• State TI Thinks Resolved
  • Locked Locked
  • Replies 11 replies
  • Answers 2 answers
  • Subscribers 63 subscribers
  • Views 2271 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.

TPS2420 - Overload Protection or Current Limiting?

Stephen Noel
Prodigy 70 points
Other Parts Discussed in Thread: TPS2420

I am using the TPS2420 eval board to verify operation with another PCB designed in-house. This system needs to operate over voltage ranges of 6-8.5 V. I find that my board/load is only able to power on via the TPS2420 from 6-7.7V, while the range 7.7-8.5V results in a FLT signal enable and the chip goes into reset/restart mode. If I start the voltage lower than 7.7V and ramp up my power supply rail connected to the Vin input of the TPS2420, the chip performs fine. Therefore I suspect the TPS2420 is faulting into either the Hard Overload - Fast Trip or the Overcurrent Protection state. My readings of Vin and Imon do not match any of the application plots in the chip datasheet, so it is difficult for me to tell which mode my load is driving the TPS2420 into. 

The steady-state current draw of this board is ~0.6A with operational current draw of up to 3A. 

The following plot shows a few signals at the vin=8.5V case, which resulted in a fault. 

My settings on the eval board are as follows:

C_CT 1uF -> 38.9ms
I_MAX 40.2kOhm -> 5A
I_FLT 50kOhm -> ~4A 

What is the fault mode the TPS2420 is going into? Does the chip read the in-rush current at higher Vin voltages as a a "real, continuous failure" as presented in the datasheet? What can I change on my board/load or, preferably on the TPS2420 surrounding circuit to mitigate this problem?

Edit: Here is a long duration plot of the 8.5V Vin case.

Edit #2: We decided to short the CT capacitor pad in order to bypass the fault detection capability. In addition, we removed our in-line RF bead from the power rail to decrease any inductive loading. We then tested the turn on over the whole range of Vin's. Below are the results for Vin and Imon across the Vin range. With the CT capacitor shorted, the instrument is able to power on within 1050 ms. The cases which resulted in a fault prior to this test take 1050ms to reach Vout=Vin, while the successful "PG" cases from before (<7.6V) reach this state immediately. 

  • 0
    TI__Guru** 105361 points
    Hi Stephen,

    Your parameter settings looks correct. Please let me know the following details.
    what is the load resistance and output capacitance values?
    can you send me the Vin, Vout and Iout waveforms for long interval say 300ms

    Best Regards,
    Rakesh
  • 0 in reply to Rakesh
    Prodigy 70 points

    By measuring the resistance and capacitance across the rails of the load: 413 uF and 1.37 kOhms. 

    I will send you that plot shortly. 

  • 0 in reply to Stephen Noel
    TI__Guru** 105361 points
    Ok Stephen.
  • 0 in reply to Rakesh
    Prodigy 70 points
    Sorry for the delay; I included the plot in my original post above.
  • 0 in reply to Stephen Noel
    TI__Guru** 105361 points
    Hi Stephen,

    It looks like overcurrent shutdown. The device is ON for fault timer period of around 40ms and then turned off.
    You can check whether this issue repeats with out any load.
    Also monitoring actual current Iout instead of Imon helps in debugging the issue.

    Regards,
    Rakesh
  • 0 in reply to Rakesh
    Prodigy 70 points

    Rakesh,

    The issue doesn't occur without the load on the eval board.

    Can you explain what the chip is doing in the first 2ms and then until the 40ms shutdown?

    Unfortunately we do not have current measurement probes in our lab, so I can't get you that data. 

  • 0 in reply to Stephen Noel
    TI__Guru** 105361 points
    Stephen,

    The behavior is load dependent. In your case, when the device is enabled, it sees current more than the programmed limit. Hence, the control unit starts the fault timer. If the load current remains more than programmed limit for the duration of fault timer 40ms, the device treats as overcurrent and shutdown.

    Regards,
    Rakesh
  • 0 in reply to Rakesh
    Prodigy 70 points

    Is the device also current-limiting the load, as the Vout data seems to show? It's difficult for me to imagine our board pulling 5+ amps for 40ms+. Previous testing (prior to integration with the TPS2420) showed a large in-rush current spike lasting for less than 2ms and then relatively low current draw after that. 

    What changes can we make from this point on the TPS2420 circuit? What about our load? I already tried decreasing the load capacitance by 20% without any noticeable change in the shutdown behavior of the TPS2420.

  • 0 in reply to Stephen Noel
    TI__Guru** 105361 points
    Stephen,

    The device provides circuit breaker type of protection and has not current limiting.
    What kind of load are you using for testing? If it is electronic load, use it in constant resistance mode.
    Can you resolder the device, just to make sure that power pad is soldered properly.
  • 0 in reply to Rakesh
    Prodigy 70 points
    Rakesh,

    Why does Vout drop to ~1.5V in the fault cases then? The datasheet does not show a Vout voltage drop (Figure 23) associated with Overcurrent Shutdown. Why is this happening in our case?

    I am not testing the TPS2420 board with an electronic load, all of my data is from the actual load.

    I don't have the capability to re-solder the chip at my lab.

    I attached some additional plots to my post under "Edit #2":

    We decided to short the CT capacitor pad in order to bypass the fault detection capability. In addition, we removed our in-line RF bead from the power rail to decrease any inductive loading. We then tested the turn on over the whole range of Vin's. Below are the results for Vin and Imon across the Vin range. With the CT capacitor shorted, the instrument is able to power on within 1050 ms. The cases which resulted in a fault prior to this test take 1050ms to reach Vout=Vin, while the successful "PG" cases from before (<7.6V) reach this state immediately.

    A few questions:

    1. Does this give you any more insight into what is happening between our load and the TPS2420 board or changes we can make on the TPS2420 sub-circuit or our load?

    2. Could we just use a larger CT capacitor (2.7uF) to ensure our load can power on at higher Vins? What risk are we taking on in this case?
  • 0 in reply to Stephen Noel
    TI__Guru** 105361 points
    Hi Stephen,

    In fault case, the output voltage drop depends on the amount of overload.
    From your test waveforms, it looks the device is entering into power limit mode above 7.7V Vin. The output is around 1V for 1000ms long interval. Unless the add-on PCB is loading excessively, the output should ramp.

    what exactly is the circuitry on PCB designed in-house. Can you provide V-I characteristics of the board (by powering from bench power supply).

    Regards,
    Rakesh