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.

F29H850TU: Supply design

Part Number: F29H850TU
Other Parts Discussed in Thread: LP8758-E3, TPS3840, UCC25800-Q1, TPS65261-1

Hi,

Does TI has a recommendation for a supply design for the F29H850TU that is based on already available DC/DC converter?
I'm looking for a design for an industrial environment with 24V nominal input voltage (no functional safety requirements). Beside the needed voltages for the F29 (3.3V and 1.25V) with the mandatory sequence and slew rates we also need a +5V.
I had a look to the TPS650366 (TPS65036604RAYRQ1 or TPS65036605RAYRQ1) but it looks as this PMIC is not yet available or there are still some documentations missing (e.g. OTP parameter).

Thanks and best regards,
Patrick

  • Hi Patrick,

    TI does not publish a specific reference design for powering the F29H850TU from a 24V industrial input. However, a workable architecture can be assembled from existing TI DC/DC converters. Here's what you need to know:

    F29H850TU Power Requirements

    From the datasheet 1:

    Parameter Min Max Unit
    VDDIO/VDDA (3.3V) slew rate 3 100 mV/µs
    VDD (1.25V) slew rate 2 100 mV/µs
    Total VDD decoupling 10 22 µF
    VDDIO decoupling per pin — 0.1 µF

    Sequencing requirement: VDDIO (3.3V) must ramp first, then VDD (1.25V) follows. The delay between the two ramps is specified with a minimum of 0 µs (i.e., VDD cannot lead VDDIO) 1.

    Recommended Architecture

    Since no single TI PMIC accepts 24V directly and provides all three rails with sequencing, a two-stage approach is practical:

    Stage 1: 24V → 5V (Wide-Input Buck)

    • LM5145 — Synchronous buck controller, 6–75V input, suitable for 24V industrial with wide transient tolerance. Generates your 5V system rail.
    • LM25145 — Similar wide-input buck controller, 6–42V input range.

    Either device handles the 24V input directly and produces a regulated 5V output.

    Stage 2: 5V → 3.3V + 1.25V (PMIC with Sequencing)

    From the 5V intermediate rail, use a multi-output PMIC with integrated sequencing:

    • TPS65220 — 3 buck converters + 4 LDOs, 2.5–5.5V input, integrated power sequencing. Can generate both 3.3V and 1.25V with proper startup order.
    • TPS65219 — 3 synchronous bucks with integrated sequencing, similar input range.
    • LP8758-E3 — 4 buck converters with adjustable slew rate control, which directly addresses the F29's slew rate requirements.

    The LP8758-E3 is particularly relevant given its explicit slew rate control feature.

    Additional Design Note

    The F29H850TU's internal VDD POR trips below the minimum recommended VDD (1.14V), so it cannot be relied upon for supply supervision. For robust operation, add an external voltage supervisor on the VDD rail 2. Since you have no functional safety requirements, a simple supervisor (e.g., TPS3840) monitoring VDD driving XRSn is sufficient.

    Follow up: What are your estimated current requirements on each rail (5V, 3.3V, 1.25V), and what is the expected input voltage range around the 24V nominal?


    1. F29H850TU Datasheet - Power Sequencing & Electrical Specs
    2. E2E: F29H850TU VDD POR Availability
    3. F29H850TU Migration Guide

    Best Regards,

    Zackary Fleenor

  • Hi Zackary,

    Thanks for the detailed solutions for a F29H850TU supply.
    In the meantime I have worked-out a posible solution as well. This is based on the TPS65261-1 (we already use this part). With this part I plan to create a 3.3V, 1.25V and a 5V rail (comming up in this sequence). The TPS65261-1 will be supplied by a 12V rail. This 12V rail is created by an LLC converter based on the UCC25800-Q1 as we need a high isolation between 24V (input) and the F29H850TU circuit.
    Back to your questions:

    Current requirements:
    - 1.25V: 0.82A (worst case)
    - 3.3V: 0.78A (worst case)
    - 5V: <0.1A (not yet completely defined)

    Input range:
    - VIN nominal: 24VDC
    - VINmin: 18VDC (we might need to increase this value due to other restrictions, maybe VINmin will be 21VDC)
    - VINmax: 28VDC


    Thanks for the design note concerning VDD supervision. Unfortunately this is always a bit of a hassle as there need some tight tolerances to be fulfilled.
    We have used the TPS3808G01DBV for this kind of tasks but be often needed to raise the nominal VDD voltage a bit to not have overlapped tolerance fields.

    Regards,
    Patrick

  • Hi Patrick,
    Thanks for sharing the details — your TPS65261-1 + UCC25800-Q1 approach makes a lot of sense, especially given the isolation requirement between the 24V input and the F29H850TU circuit, and the fact that you're already familiar with the TPS65261-1.
    A few thoughts on your design:
    Current requirements vs. TPS65261-1 capability Your worst-case loads (1.25V @ 0.82A, 3.3V @ 0.78A, 5V @ <0.1A) are well within the TPS65261-1's output current capability, so you should have good margin — that also gives you headroom if the 5V load requirement grows once it's finalized.
    Sequencing Good to confirm the TPS65261-1 gives you the 3.3V → 1.25V → 5V startup order you need. Just double check the delay between the 3.3V and 1.25V ramps meets the F29H850TU's minimum 0 µs requirement (VDD/1.25V cannot lead VDDIO/3.3V).
    Isolation via UCC25800-Q1 LLC → 12V Makes sense for your isolation need. Worth confirming the 12V rail's line/load regulation and ripple stay within the TPS65261-1's input tolerance across your full VIN range (18–28V input to the LLC stage), especially near VINmin if you end up needing to raise it to 21V.
    VDD Supervision (TPS3808G01DBV) Understood on the tolerance-stacking challenge — this is a common pain point. A couple of options to ease the margin squeeze without raising nominal VDD too much:
    • Consider the TPS3808G01DBV's adjustable threshold version (external resistor divider) instead of a fixed-threshold variant, if you aren't already using it — this lets you fine-tune the trip point rather than relying solely on a fixed percentage tolerance.
    • Alternatively, a device with tighter accuracy (e.g., ±0.5% class supervisors) may reduce the amount you need to raise nominal VDD to avoid tolerance overlap.
    Best Regards,
    Zackary Fleenor
  • Hi Zackary,

    I'm not quite shure if I understand the "0us requirement" correct. Ad of my understanding VDDIO (3.3V) must come up before VDD (1.8V) .
    Attached is the power up sequence I have realized with some Evaluation boards. Maybe you can review if that is ok for the F29H850TU:


    CH1 (yellow): 12V rail, CH2 (red): VDDIO (3.3V), CH4 (green): VDD (1.8V), CH3 (blue): Master Enable for TPS65261-1 

    VDDIO (3.3V) comes up around 1.7ms prior to VDD (1.8V). Slew rate of VDD and VDDIIO is around 10mV/us.

    Regards,

    Patrick

  • Hi Patrick,

    Your power-up sequence looks correct for the F29H850TU.

    Sequencing Requirement Clarified

    The "0µs minimum" simply means VDD must never start ramping before VDDIO — they can start simultaneously (0µs delay) or VDDIO can lead by any amount. Your captured waveform shows VDDIO (3.3V) leading VDD by ~1.7ms, which satisfies this requirement with comfortable margin 1.

    Your Measured Parameters vs. F29H850TU Specs

    Parameter Your Measurement F29H850TU Requirement Status
    VDDIO ramps before VDD Yes (~1.7ms lead) VDDIO must not lag VDD (≥0µs) Pass
    VDDIO slew rate ~10 mV/µs 3–100 mV/µs Pass
    VDD slew rate ~10 mV/µs 2–100 mV/µs Pass
    Your design meets all power sequencing and slew rate requirements 2.

    TPS65261-1 Automatic Sequencing Note

    With MODE tied high, the TPS65261-1 uses a 1024-cycle delay between sequential buck startups 2. At 600 kHz switching frequency, this translates to ~1.7ms — consistent with what you're observing. Since Buck1 (3.3V) starts first and Buck2 (1.25V) follows after this delay, VDD can never lead VDDIO under normal operation.

    One Observation

    Your scope annotation mentions "VDD (1.8V)" — the F29H850TU specifies VDD at 1.25V nominal (min 1.14V, max 1.32V). Please confirm your output is set to 1.25V rather than 1.8V, as operating at 1.8V would exceed the absolute maximum rating for VDD.

    Follow up: Can you confirm the VDD rail is configured for 1.25V (the scope label shows 1.8V)? Also, have you verified the 12V LLC output regulation and ripple across your full 18–28V input range to ensure the TPS65261-1 input stays within spec?


    1. F29H850TU VDD POR Availability - E2E Forum
    2. TPS65261-1 Datasheet - Power Sequencing

    Best Regards,

    Zackary Fleenor

  • Hi Zackary,

    Thanks for that detailed review. I appreciate that.

    VDD rail
    "VDD (1.8V)" is a typo. Too many voltages in my head ;-). It is of course 1.25V.

    LLC output/12V rail
    As we use an additional buck converter at the LLC secondary side the 12V regulation and ripple is fine.
    Due to other restrictions VIN min will be most likely 21VDC instead of 18VDC.

    Best regards,
    Patrick