• State TI Thinks Resolved
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 62 subscribers
  • Views 97 views
  • Users 0 members are here
Support feedback
Options
Options
Related

LM1117-Q1: ThetaJA of LM1117-Q1(KVU) vs TLV767-Q1(SON)

Yoshikazu Kawasaki
Intellectual 961 points

Part Number: LM1117-Q1
Other Parts Discussed in Thread: TLV767-Q1, , TPS7B4253-Q1

Hello,

My customer has been investigating low cost LDOs from 16V(max) to 5V/1A.  If I searched ti.com, I found TVL767-Q1 and LM1117-Q1 would be the candidates.  It needs to have low thermal resistance for this case, so I checked ThetaJA of those and found that the bigger package LDO(LM1117-Q1 in KVU has 67.2degC/W) has higher than smaller package LDO(TLV767-Q1 in SON has 51.9degC/W).  Would you please tell me the reason why it is because it should be opposite normally like this site shown below?  I don't see advantages on LM1117-Q1 comparing with TLV767-Q1 because the sample price is also more expensive.

[FAQ] Thermal Dissipation from an LDO Perspective - Power management forum - Power management - TI E2E support forums

Best Regards,

Yoshikazu Kawasaki

  • 0
    TI__Genius 9830 points

    Hi Kawasaki-san,

    It looks like the TLV767-Q1 DRB simply has less thermal impedance between than junction and the PCB the LM1117-Q1 in this case, leading to an overall reduced RθJA. While bigger packages are generally better, this is not always true. Thermal metrics can can be affected by die size, pass element size, mold compound between die and thermal pad, etc. Sometimes these factors are more significant that the package-PCB surface area connection.

    The metrics are simulated on a board compliant with JESD51-7. The standard PCB lacks some of the features we recommend to improve thermal performance, such as thermal vias (would have marginally more significant effect on a KVU package).

    11W ( (16V-5V)*1A ) of power dissipation is alot. At an ambient temperature of 85'C, you'd need an RθJA better than 3.6'C/W to avoid heating the part past its recommended range. That is a very hard thing to accomplish. 

    The LM317KCS may also be a viable candidate here.

    Best,

    Gregory Thompson

  • 0 in reply to Gregory Thompson
    Intellectual 961 points

    Hello Gregory-san,

    Thank you very much for your quick reply.

    My customer asked me, so I'd like to know exactly the reason why LM1117-Q1 has higher ThetaJA than TVL767-Q1.  I need to tell them the technical reason.  Would you please tell me that?

    Yes, the power dissipation is quite high, but it doesn't seem to have the same condition of 16V input with 1A output.  I'll propose tracking LDO like TPS7B4253-Q1 as well, but it needs to be low cost to beat competitors.  In this case, it doesn't have to be Q1 devices, but they'll definitely ask failure analysis if they get failures, so I need to propose Q1 devices from TI which also means it's quite difficult to beat competitors.  It seems some competitors accept failure analysis even non-Q1 devices.

    Best Regards,

    Yoshikazu Kawaaski

  • 0 in reply to Yoshikazu Kawasaki
    TI__Genius 9830 points

    Hi Kawasaki-san,

    Ultimately, it's because we performed a simulations compliant to the same JEDEC standards for both parts and the thermal metrics calculated from the results were different. Unfortunately that's the most technical reason we have. The lower thermal impedance along the primary thermal pathway previously mentioned helps explain the result, but this is not a rigorous technical reason.

    I see, I had assumed the 16V was the input voltage for the conversion.

    Regarding failure analysis, I talked to the team and we have performed failure analysis on catalog LDOs if there is a concern that it is a systematic issue.

    Best,

    Gregory Thompson