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.

TPS544C25: TPS544C25 over current

Part Number: TPS544C25

Hello Team,

I do see some TPS544C25  triggering the over current events.

The current I can measure over PMBus is in the range of 20-24A while over current is set to 40A.

Thus from "steady" current I do not see the over current.

PMBus bandwidth does not allow to measure/detect peak as fast as a comparator would do, so I'm not surprised.

Nevertheless since the steady current is below the OC threshold it meas that is a current peak that trigger the OC event.

The questions:

  • Are there ways to measure the current peak that are seen by the OC block?

I do not see anything going out

  • What would you recommend to measure the current peak?
  • Are there any internal settings/register to enable some delay to allow short peak? Detection window or similar?

only what would last longer would be detected as over current.

Thanks,

SunSet

  •  

    The TPS544C25 uses the voltage drop on the low-side FET during the low-side FET conduction time (often called the OFF-time) to measure the average inductor current for both Telemetry and OC fault.  While there is no way to measure the exact voltage seen by the TPS544C25, you can approximate it.

    1) Direct measurement of the SW node during the low-side FET ON-time.

    While you would need to calibrate the measurement to the Rdson of the low-side FET, measuring the voltage from SW to PGND using a differential probe can provide a measurement of the current limit detected by the TPS544C25's current sense circuit.  The internal circuit will sample the SW - PGND voltage during the first half of the low-side FET on-time, then latch the voltage at the mid-point for the rest of the switching cycle.  This held value is then passed to the ADC and current limit circuitry to obtain the average current.

    2) Direct measurement of the Inductor current

    Adding a small lead in series with the output terminal of the inductor, it is possible to take a direct measurement of the inductor current, again measuring the current at the mid-point of the low-side ON time for the current measurement.

    3) R-C Observer measurement

    Adding an R-C across the inductor it is possible to use the inductor's DCR as a sense element.  Connect the resistor to the inductor's SW terminal and a capacitor to the inductor's VOUT terminal, and the differential voltage on the capacitor will be proportional to the inductor ripple current.

    At low frequencies, the capacitor voltage will exactly equal the average current times the DCR of the inductor.  To get the pulse by pulse AC current to match as well, you need to size the RC time constant to be equal to the L/DCR time constant of the inductor.  Select a convenient capacitor value, such as 100nF, and calculate the resistance needed so R = L / (DCR x C)

    With that in place, a differential measurement of the voltage across the capacitor will represent the AC ripple current in the inductor, assuming the inductor does not start to saturate.

    Unfortunately, the TPS544C25 does not support IOUT_OC_FAULT_RESPONSE settings where an OC fault is ignored for a time.

    The other thing to consider is that it is not the low-side FET OC fault that is getting triggered, but the high-side current limit.  To sense the current in the high-side FET, the TPS544C25 compares the SW voltage during the high-side FET "ON" time to the VDD pin voltage.  If the VDD pin is connected to VIN away from the TPS544C25, or has an R-C filter between VDD and VIN, the AC ripple on the input voltage during the ON-time can cause the high-side OC to trigger prematurely due to the external drop between VDD and VIN.

    Check the schematic and layout.  If there is a filter resistor between VDD and VIN, measure the AC ripple voltage on VIN and the current flowing into the VDD pin.  Increase the VDD filter resistor so that the drop from VIN to VDD due to the DC current into VDD is greater than the AC ripple observed on VIN to offset any error to the high-side short-circuit protection.

  • Many thanks.

    Very exhaustive.

    Regards,

    SunSet

  •  

    Glad to be of assistance.  I will close this thread now.  If you have any additional questions, please feel free to post a new thread.