• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 3 answers
  • Subscribers 63 subscribers
  • Views 119 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.

ISO5852S: ISO5852S: Clarification on VIORM change and long-term reliability

pepa Zdepa
Prodigy 10 points
Part Number: ISO5852S
Other Parts Discussed in Thread: UCC21710, , UCC21750

Tool/software:

Hello TI Team,

We are evaluating the ISO5852S for a new design and have a question regarding its long-term reliability, prompted by a change we noticed in the datasheet specifications over time.

An older datasheet for the automotive version (SLLSEQ2A, from 2016) specified VIORM = 2121 V. However, the latest datasheet for the industrial version (SLLSEQ0C, from 2023) specifies a significantly higher VIORM = 2828 VDC/ 2828 Vpk.

Our application requires functional isolation with a continuous working voltage of 2500 VDC. We do not need to meet the formal VDE certification for this design.

Given the historical change in the specification, we would like to ask for clarification:

  1. Was the increase from 2121V to 2828V the result of a physical change/improvement to the component, or was it based on new characterization and re-certification to a different standard?

  2. Can we confidently rely on the ISO5852S for long-term reliability (e.g., a 15+ year lifetime) at a continuous 2500 V<sub>DC</sub>, or is this value primarily a certification limit rather than a physical operational limit?

We want to ensure we understand the basis for this change before committing to this part for our design's lifetime.

  • 0
    TI__Mastermind 23176 points

    Hi Pepa,

    I am trying to find more history, but there is a pending update on most of our datasheets due to a new VDE standard. The Visom will be increased from 8kV to 10.4kV to reflect the updated test condition of IEC 60747-17 vs. VDE 0884-11, which was the old standard that required 1.6x of margin for this spec. Also, there was an adjustment in the TDDB lifetime standard, which affects Viorm:

    Previously: 135 years

    Current: 34 years 

    The different Viorm is based on the less conservative guidance to preserve the isolation capacitor lifetime projection. In the near future, this may be increased further to the level that creates a lifetime of only 24 years. This device is from a process called LBC8LVISO.1, and the lifetime vs. Viorm working voltage rating is determined from this graph.

    Best regards,

    Sean

  • 0 in reply to Sean Cashin
    Prodigy 10 points

    Hi Sean,

    Thank you again for your very helpful explanation on the ISO5852S VIORM rating change.

    Based on that discussion, I have a specific follow-up question regarding the UCC21710. Its datasheet currently specifies VIORM = 2121 Vpk.

    Our new application requires functional isolation with a continuous working voltage of 2500 VDC. This is clearly above the official VIORM rating in the datasheet, but your previous explanation gives us hope that the part's physical capabilities might allow for this.

    We understand that you cannot officially approve operation outside the datasheet specifications. However, could you provide any insight or share any characterization data regarding the long-term reliability of the UCC21710 when subjected to a continuous 2500 VDC?

    Specifically, we are trying to understand if the 2121 Vpk limit is based on older, conservative lifetime models (like the ISO5852S was), or if there is a fundamental limitation in the UCC21710 that prevents reliable operation at 2500 VDC.

    Any guidance you can offer would be critical for our design assessment.

    Thank you.

  • 0 in reply to pepa Zdepa
    TI__Mastermind 23176 points

    Hi Pepa,

    UCC21710 is a newer design with smaller dimensions, from a process called LBC8LVISO.2. Here is its isolation capacitor lifetime projection chart, in units of Vrms, not Vdc:

     

    If you use this device at 2500V DC instead of the rated 2121DC, then the Time Dependent Dielectric breakdown is projected to cause a >1ppm failure rate 

    after a lifetime that decreased logarithmically from 54 years to 9.4 years: 54/ (100^((2500-2121)/1000)

    Best regards,

    Sean

  • 0 in reply to Sean Cashin
    Prodigy 10 points

    Hi Sean,

    Thank you for your previous response. I am re-reading it and would like to ask a few follow-up questions to ensure my understanding is correct, especially regarding some datasheet comparisons.

    1. VIORM Change Rationale: Am I correct in understanding that the V_IORM for the ISO5852S was increased to 2121 Vpk bipolar primarily to align with newer industry standards?

    2. Datasheet Discrepancy (External Clearance): I've noticed a difference in the documentation. The ISO5852S datasheet (SLLSEQ0C – AUGUST 2015 – REVISED MAY 2023) specifies an External Clearance of 14.5 mm. However, the ISO5852S-Q1 datasheet (SLLSEQ2A – SEPTEMBER 2016 – REVISED DECEMBER 2016) states it is only 8 mm. Since both parts use the same SOIC (DW) 16 package, could you clarify which value is correct or why they differ?

    3. Datasheet Discrepancy (VIORM): Similarly, the V_IORM in the standard ISO5852S datasheet (rev. MAY 2023) is listed as 2828 Vpk bipolar, while the ISO5852S-Q1 datasheet lists 2121 Vpk bipolar. Are both of these values correct for their respective parts?

    4. Barrier Lifetime Comparison (ISO5852S vs. UCC21750): My final question is about comparing isolation barriers. Do I understand correctly that the SiO₂ barrier in the UCC21750 is considered less robust than the barrier in the ISO5852S? Specifically, for our application with a continuous working voltage of 2500V DC, would this imply that the expected barrier lifetime of the UCC21750 is significantly lower compared to the ISO5852S?

    Any clarification you can provide on these points would be greatly appreciated.

  • 0 in reply to pepa Zdepa
    TI__Mastermind 23176 points

    Hi Pepa,

    1. The new 2828 Viorm was adjusted to reflect a shorter lifetime guidance (135 years down to 34). This was an internal TI decision. In the near future, this may be increased further to reflect a lifetime of only 24 years, due to a new industry standard. 

    2. I don't know why this datasheet says 14.5 mm. The device should only have 8mm of creepage and clearance. I will try to correct this. The Q1 datasheet says 8mm and they are identical devices except for additional testing.

    3. The Q1 datasheet seems to still base this 2121Viorm off of a 135 year lifetime (1500Vrms). This was the first isolated gate driver made by our product line, and later devices have specified Viorm less conservatively. The industrial version appears to have been updated in 2015 to remove this margin, while the automotive version retains the original Viorm. 

    4. That is one interpretation. The distance through the isolation is related to the process technology, and newer process technologies have smaller features, lower cost, and often higher performance. The isolation voltage is indeed lower, but that is not often the limiting factor in a design. Newer devices have an even smaller 17um distance through isolation. 

    Best regards,

    Sean