected the laser to a li-ion battery and the TEC to a power supply and started cranking up the diode current. at about 350 mA, it was about as bright as my 40 mW atlasnova pointer, which can barely make dents in electrical tape. Now at 400mA, it cleanly and quickly slices electrical tape and is quite a bit brighter than the 40mW.

I wonder how much the diode can take? I'm going to fix an LED in front of it and look for power variations that might hint at diode damage. So far, I'm keeping the cold side of the TEC at about 15 degrees to avoid condensation.

Soon I'll post mW readings.

What do you think?

laser_tec.jpg


:rock:
 
Last edited:

Athoul

Enlightened
Joined
Oct 5, 2005
Messages
391
Re: TEC - cooled 5mW pointer, just for fun.

Wow, what a great project you have going there. :)

Now you have a tabletop DPSS laser unit :D You did a great job with the heat distribution though. I bet you will be able to crank it up quite a bit beyond the normal limits and look forward to seeing your results.
 

shawndoe

Newly Enlightened
Joined
Nov 9, 2005
Messages
25
Re: TEC - cooled 5mW pointer, just for fun.

Hi,

That is to cool. I like the addition, simple and easy to implement. Have you tried this with other modules? I seem to remember that crystal efficency is related to temp. Warmer then room temp was better, but I may be wrong on this.

Also, the wavelength of the diode changes with temp. Maybe cooling the diode is shifting it's frequency. Try setting a steady current then slowly increase cooling and look for a peak in efficency.

Of course your know all this, did I miss anything?

Great experiment.
Shawndoe
 

tobjectpascal

Banned
Joined
Feb 19, 2006
Messages
177
Re: TEC - cooled 5mW pointer, just for fun.

it's a shame peltier devices use so much power, the small one i have uses 1.8amps x 12 volts, if they made them small enough and get the draw to around 300 - 500ma at say 3 - 6 volts, it might be possible to make a portable really high powerfull laser.
 

dr_lava

Enlightened
Joined
Feb 8, 2006
Messages
412
Re: TEC - cooled 5mW pointer, just for fun.

You're right, shawndoe about changing the cooling and looking for a peak efficiency. I'll do that eventually.
The efficiency is related to the wavelength match between the diode and crystal, which is one of the reasons different crystals have different optimal temperatures.. The Nd:YVO4 is more efficient cooled and the KTP is more efficient warm, from what I have read, but I'm not going into a triple-temperature controlled laser right now!

I measured with my green diode and it's a steady 440uA right now. From previous known power tests this translates to about 60-70mW. Not bad for an 'average' laser. 'Average' lasers under normal circumstances only output 15-20 mW when tweaked.

tobject - absolutely, however, I only need to draw about 1 Amp with this peltier to maintain temp. The problem with portables is that a heat dump is still needed; with a TEC, it needs to dump the laser heat AND the TEC power now.

caught_green.jpg


It's that good stinging feeling :)
 
Last edited:

pixar

Newly Enlightened
Joined
Apr 28, 2006
Messages
88
Re: TEC - cooled 5mW pointer, just for fun.

440ua should indicate apporx 150mw +/- %5-10, so why the 60-70mw?
 

dr_lava

Enlightened
Joined
Feb 8, 2006
Messages
412
Re: TEC - cooled 5mW pointer, just for fun.

pixar said:
440ua should indicate apporx 150mw +/- %5-10, so why the 60-70mw?

Sorry, that is a widely held misconception. Please help in spreading the truth; read this!

http://argonholo.webhop.net/electronics/greenLEDs.html


His LEDs put out anywhere from 3.6 to 11.5 uA/mW. Mine is about 7.5uA/mW

my measurements ranged from 440 to 485 uA, depending on TEC temperature.
 
Last edited:

RickSCV

Newly Enlightened
Joined
Jun 13, 2006
Messages
11
Re: TEC - cooled 5mW pointer, just for fun.

I've seen a half amp pelt. device somewhere off one of the taiwaniese electroics websites... I have no idea though how they operate or what results they give, but a little cooling is better than no cooling. I have an old one these attached to a socket 7 cpu heatsink. I actually had though of trying this same project with my new newwish laser Im about to scrap since the casing was nearly destoryed getting into it :)
 

dr_lava

Enlightened
Joined
Feb 8, 2006
Messages
412
Re: TEC - cooled 5mW pointer, just for fun.

OK, little update.. I wrote a control program for the TEC and put a thermistor in with grease right at the back of the diode (the thermistor is only about 1mm dia.). I then set up a little simple data acquisition with a PIC 16F872 and the computer's serial port.

Then I ran the experiment of decreasing the temperature very slowly and monitoring the laser's output. Here is a plot of two runs of this:

tec_temp_power.gif


No, It's not a Fraps capture ;)

So, the ideal temperature for this diode and crystal combination seems to be 21.5 Degrees C when it is driven at 400mA.

What could cause such a big jump as seen in this graph? I have some theories but I'll let youall give it a shot. Post below!

The next experiment will be to try this at a different set current and see how much the plots differ. In the end, a 3-D surface could be generated of pump current and temperature. That would be pretty cool to see.


Note: each pump diode and crystal combination is different, so this plot is unique to this laser and doesn't mean that all lasers are most efficient at 21 Degrees!
 

Corona

Enlightened
Joined
Aug 1, 2005
Messages
246
Location
Chicago Area
Dr_Lava said:
What could cause such a big jump as seen in this graph? I have some theories <clip>

Interesting. Looks like something unsavory is happening at +/- 4 degrees from your optimal power point. The relatively large downward excursions at ~18 and ~26 degrees seem repeatable.

I'm guessing it's optical alignment, and not laser diode, related. All that brass and glass wandering around...

As a control, you might plot the detector response over time and see how much that drifts (against your optimal power temperature). Just to prove that there is no time-related shift as the detector heats up ;-)

A theory on the "peakyness" - the pump LD center frequency is temperature dependant and is it possible that the green output varies in wavelength enough to span the detector's passband.

Also the "crystal set" has an optimal center passband which will have a largely peaky response.

Of course there is a physical upper limit to pump current; the bonding wires and die size of the LD as well as the facet area determine where it goes <poof>. 400mA seems about as far as I would go, having killed a few LDs at 450mA (unchilled)...

All that aside - a very cool project! I've been wanting to do this for awhile; have a few Peltiers laying around - fans and heatsinks are NO problem for me :D

Good work!

PS - that's what MY bench looks like!!
 

msnyder

Newly Enlightened
Joined
Mar 29, 2006
Messages
33
Location
Palo Alto, CA
Re: TEC - cooled 5mW pointer, just for fun.

RickSCV said:
USB ports typically supply 5v at 500ma right? If so then this unit could be dissassembled for its peltier for a portable project correct? If not then atleast I'll be able to cool my beer with it.

http://www.thetravelinsider.info/travelaccessories/coolitchiller.htm

Saw those at Fry's last night. If you happen to live in Fry's territory,
you wouldn't have to wait or pay for shipping...
 

dr_lava

Enlightened
Joined
Feb 8, 2006
Messages
412
Corona said:
I'm guessing it's optical alignment, and not laser diode, related. All that brass and glass wandering around...

As a control, you might plot the detector response over time and see how much that drifts (against your optimal power temperature). Just to prove that there is no time-related shift as the detector heats up ;-)


I can see in the beam scatter the exact times that it gets brighter and dimmer (for the large changes) , so at least that part of it is not LED detector related. I,too, believe it's optical alignment related, as something related to the cavity is shifting or aligning to get the response jump.

Here is a plot of 36 minutes of runtime at a peak temp. setpoint. ( I did a test at work with a 40mW laser and re-did the diode calculation readings - indicating a new high of about 100mW! The speed that it burns through tape seems to substantiate this.


laser_fixedtemp.gif


Note that this is done entirely without optical feedback,only temperature feedback, so the stability is pretty good. After cooling and restarting the system, peak output (97mW) was achieved again, so the slow power reading decrease is caused by some as-of-yet unknown factor. I'll have to monitor my power supply for voltage sag and the laser current for sag as the driver board slowly heats (the whole room heats up!) So it's not impossible that it's detector diode related.
:popcorn:
 

Athoul

Enlightened
Joined
Oct 5, 2005
Messages
391
You might also be seeing the results of what occurs when the diode red shifts by heating up, and blue shifting by cooling down. If the diode strays from 808nm in either direction it will cause the crystal to not lase as efficiently. You graph illustrates the effects it would have by being to "cool" and to "hot." In the case that it is to cool, the diode may be 805-806nm..and when it gets to hot, 810nm or so.
 
Top