• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 2 answers
  • Subscribers 63 subscribers
  • Views 383 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.

LMG3410R070: High side transistors of a full bridge indicating over current protection under no-load condition

Marcel Kossek
Prodigy 70 points
Part Number: LMG3410R070
Other Parts Discussed in Thread: ISO7762

We recognised an interesting fault occurrence while testing a full bridge converter with the LMG3410 under no-load condition. 

The trip point of the fault indicated by the FAULT Pin is frequency und slew rate dependent. For low frequencies < 250kHz the fault never occurs. For a frequency ~250kHz the high side transistor of the each half bridge shuts down after about 1 minute of operation and a fault is indicated. For higher frequencies tested up to 500kHz the operation time to failure decreases rapidly to a few seconds. For steep slew rates this time decreases as well which sounds like an EMI problem. But for 500kHz and flat slew rates about 20V/ns the fault is trigger as well. Therefore EMI becomes more implausible.

Because the fault can be reset by interrupting the PWM signal, i guess the fault conduction is due to over current protection as mentioned in the data sheet.

Do you have any idea why the OCP fault is triggered and how to solve it?

  • 0
    TI__Expert 3225 points

    Hello Marcel,

    Thanks for asking. Yes it sounds like that you have met an EMI issue. There can be a OCP signal which interrupt the operation of your circuit. 

    1. The OCP signal could be fake, which means that the signal is generated by noises typically from common mode interference. In this case, adding some filter capacitors along the fault signal lines would be helpful. RC types of filter is a good option.

    2. There could be chances that when you have high dv/dt, the input PWM signal is contaminated, which would lead to a shoot-through case. Please check the common mode rejection capability and any path for CM noises. For example, the CM noise can be transmitted across your isolator/optocoupler, or the isolated power supply, and cause certain ground bouncing. Please make sure you have enough CM chokes to filter out the noise. 

    Hope it helps. Thanks!

    Best regards,

    Lixing

  • 0 in reply to Lixing Fu
    Prodigy 70 points
    Hi Lixing,

    thanks for your answer!
    Refering to:

    1. I thought the fault-Pin is a push/pull output. How can noise from the outside influence the behaviour inside the LMG3410 to trip a fault? I did not test this method, but i will do so soon.

    2. I chose the NXE2 (Cio = 2pF)as isolated DC/DC converter for the LMG3410 and the ISO7762 (Cio = 0,9pF) for signal isolation with relativly low input to output capacitances. I added CM chokes on the input side of the NXE2s, but with no change in fault occurrence.

    After some more tests I can say that the fault is dependent on the input voltage and the switching frequency. Influence of the slew rate is very small, so the fault occurs for 300kHz and 400V with a slew rate of 90V/ns as fast as for 20V/ns.

    kind regards Marcel
  • 0 in reply to Marcel Kossek
    TI__Expert 3225 points
    Hi Marcel,

    Thanks for the follow up. Please see my answers as below:

    1. Yes the fault-pin is a CMOS push/pull output. This external influence cannot affect the internal operation. but just in case you have external protection from your controller. If not, you should be fine on this.

    2. I did check the datasheet of both parts that you used. Yes they are very low in capacitance and should not be a major factor to affect your EMI in the case of low dv/dt like 20V/ns.

    3. We have run the devices at hard switching with high frequency as well, but have not observed this issue yet. Sounds like it may be something related to layout or design. Please make sure that you have your layout, especially your Vneg capacitor, VDD capacitor as close as possible to the pins. Also could you please share your running condition like current on the devices?

    Thanks and regards,
    Lixing