Being able to thermally stabilize an LED is useful for any sort of high accuracy sensing application. I was looking for an LED module that would incorporate a thermistor and something to adjust the temperature like a TEC.

I was able to find this company: https://www.laserscom.com/products that have the three packaged together with a fiber optic connector coming off of it and am wondering if I can find more convenient options (In terms of location/availability). But overall, I'm finding it surprisingly hard to find anything like it. I tried looking in the different optoelectronic categories at digikey and looking at the product list for some big LED companies but really didn't find anything like it.

It seems like something that would exist based on reading application notes and datasheets of PID controllers intended to work in tandem (http://www.analog.com/media/en/technical-documentation/data-sheets/ADN8835.pdf). Am I just not using the right search terms to find it or are LED+thermistor+TECs typically bought separately and put together?

  • \$\begingroup\$ I've never heard about LEDs with built-in compensation. The reason for this, I think, is that having the two separate gives designers wide range of options while keeping the price down. Maybe look at this from different direction? If the goal is to stabilize LED output then the output must be the source of feedback, which would cover temperature drift as well as manufacturing differences in LEDs. Pretty much like they do it in linear optocouplers. Of course then you need stable photosensor, but it could be easier to do \$\endgroup\$
    – Maple
    Jun 5 '18 at 19:57
  • \$\begingroup\$ We've always bought the LED (some come with a thermistor) and TEC seperately. \$\endgroup\$
    – D Duck
    Jun 5 '18 at 21:08
  • \$\begingroup\$ What's the application? What kind of environment, hot, cold? Just one single LED? LED Vf, watts, part number? \$\endgroup\$ Jun 6 '18 at 13:34
  • \$\begingroup\$ Okay, sounds like separate is the norm. Thanks for the insight! @Maple Serving on the output works great for most of applications. However, I have one where I'm trying to simulate a decaying fluorescent signal with an RC circuit and the non-linear change in Output vs Input current at different temperatures is too significant for me. \$\endgroup\$
    – Colin
    Jun 8 '18 at 17:14
  • \$\begingroup\$ @Misunderstood The application I mentioned in another comment. It's running in an indoor environment, and I would expect it to be in the range of 15-30°C. It would be a few different channels with one LED, LED Vf ~2.1V, up to 100mW, probably something similar to EPITEX L625-36 \$\endgroup\$
    – Colin
    Jun 8 '18 at 17:17

It's running in an indoor environment, and I would expect it to be in the range of 15-30°C. It would be a few different channels with one LED, LED Vf ~2.1V, up to 100mW, probably something similar to EPITEX L625-36

The best way to keep LEDs running cool is to use an LED with high efficacy and a thermal pad.

The Epitex L625-36 @ 50 mA, 2.2V Vf = 110 mW, outputs 23 mW of light. Efficacy = 23/110= 20.9%

The problem is cooling an LED embedded in 3 mm of epoxy resin with no thermal pad.

Now if you were to use a Cree XP-E P/N XPEBRD-L1-0000-00901
It has a photon output of 2.10 µmol/s (400 mW Radiant Power) @ 350 mA x 2.2V Vf = 770 mW total power dissipation. Efficacy = 400/770 = 52%.

52/20.9 = 249% so you will have ~2.5x less heat to deal with due to efficacy.

Plus the total power dissipation will be lower due to a ~10% lower Vf at lower current.

Less heat flux makes thermal management easier.

15°C variation in ambient is the issue. If ambient were kept constant there would be no LED junction temperature deviation once the system reaches thermal equilibrium. Red LEDs are more susceptible to variation in temperature that all other colors except amber.

Temperature vs. Irradiance

Rather than thermal management you may want to consider using a photo diode rather than thermistor to keep the irradiance constant rather than temperature.

The circuit below (yellow) will regulate LED temperature but does so by changing the current which affects irradiance. But if you replace the thermistor with a photo diode it will keep the irradiation constant.

This circuit incorporated with a linear constant current regulator would keep it very simple with very little ripple current.

LED Thermal Management circuit


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.