Because of the holidays, TI E2E™ design support forum responses will be delayed from Dec. 25 through Jan. 2. Thank you for your patience.

  • State TI Thinks Resolved
  • Locked Locked
  • Replies 13 replies
  • Answers 4 answers
  • Subscribers 31 subscribers
  • Views 2065 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.

TAS5624A: Heatsink mounting recommendation : Maximum torque force that IC can sustain

Wee Keat Chin
Intellectual 435 points
Part Number: TAS5624A
Other Parts Discussed in Thread: TAS5624

My customer facing some market feedback due to D-amp IC broken.

This D-amp IC broken caused by mechanical heatsink vibration which make some IC pin break.

Customer would like to understand how much kg/f can stand for screw torque or maximum load (kg) withstand when the heatsink mounted on top of the IC.

Customer wants to study to improve vibration to avoid broken pin IC with strengthen the screw torque of the heatsink.

Please provide heatsink mounting data(screw torque) /recommendation for HTSSOP or similar IC package.

Thanks.

Best regards,

Chin

  • 0
    TI__Mastermind 24835 points
    Hi Chin,
    We are looking into this and Adam will get back to you later.
  • 0
    TI__Guru 63350 points

    Chin,

    On our EVMs we use M3 screws and 60-80 Newton Centimeters of torque to tighten the screws. This is only a rough estimation as we have not mechanically tested the heatsinks, PCBs, or devices.

    We use 60-80 Ncm because it is the minimal number that gives us good thermal performance.

    Regards,

    -Adam

  • 0 in reply to Adam Sidelsky
    Intellectual 435 points

    Hi Adam,

    Thanks for your reply.

    I will inform customer to control the torque force at range 60-80 Ncm for optimum thermal performance.

    Just to confirm that TI doesn't have measured/simulated data for maximum torque, am I right?

    Thanks.

    Best regards,

    Chin

  • 0 in reply to Wee Keat Chin
    TI__Guru 63350 points
    Chin,

    We only have this recommended value from our own experience. We have not tested a maximum value.

    Regards,

    -Adam
  • 0 in reply to Adam Sidelsky
    Intellectual 435 points

    Hi Adam,

    Sorry to trouble you again.

    My customer has additional questions. Please refer to attachments for details of failure.

    TAS5624A IC pin lead broken .xlsxTAS5624A IC pin lead broken .pdf

    1) Customer wants to know what is the maximum displacement of pin lead when overstress happens?

    Is it possible to simulate ? Could you please consult your IC package designer?

    2) Is it possible the vibration of IC or heatsink at resonance frequency caused the IC pin lead broken?

     Customer actually has other model using same IC but different size of heatsink and didn't encounter the problem so far.

     So customer suspect the IC vibrate at resonance frequency and overtress condition causing the pin lead broken.

    3) From the part spec, the height between pins and IC mold body is ranges 0.05 – 0.15. May I know the design purpose ? Is it for soldering or coplanity consideration ?

    Thanks.

    Best regards,

    Chin

  • 0 in reply to Wee Keat Chin
    TI__Guru 63350 points
    Chin,

    I am not sure if we can answer these questions but let me ask.

    Regards,

    -Adam
  • 0 in reply to Wee Keat Chin
    TI__Guru 63350 points
    Chin,

    1. Unfortunately this isn't something we can simulate. We only torque the heatsink enough to maintain thermal contact at the pressures I mentioned above.

    2. I don't think vibration is the cause.

    3. My thought is this is tolerance. If the pins are a small amount below the mold body, we can guarantee that they will be level with the PCB including any tolerance variation.

    Regards,

    -Adam
  • 0 in reply to Adam Sidelsky
    Intellectual 435 points

    Hi Adam,

    Thanks for your reply.

    i) Based on your explanation, since there has no data for the pin lead displacement due to stress from top, does it mean that this parameter is not a consideration during IC package design?

    ii) If vibration is not the cause, could you suggest any possibility that could lead to such failure? Customer needs to find out the root cause.

      [The Damp IC is still functioning despite of the broken of pin lead and failure only happened to the specific model with different heatsink and heatsink bracket]

    Please give some hints.

    Thanks.

    Best regards,

    Chin

  • 0 in reply to Wee Keat Chin
    TI__Guru 63350 points

    Chin,

    In the following picture, is the pink piece a thermal pad? Please let me know the Thermal pad part number.

    Once mounted in position, does the heatsink touch the PCB? Or is it only touching the top of the DAMP?

    Also, what is the screw model number or type?

    Regards,

    -Adam

  • 0 in reply to Adam Sidelsky
    Intellectual 435 points

    Hi Adam,

    Besides touching the top of DAMP, the heatsink touches the PCB through heatsink bracket.

    Please check attachments for the assembly of heatsink, thermal pad and screw drawing and datasheet.

    D-AMp Heat Sink construction.pptx

    Do inform if you need more info.

    Thanks.

    Best regards,

    Chin

  • 0 in reply to Wee Keat Chin
    TI__Guru 63350 points
    Chin,

    I understand the drawing, thank you.

    It looks to me that even though the heatsink bracket is supported on the PCB surface, the large mass of the heatsink itself is still unsupported. This could mean that the heatsink mass itself could flex the heatsink bracket during vibration and cause damage. Does the system typically see vibration in the normal use case?

    Can the customer check with the silicon pad manufacture and the heatsink manufacturer about the force needed to achieve correct thermal performance?

    Has the customer tried reducing their heatsink screw torque and checked for thermal performance?

    Regards,

    -Adam
  • 0 in reply to Adam Sidelsky
    Intellectual 435 points

    Hi Adam,

    Your inference could be correct as customer found out that some of the protruding supports on the heatsink bracket are out of specs(height shorter).

    Hence the heatsink is not fully supported and sitting on the PCB. This failure is so rare as it only happened in one model, coincidentally the heatsink and heatsink bracket used is different from others.

    Customer now is simulating to reproduce the failure. Since vibration is the suspected culprit, their study approach could be finding the resonance frequency.

    Customer wants to understand whether TI happens to encounter similar failure or comments about the resonance frequency that could lead to break the IC pins.

    For the correct thermal performance, do you mean that the minimum torque required in order to achieve the thermal conductivity as stated in the datasheet?

    Example refer to the silicone pad, the thermal conductivity is 1.7W/mK [ISO 22007-2]

    Thanks.

    Best regards,

    Chin

  • 0 in reply to Wee Keat Chin
    TI__Guru 63350 points

    Chin,

    TI uses a heatsink such as the following for the TAS5624:

    https://www.digikey.com/product-detail/en/advanced-thermal-solutions-inc/ATS-TI1OP-519-C1-R3/ATS2180-ND/5125225

    As you can see from the image, the heatsink has feet that touch the PCB on both top and bottom of the IC, also no thermal pad is used, the heatsink rests directly on the IC with thermal paste applied.

    Since we have two square feet touching the PCB surface and no thermal pad, this heatsink is very sturdy.

    We have not heard of any issues like this from other customers.

    The minimum torque mentioned is based on our findings for the amount of torque needed to give us adequate thermal performance in our application.

    Regards,

    -Adam