The big TEC driver thread!

[Solarfire]

Perfect work !!! Iam searching for a TEC driver like that a long time. But theres one thing i miss. A Tec need more current for cooling than for heating. I think it would be better to control the current limit of heating and cooling individually. Iam also interested in 2 of these tec drivers.

Current limiting is only necessary for TECs that are rated for less than 12V or less than 3 Amps the PID will take care of the rest. In order to get an equal response for heating and cooling 2 sets of PID parameters would be necessary but the response time is plenty fast enough as it is. The graphs you see above are the results of a rough setup and were not further tuned due to missing revisions of the PWM filter on this prototype build. As soon as the pre-series boards get here with these filter changes implemented, some more response graphs of TECs from 4W to 30W types will be posted.

Here's a thread to a GB Interest poll http://www.photonlexicon.com/forums/...TEC-LD-drivers

[Eu1eOne]

I need a tec controller with more power than 3Amps. 48V @ 10A would be fine. I need a tec controller for my optical base plate.

[RedlumX]

Inspired by this thread I developed a highly stable TEC controller as Arduino shield. This is meant to stabilize laser diodes for holography, for show applications this would be overkill. Details are here: http://hololaser.kwaoo.me/electronic...duino-TEC.html

[dave]

Looks nice

Perhaps the xmega might be a better choice for faster PWM? or another nice option is the STM32F446 which would allow 12bit PWM at just shy of 90kHz, and the NUCLEO-F446RE has an Arduino form factor

[RedlumX]

Looks nice

Perhaps the xmega might be a better choice for faster PWM? or another nice option is the STM32F446 which would allow 12bit PWM at just shy of 90kHz, and the NUCLEO-F446RE has an Arduino form factor

Hello, yes I know but I wanted to keep it cheap and simple; the speed of the Uno is enough. What prevented me to look in more detail is that those all run on 3.3V which brings a ton of other inconveniences wth it.

[planters]

Do you really need a controller? The TEC acts as a simple resistor. Unlike a diode, it will self limit its current depending on the voltage. I would suggest you mount the TECS and using a lab supply of if you do not have access to one with that kind of power capability, then try a couple of different voltage fixed supplies, narrow down the good enough, best voltage requirement and install that voltage directly to the TEC from a fixed power supply and you're done.

[catalanjo]

@"and you're done. " ......not quite ... might be best to add a cheap thermostat to protect the LD (just in case), for when the peltier fails.

Cheers

[RedlumX]

Well, to achieve millikelvin stability some sort of feedback loop seems quite essential...

[planters]

RedlumX,

Just got around to reading through your write up on the controller. Nice work and an impressive result.

Additional considerations for extremely stable LD operation would be:

A hermetically sealed container for the diode should be built or better yet a vacuum chamber to isolate the diode from air currents and fluctuations in room temperature.

A highly stable driver is necessary because all LDs shift their wavelength with drive current. This effect is more pronounced in the multimodes, but it is still a large effect in single modes compared to the effects of temperature at the mil-Kelvin level.

Ultimately, a high dispersion spectrograph (specially built by you to greatly expand only the wavelength of interest) illuminating a linear CCD array with a feedback loop to the current and or temperature could help you to directly stabilize the wavelength.

[dave]

RedlumX,

Just got around to reading through your write up on the controller. Nice work and an impressive result.

Additional considerations for extremely stable LD operation would be:

A hermetically sealed container for the diode should be built or better yet a vacuum chamber to isolate the diode from air currents and fluctuations in room temperature.

A highly stable driver is necessary because all LDs shift their wavelength with drive current. This effect is more pronounced in the multimodes, but it is still a large effect in single modes compared to the effects of temperature at the mil-Kelvin level.

Ultimately, a high dispersion spectrograph (specially built by you to greatly expand only the wavelength of interest) illuminating a linear CCD array with a feedback loop to the current and or temperature could help you to directly stabilize the wavelength.

Perhaps you should check out the rest of his site Eric http://hololaser.kwaoo.me/argonlaser.html I think even you could learn a thing or two from Wolfgang Very cool site.