• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 64 subscribers
  • Views 132 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Original question:

Thermal Management

Thermal Management - Temperature Sensor

Rodrigo Vernini
Intellectual 290 points
Other Parts Discussed in Thread: LM317

i have been developing a temperature/humidity sensor for a specific enclosure. As the project evolved the hardware got more features, and i've been having some issues with my thermal management.

My first project was aligned with common PCB thermal management best practices, with a solid cooper plane on entire PCB, had some problem with the sensor heating ~1°C.

The second project was with a solid polygon on top half of the PCB, and the sensor on the bottom of the PCB, without a polygon. That was a nice solution, until a client installed the product on the ceiling, in a horizontal position. This caused a increase of 4°C on the sensor, as it had no air flow on it.

Now i'm reprojecting the PCB, and one thing i've been looking at is for components enclosures with the lowest thermal resistance, see an example of a LM317 thermal resistance:

LM317 thermal management

On my first thought i would choose the PK enclosure, as it has the lower temperature increase per watt. But that could be the same mistake as my first project design (or not, i would like to understand that!), it could just dissipate it's heat better, but, after all, generate the same amount of heat inside the enclosure.

So, in this application, in an enclosure that does not have air flow, and all heat generated is "accumulated" is there any difference by choosing differents enclosures? Or the efficiency on dissipating electrical power on thermal is always the same?

  • 0
    TI__Guru 71526 points

    Hello Rodrigo, 

      We are not clear on what you are asking to accomplish with the temperature measurements you are describing. 

    Are you wanting to measure the temperature on different pcboards, or the temperature in different types of enclosures?  

    And is the LM317 power dissipation just an example, or is the LM317 in your design?  If so, how does it fit into this evaluation? 

    Drawings/diagrams of what the actual design looks like would be helpful.

    Regards,

    ~Leonard 

  • 0 in reply to LeonardEllis
    Intellectual 290 points

    Hi Leonard.

    Sorry for the long story.

    My question is:

    Does different linear regulators generates the same amount of heat (with the same current, Vin, Vout)? Not talking about IC temperature, talking about total heat generated (included the dissipated heat).

    If no, does different packages from the same IC generates the same amount of heat?

  • 0 in reply to Rodrigo Vernini
    TI__Genius 14245 points

    Hey Rodrigo,

    If the Vin, Vout, and Iout are all the same then the total power dissipation will be the same which means that the heat generated will be the same but as you are alluding to the IC temperature can vary between parts/packages based on how well that part/package dissipates the heat to the environment (mostly through the PCB). So choosing a package with the lowest thermal resistance could result in your PCB heating up more. This is obviously a design trade off, since you don't want to heat your PCB any more than necessary but also don't want the LDO to get so hot that it is out of spec (or wears out prematurely since it is always running at high temperatures). 

    So there are a few things I can think that you might be able to do

    • Reduce the power dissipation of the LDO, your output voltage and load current are probably fixed requirements so that probably means lowering your input voltage while keeping the LDO out of dropout (and with head room for good enough PSRR for your application)
      • If you have a large Vin to Vout differential then you may need a switching converter to improve the efficiency of your power solution. 
    • Further refine the PCB such that you can still dissipate the same amount of hear but keep it away from your sensor (sounds like you already tried this)
      • this can be done with metal routing as you mentioned, but also adding a heat sink to the PCB may prove helpful.
    • Counterintuitively you could choose an LDO package which has higher thermal resistance in order to reduce how much heat is transferred to the PCB. However you will need to be careful not to cause the LDO to get too hot that it goes into thermal shutdown, or doesn't meet your application needs and that it doesn't wear out prematurely (The JEDEC standard for IC's is 9 years at ~55C, so operating above a junction temp of 55C can result in earlier wear out).