• State TI Thinks Resolved
  • Locked Locked
  • Replies 8 replies
  • Answers 2 answers
  • Subscribers 29 subscribers
  • Views 322 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

SN754410: Pulse switch failure

Liu Yang
Prodigy 10 points
Part Number: SN754410

Hello, dear experts：

We have used SN754410 and maintained the same circuit design for many years, but occasionally the chip conversion failure occurred.

We use SN754410 to convert 3.3V-LVCMOS level input pulses into 15V level PWM signals and send them to the IGBT driver card.
The input signal adopts a 3.3V open-drain circuit with a 510Ω pull-up resistor, please refer to the following circuit:

LED4_5V is actually 3.3V, and the pull-up resistor is also 510Ω.
The PWM signal in our product occasionally fails. After no-load test, it is found that the output signal of 1 of the 4 outputs is often low, but the corresponding input signal still has the expected LVCMOS level pulse (2.5KHz, 50% Duty). ). After a period of time, this Pin can resume normal 15V pulse output.
The pin with the output problem is not fixed yet, sometimes it is 2A, sometimes it is 3A.

LED4_5V is actually 3.3V, and the pull-up resistor is also 510Ω.
The PWM signal in our product occasionally fails. After no-load test, it is found that the output signal of 1 of the 4 outputs is often low, but the corresponding input signal still has the expected LVCMOS level pulse (2.5KHz, 50% Duty). ). After a period of time, the Y can resume normal 15V pulse output.
The pin with the output problem is not fixed yet, sometimes it is 2A, sometimes it is 3A.
I would like to know, under what circumstances will SN754410 fail to convert, I am not sure if this performance of the chip means that it has been damaged.
My English is not very good, thanks for the help

  • 0
    TI__Guru 56650 points

    Yang,

    "but occasionally the chip conversion failure occurred"

    Did that happened on certain boards? Or could be different board at different time? After the issue happened, did it recover?

    BTW, I didn't see any risk for SN754410. Do you have good board and bad board? Can you swap the device between the goad board and bad board? 

    Regards,

    Wang Li

  • 0 in reply to Wang5577
    Prodigy 10 points

    Li，Thanks for your reply！

    There are three types of our circuit boards, and 754410 are used in these different boards. At present, similar problems have occurred in each board.
    There is a DIP socket on our board. When the control system detects the IGBT overcurrent, it is finally found that the 754410 has sent the wrong PWM. At that time, we replaced it directly, and the new 754410 performed well.
    The following are photos of two boards. One of the boards, we deliberately removed the DIP socket, the test found that the chip is still unstable.

    Yesterday’s test result was that 2Y can recover the pulse by itself. After powering on for one night, I found that 2Y has continued to output low level. The 2A input 3.3V pulse is still correct. Even if the power is turned off and on repeatedly, the 2Y output is still 0V.

    The temperature of the chip noticed yesterday does not seem to be high because there is no load.

    Now I suspect that it may be due to some system-level reasons in the product. For example, 754410 is overloaded for a short time, or other possibilities, the chip has been slightly damaged, the work is no longer stable, and over time, it will cause complete damage.

    But I am not sure whether this assumption holds. If it is the internal thermal protection of the chip, why is only one of the pins output incorrectly, but the other pins output normal.

    We have encountered several problems now, all of which were on the customer site. In the factory test, the impact of environment and time was not obvious. Now we can’t locate the cause, and we can only choose to temporarily replace a new one.

    Regards,

    Liu Yang

  • 0 in reply to Liu Yang
    TI__Guru 56650 points

    Yan,

    Thanks for more detail info.

    "At present, similar problems have occurred in each model"; "the new 754410 performed well." and "only one of the pins output was incorrectly" 

    Let us confirm the device damage first. After 2Y output is 0V, would you measure the 2Y pin output resistance and compare it with a normal device.

    Were all the damages devices failed on 2Y output on three different models?

    Can we increase VCC2 capacitor and add 0.1uF on VCC1? And then, monitor the failure rate.

    Regards,

    Wang Li

  • 0 in reply to Wang5577
    Prodigy 10 points

    Li，Sorry for the late reply message because of the weekend.And thank you very much for your tips.

    I only have two boards that clearly have problems at the moment, which is exactly the test process mentioned last week. And 2Y is not shorted to 0V. We compared the good board and the bad board. The resistance range of 2Y to 0V is 4.5MΩ~5MΩ.

    I am sure that the previous board does have 2Y output damage. 2Y corresponds to the drive of the U-phase lower half-bridge, but we have also found that the V-phase half-bridge drive (4Y) is also problematic, so we can only say that there is a high probability that there will be a problem with 2Y. , It needs further attention.

    0.1uF capacitors are always connected near VCC1 and VCC2. I noticed very early that 1uF bypass capacitors are recommended for VCC2, but I use 10uF capacitors near the output pin of the 24V/15V switching power supply (VCC2 power supply).

    Now the 2Y should be completely damaged. I think it has little to do with the capacitor. After all, the 0.1uF design has been in use. But I can try to change to 1uF. Please tell me, why do you suspect that the problem is caused by the capacitor?

    • "Place device near the load to keep output traces short to reduce EMI. "
    • "These devices have limited built-in ESD protection. The leads should be shorted together or the device placed in conductive foam during storage or handling to prevent electrostatic damage to the MOS gates. "
    • "In the thermal shutdown mode, the output is in a high-impedance state regardless of the input levels. "

    The only factors that can be found in the data sheet are the above, and it does not explain the specific failure behavior of the chip.I have to keep doing some tests.

    Regards,

    Liu Yang

  • 0 in reply to Liu Yang
    TI__Guru 56650 points

    Yang,

    If we confirmed the 2Y damage, we can send few samples back to do failure analysis to see what internal circuit gets damage.

    VCC2 would have a spike current when the 2Y is from low to high. The VCC2 capacitor charges the 2Y output. When the current is small, we may not see a high voltage spike on VCC2 pin due to the parasitic inductance between the input 10uF capacitor and VCC2 pin. But, we got damage and VCC2 capacitor is only difference from the reference circuit, I want to double check it.

    Regards,

    Wang Li

  • 0 in reply to Wang5577
    Prodigy 10 points

    Li，

    Let's try to test the 1uF capacitor first, and find other problematic SN754410 for further testing, and then select a few chips and send it to you?
    Can I ask how to send it to you?

    Regards,

    Liu Yang

  • 0 in reply to Liu Yang
    TI__Guru 56650 points

    Yang,

    I will send you an email to talk about FA job.