• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 64 subscribers
  • Views 725 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

LM2594HV 12V o/p thermal shutdown query

Phil Dixon
Prodigy 70 points

When thermal shutdown (or even temporary output short) the output voltage drops to something like 0.5V but the inductor and the associated capacitor get very hot and prevent the regulator from recovery even after a power cycle, or at least until the inductor has cooled significantly.  It would appear that the thermal shutdown protects the IC but does something odd to induce the inductor/cap to get very hot.  I believe I'm using a bog-standard application circuit but is there something I'm doing wrong to cause this problem?

  • 0
    TI__Genius 16025 points
    Phil,

    Please share your layout and schematic. It will whittle down the possibilities and get us to a solution much faster.

    If your ambient temperature is above the thermal threshold of the IC, then thermal shutdown will protect the IC by design. However if your design is at 25C and you are seeing the part thermally shutdown while you power it up, then we need to investigate your design.

    Thanks,
    Anston
  • 0 in reply to Anston Lobo
    Prodigy 70 points

    I will try to post a schematic and layout later but in the meantime.

    I have tried adding a small heatsink to the device and it does appear to have stopped the shutdown.  I'm thinking that the device needs more ground plane heatsinking to prevent this problem.  The device does appear to get pretty warm considering that I'm not actually drawing any current most of the time, this is a power supply for an alarm buzzer that is very rarely activated. 

    When not loaded the IC gets about 35degC above ambient with a small heatsink but with no heatsink (and pretty much no ground plane either) it would appear that if the ambient gets above about 40degC then the shutdown occurs.

    I must admit I had not expected this and didn't spend any time thinking about proving some grounded plane to heatsink this part, I notice though that the examples in the datasheet are looking for 4sq" of 2oz copper, I know my current requirement s much lower than the example but even so I think using an example of 4sq" must e saying something.  I would say though you do have to wonder who would use a small soic 8 device and then give it that much board space!

    I'm going to look at adding/increasing the ground plane and possibly using the On/Off pin to ponly trun on the device for the short periods the buzzer is active.

  • 0 in reply to Phil Dixon
    TI__Genius 16025 points
    Hi Phil,

    I would highly recommend using Webench for your design. It will in, checklist style, ensure you've covered all the design considerations.

    Now, even when the part is in light load it is dissipating a small amount of energy. Without any heatsink, that heat just builds up incrementally until you see the issues that you mentioned. The heatsink doesn't need to be as large as you mentioned, because it is dependent on the load requirements.

    I would suggest using a more recently released part only because the light load power dissipation would be lower and the efficiency much higher leading to an overall cheaper and leaner design.

    But to answer your question you definitely need a GND plane heatsink. Further recommendations when I have a schematic/Layout.

    Thanks,
    Anston
  • 0 in reply to Anston Lobo
    Prodigy 70 points

    On re-reading the datasheet I see that the NC pins should also be connected to the ground plane to remove heat as well as generous ground plane under the IC.

    I'm quite embarrassed to say that I originally made the assumption that a switching regulator wouldn't dissipate any energy, or at least so little that heatsinking wouldn't be required, my original pcb layout was extremely low on ground plane around this device. 

    I tested my board by adding a small heatsink to the top of the IC and things got much better. 

    I'm now updating the pcb to add significant ground plane to this area so this should now be fixed. 


    Thanks for your help Anston.