• State TI Thinks Resolved
  • Locked Locked
  • Replies 17 replies
  • Answers 1 answer
  • Subscribers 66 subscribers
  • Views 1279 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.

Original question:

LM5176: After replacing short circuit input capacitor (1 of 15), circuit no longer starts (issue caused by non-resistive...

LM5176: LM5176PWP isn't starting

DH
Intellectual 310 points
Part Number: LM5176

Hi,

I'm currently debugging a new board with the LM5176PWP and unfortunately I see an issue I can't solve. For me it looks like all conditions or voltage levels of pins connected to external circuits seems to be fine but there is no output signal. I see that soft start stays at zero, the low side MOSFETs of the input stage in constantly on, but this condition is stable without any change.

* input voltage is 48V,

* input current is small,

* output voltage nearly zero,

* both current measurements are nearly zero, I'm measuring at the output,

* EN/UVLO is about 1.6V,

* RT constantly 1.25V, there is a 62k resistor to GND

* Mode = 2V, there is a 91k resistor to GND. 

* VCC = 7.4V

The voltage at the soft start pin stays low all the time, also during startup. Do you know a reason for this behavior. No MOSFETs are switching in PWM mode, not at startup and not during constantly powered condition, only the low side gate 1 is constantly active.

The datasheet mentioned a short output protection. It's not short but I don't find much information related to this feature.

I found here at E2E a user with more or less the same issue, but unfortunately this post is locked and a can't add my question there. On the other hand, the users initial post explained the same situation, that there is no output although all pins have right signal, but the conversation ends with a discussion about output capacitors...

Maybe somebody have seen the same behavior and can give me an idea where to start with.

Thank you so much,

DH

  • 0
    TI__Mastermind 46905 points
    Hi DH,

    did you see this behavior in every system board? I want to check if there is caused by the assembling issue.
    could you also share the schematic of the circuit?
  • 0
    TI__Mastermind 40080 points

    Hi DH,

    Also please share with us your Vin range, Vo and Io so that we could check the parameters first.

  • 0 in reply to Helen chen
    Intellectual 310 points

    Hi,

    thank you Helen and Jasper for your thoughts. Vin is 48V, Vo is currently zero but designed to be controllable up to 54.6V. Io is 12A peak, but as I said, my issue is that there is currently no reaction of the IC, so the power components should be not involved.

    Yes, I have another board in the lab I will test asap. What I doesn't understand is, that the soft start pin stays low and nothing happens. Is there a reason soft start didn't rise. The only reason I see is the current limitation, but the voltage between the current measurement pins is nearly Zero. As I said, there is no output current/voltage.

    The system is controlled by microcontroller and I'm wondering if there is an error  state, the IC latches off?

    Please find the circuit attached.

    Regards,

    DH

    LM5176PWP circuit.pdf

  • 0 in reply to DH
    TI__Mastermind 40080 points

    Hi DH,

    It looks like the SS cap discharges because of something wrong triggers the protection. You could catch the SS pin waveform right after turn on to check it.

    By the way, the expert for this part is currently out of office and will address your issue after the  new year holiday.

  • 0 in reply to Helen chen
    Intellectual 310 points

    Hi Helen,

    that's exactly what I thought, but nothing happens at the SS pin, also not during startup. The voltage rise to about 100mV with En/UVLO and stays at this level. Please find a plot attached.

    Regards,

    Blue is CC and Magenta EN/UVLO. Yellow is the output voltage. All levels are constant.

  • 0 in reply to DH
    Intellectual 310 points

    Hi,

    today I checked another board and it looks like the LM5176 get broken initially at first switching on.

    I made a test and change the output voltage divider to about 27V and I solder out the boost MOSFETs. In this configuration everything works pretty well.

    Then I enabled the output MOSFETS again and the system works as well es expected. Last step was soldering the voltage divider back to 54.6V and after this step, the system went in the same issue as before.

    I didn't measure the output voltage but maybe the issue is a voltage overshoot that destroy the IC?

    There are the VIsns and the VOsns connections of the IC that are measuring the input and output voltage without a divider. TI specifies, that at high input voltages (>40V), a 2k resistor in series to VIsns is required. What's the reason for that resistor? Maybe the same resistor is necessary at VOsns?

    Regards,

  • 0 in reply to DH
    TI__Mastermind 40080 points

    Hi DH,

    The abs max voltage on the BIAS pin is 40V, and the SW1and SW2 abs max transient voltage is 65V, so 54.6V output voltage is really too high. You can check which pin is damaged......

    The expert of this part will address your issue after the new year.

  • 0 in reply to Helen chen
    Intellectual 310 points

    Hi Helen,

    thank you for your response. Vbias is connected to an external 12V supply (currently not active) and the voltage at SW1 ans SW2 is limited to <60V. I measure no voltage higher than the allowed voltages. 

    I make some quick tests with a 2k resistor in series to VOsns and I see no issue since then (~3 hours debugging with switching several times on and off). It would be great if your expert can provide me a feedback related to the VOsns pin and if a resistor at this pin as specified at VIsns will make the design work. What's the reason for the specified 2k at Visns?

    Regards,

  • 0 in reply to DH
    TI__Mastermind 21720 points
    DH,

    Regarding the schematic there are a few things that I would like to have you try with out the 2kOhm resistor on the VISNS pin

    1. Set R421 to 0Ohm
    2. Set R407 to 0Ohm
    3. Set R408 to 0Ohm

    Another think I notice is that the slope capacitor looks large at 1.8nF. Please use the LM5176 quick-start calculator to check the component values based on the application parameters.

    www.ti.com/.../snvc208

    Thanks,

    Garrett
  • 0 in reply to Garrett Roecker
    Intellectual 310 points

    Hi Garrett,

    thank you for your quick response and providing the excel file. I used the calculations in the datasheet but the excel file provides me a slope capacitor of 1.8nF as well. The reason for this high value seems to be the relatively high inductance of 18u. 

    The resistor R421 is necessary for voltage feedback. R407 and R408 are working as high side gate resistors. The general idea of placing them in the source connection is to limit switching speed and charging of the bootstrap circuit in combination. I will check if the behavior will change without this resistors.

    Switching speed is very low, do you think the 1R0 resistors may cause any problem?

    Regards,

    DH

  • 0 in reply to DH
    TI__Mastermind 21720 points
    Hi DH,

    Sorry, I didn't mean R421, I meant to type R412. Please try making this 0Ohm.

    Typically if the charge rate of the boot capacitor need to be changed the resistor is placed in series with the boot diode. This will only slow down the charging of the capacitor. If the charge and discharge of the boot capacitor needs to be changed the resistor is placed in series with the boot capacitor. This is why I am recommending shorting R407 and R408.

    When you say the switching speed is slow, what is the rise and fall times of the switch nodes?

    Thanks,

    Garrett
  • 0 in reply to Garrett Roecker
    Intellectual 310 points

    Hi Garrett,

    sure, you are absolutely right with the resistor in series to the diode. The version I used in this circuit is (at least for me) an alternative that limits the current in both paths with only one resistor.

    Sorry, switching frequency is low (~130kHz), the switching speed is comparatively fast because the TI MOSFETs we are using are quite fast as well and provides small capacitors.

    We will try changing the resistors tomorrow in the lab.

    Do you think the 2k resistor is not necessary at Vosns? Why is it necessary at Visns?

    Thank you for your great support.

    Regards,

  • 0 in reply to DH
    Intellectual 310 points

    Hi Garrett,

    Today I made the test considering your recommendation but without a positive result. What I have done was

    * Patch a board to your recommendations

    * Connect LDRV1, LDRV2, Vin and Vout to a scope

    * Switch on

    While initial powering I see that there is no over-voltage at the input nor the output. The input voltage rise to the specified 48V and the output voltage rise to about 54V before the IC fails again. After that fail, the condition is exactly as described in my first post.

    After that I took exactly the same board, change (repair) the IC and add the 2k resistor at VOsns. Since then I can switch on/off without a problem. 

    As assumed earlier, I'm pretty sure that the VOsns pin is involved in this fail situation. Can you explain me more details about this pin please. 

    * What is the functionality of this pin. The datasheet said, that it's connected to the slope compensation, but there are no further information.

    * Will a 2k resistor be a problem for the circuit? Will an additional capacitor be necessary or helpful?

    * What's the difference with this resistor?

    We noticed, that the EVM has an unmounted RC combination at this bin. What's the reason for that?

    Thank you again for you support. I know, this are a lot of questions, but it's very important to find a quick and stable solution. In worst case, we have to switch the IC soon.

    Regards, 

  • 0 in reply to DH
    Intellectual 310 points

    Hi Garrett,

    hope you are doing well.

    What's the status on your side? Do you find the time to think about our investigations and my questions above?

    Regards,

  • 0 in reply to DH
    Intellectual 310 points
    Hi Garret,

    Hope you are doing well. I need your feedback related to the 2k resistor? We don't have damaged ICs till we add this resistor but I'm sure you understand, that I need a feedback from your side to use this solution is the next design loop.

    Regards,
  • 0 in reply to DH
    TI__Mastermind 21720 points
    Hi DH,

    I looked into the design and there should be no reason at adding the 2kOhm resistor to the VOSNS pin would change the operation. This might be layout related. Adding the 2kOhm resistor might be helping to filter out noise on the VOSNS pin. Is the layout of the VOSNS pin tight? Is it way from the switch nodes?

    Thanks,

    Garrett
  • 0 in reply to Garrett Roecker
    Intellectual 310 points
    Hi Garrett,

    yes, the layout is tight and we have to made some changes, but my real issue is, that we add this resistor without many knowledge of the internal circuit of the IC and can't explain why it's fixing this problem.

    As mentioned in another post a view minutes ago, it looks like we have another issue we can't explain. Sometimes it seems that the converter is working in reverse direction. What we see is, that the input voltage (output voltage of our lab power supply) increase to values more than the voltage we set. We set 48V, the MOSFETs are designed for a voltage of 60V max and the voltage that occurs is ways higher.

    I'm not sure if it's wise to use this IC for mass production without understanding why all this problems occur. I hope you understand my concerns.