Laser controller system.

[ARG]

I'm working on a new project to stabilize diode power & temperature for pumping purposes.

I've come up with this:
<Snip, found a mistake>

It has input for a photodiode and an I2C temperature sensor.

The diode and TEC will be driven with analog drivers and with the feedback from the photodiode and I2C temp sensor the two DAC's will change the modulation voltage to the analog drivers.

It also has a display for information and buttons to adjust laser power & temp.

The physical setup will look like this. There will be a piece of glass mounted on a brass mount at a 45 deg angle to the beam which will reflect a small portion of the light into the photodiode. The rest of the light will be transmitted through the glass.

The diode mount is a 12mm CNC rail holder with a I2C temperature sensor I bolted on and a TEC to keep the system at a constant temperature.

The heatsink is overkill since this is a low power system, but it will keep the baseplate cool from the heat of the TEC & drivers.

Let me know what you think about my idea!

 

[djQUAN]

I would mount the temp sensor closer to the diode as possible for more accurate diode temp reading. I'm assuming you'll use springs to mount the 12mm mount to the base to have constant pressure on the TEC? Otherwise, the design looks fine.

 

[ideas]

Euh.. small error.. you bridge your TEC by the two metal screws allowing heat to flow from your (hot) output side to your cooling side.... => use non heat conducting screws or you're 'screwed'...

 

[ARG]

Euh.. small error.. you bridge your TEC by the two metal screws allowing heat to flow from your (hot) output side to your cooling side.... => use non heat conducting screws or you're 'screwed'...

Thought of that myself while looking at it, I'm going to replace them with nylon screws

I would mount the temp sensor closer to the diode as possible for more accurate diode temp reading. I'm assuming you'll use springs to mount the 12mm mount to the base to have constant pressure on the TEC? Otherwise, the design looks fine.

Yeah, I've had others tell me to use a smaller sensor for a faster response as well. Going to change this up.
I'm going to clamp the TEC with the screws, and put a dab of thermal paste in between.

 

[djQUAN]

since the screws are long enough, you can use nylon bushings (the type to insulate transistor tabs from the heatsink) to insulate the screws and add springs to keep constant and equal pressure on the TEC even when the assembly expands/contracts due to changes in temp much like the way they mount CPU heatsinks

edit: the 12mm clamp's thermal mass is much larger than the TO-220 temp sensor so it should be fine. Getting a small sensor for fast response may not be needed as the thermal inertia of the clamp will still be slow. You'd just need the sensor closer to the 12mm module for more accurate sensing.

 

[Cyparagon]

I've had others tell me to use a smaller sensor for a faster response as well.

That's not necessary. The temperature of the sensor will change just as quickly as the block. The larger package actually ensures that better than a smaller one in the sense that the contact area is larger. Even if it did somehow give a slower response, it will still be much faster than the TEC.

 

[Multimode]

Thought of that myself while looking at it, I'm going to replace them with nylon screws .

Hi Arg,

Just use some nylon washers (or Macor ceramic) under the head of the nut - or sandwich the nylon washer between two steel washers to prevent it form being chewed when tightening the nuts down.
Make sure the hole give sufficient clearance so the bolt does not make contact.
ATB
MM