STEVAL-DPSLLCK1 / STEVAL-DPSLLCP1 faults during dynamic load steps, 48 V output droops to 0 V while 400 V DC bus remains stable

[H.Joshi]

Hello ST Community,

I am testing the STEVAL-DPSLLCK1 3 kW full-bridge LLC DC-DC converter kit. The power board is marked STEVAL-DPSLLCP1, and I am using the STSW-DPSLLCK1 firmware.

The LLC DC-DC converter is being supplied from a STEVAL-DPSTPFC1 PFC board. The PFC stage provides a stable DC bus of approximately 400 V to the LLC input.

The converter starts normally and regulates to 48 V output at no load. After the converter is already running at 48 V, I apply dynamic load steps using an electronic load.

The issue is that under some dynamic load conditions, the DC-DC converter faults or the 48 V output voltage collapses to 0 V. The steady-state load current is well below the rated output current of the board, but the fault appears to happen during the transient load change.

My test conditions are:

• Input DC bus to LLC: approximately 400 V
• Output voltage: 48 V
• Load type: electronic load in dynamic load mode
• Converter starts at 48 V with no load, then dynamic load is applied

The dynamic load tests I performed are:

1. 7 A for 200 ms, then 27 A for 800 ms
   • Result: the converter trips with DC-DC overcurrent faults.
2. 10 A for 200 ms, then 25 A for 800 ms
   • Result: this test runs as expected.
3. 9 A to 25 A, with 200 ms / 800 ms timing
   • Result: the converter output voltage droops all the way to 0 V.
4. 9 A to 25 A, with slower 800 ms / 1800 ms timing
   • Result: the converter runs for a couple of minutes, but it is still not fully reliable.

During these fault events, the 400 V DC bus remains stable. The STEVAL-DPSTPFC1 PFC stage does not appear to fault, shut down, or collapse. The problem seems to be on the LLC DC-DC converter side, because the 48 V output collapses while the 400 V input bus remains present.

Previously observed fault codes from the fault LED blink pattern were:

• 0x0010
• 0x0020

From the STSW-DPSLLCK1 documentation, these appear to correspond to DC-DC overcurrent protections, including resonant current overcurrent and output overcurrent.

However, during the most recent failure with the 9 A to 25 A dynamic load step, the fault LED did not blink with a frequency/count pattern. Instead, the fault LED stayed continuously ON for the whole time.

My questions are:

1. What does a solid ON fault LED with no blink pattern indicate on the STEVAL-DPSLLCK1 / STSW-DPSLLCK1 firmware?
2. Since the 400 V DC bus remains stable, what condition would cause the LLC firmware or gate-drive stage to stop switching and allow the 48 V output to collapse to 0 V?
3. Could this be caused by a protection mechanism other than the documented overcurrent blink-code faults?
4. Is the converter entering a latch-off state, firmware lockup, gate-drive shutdown, brownout of an auxiliary/control supply, or another internal protection mode?
5. For dynamic load steps such as 9 A to 25 A or 7 A to 27 A, is there a recommended minimum load-step slew rate, transition time, or timing requirement for this evaluation board?
6. Are there firmware parameters that should be tuned for dynamic load operation, such as:
   • resonant overcurrent threshold,
   • output overcurrent threshold,
   • voltage loop compensation,
   • startup/ramp parameters,
   • burst mode settings,
   • synchronous rectifier timing,
   • load transient response parameters?
7. Is it safe or recommended to modify the overcurrent thresholds in firmware, or should the issue first be solved through control-loop tuning or electronic-load slew-rate adjustment?
8. What measurements should I capture to identify the root cause? For example:
   • 400 V PFC DC bus voltage during the load step,
   • LLC resonant current waveform,
   • 48 V output voltage transient,
   • output current waveform,
   • electronic-load slew rate,
   • auxiliary supply rails,
   • gate-drive signals,
   • MCU fault status at the moment of failure.

My goal is to make the board operate continuously without faulting during these dynamic load steps. Since the steady-state load current is below the converter rating and the 400 V DC bus remains stable, I would like to understand whether the issue is due to transient resonant current, output-current sensing, firmware protection thresholds, control-loop response, burst-mode transition, synchronous rectifier behavior, or another protection mechanism.

Any guidance on how to debug this issue, what the solid fault LED means, and which firmware or hardware parameters should be checked would be greatly appreciated.

Thank you.