• State TI Thinks Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 2 answers
  • Subscribers 15 subscribers
  • Views 241 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.

SN6505A: D1/D2 voltage distortion and SN6505A protection

kylin z
Prodigy 10 points
Part Number: SN6505A

Hi TI team,

We use SN6505A to make a isolation power. 

There are some abnormal boards showed the power supply is restarted in cycle.

The board works normally at the beginning and after several hours when the temperature is high(Actually the temperature is about 40℃) it works abnormally.

CH1-D1, CH4- power supply input current

Does this mean the SN6505A protected?

I checked the datasheet, there are 2 protections: 1.7A current limit and thermal shutdown.

Which protection acts here? The SN6505A’s case temperature is about 91.6℃ under abnormal situation.

The EN pin is always high.

And we find that the voltage for D1/D2 is distortion.

CH1- D1, CH4- Current to D1

Is this high current caused by the transformer saturation or the transformer primary inductance and parasitic capacitance oscillation?

It seems that the high current occurs when the MOSFET is turned off.

The SN6505A's deadtime can be affected?

I infer the temperature causes the transformer's feature changed.

Looking forward to your reply.

Thanks.

Kylin

  • 0
    TI__Guru 52080 points

    Hi Kylin,

    Welcome to TI E2E forum and thank you for sharing waveform related to your question, this is helpful.

    Your understanding is correct, the device has primarily two protection features - over-current and thermal shutdown. Over-current is a cycle-by-cycle protection where device detects over-current at the beginning of switching cycle and then turns OFF the device for that switching cycle. It continuous switching normal from next cycle and repeats the process each cycle.

    From the waveform you have shared, it looks like device is heating up and is hitting thermal shutdown limit periodically.
    As you also pointed out, the current goes upto 4A when the switch is being turned OFF. This is possibly due to transformer leakage inductance or incorrect design of the transformer. Please share the transformer part number / datasheet so that I can review and confirm if transformer is okay. Thanks.


    Regards,
    Koteshwar Rao

  • 0 in reply to Koteshwar Rao
    Prodigy 10 points

    Hi Koteshwar,

    Thanks for your reply. You're right that the transformer is mismatched. It is WE's 760390013. The VT value is low.

    I'm just analyzing this situation to see if it is because of the saturation or other thing.

    I find that when the temperature is high, the transformer current and D1/D2 voltage become distortion.

    So is that because the transformer's feature is affected by temperature? 

    Could you explain the failure mechanism?

    Thanks.

    Best regards.

    Kylin

  • 0 in reply to kylin z
    TI__Guru 52080 points

    Hi Kylin,

    Thank you for confirming the transformer part number and that it is not meeting SN6505A V-t product requirement.

    Based on this, I would expect the transformer to saturate and act as a short. This leads to a large current through transformer and device which the device detects and turns off switching for that cycle. Since the peak current is much higher than normal current, the device and transformer will heat up.

    Since the heat takes times to cool down, continuous operation of device in this mode could make the temperature of device to exceed thermal shutdown threshold and make it to shutdown till it cools down.

    The waveform you have shared show both these actions. Peak current indicating transformer saturation, short duration current pulse at the beginning of each cycle indicating cycle-by-cycle over-current limit protection working and periodic turn off every few 100ms indicating thermal shutdown.

    Let me know if you have any further questions, thanks.


    Regards,
    Koteshwar Rao

  • 0 in reply to Koteshwar Rao
    Prodigy 10 points

    Hi Koteshwar,

    Thanks for your explain.

    One more question about " short duration current pulse at the beginning of each cycle indicating cycle-by-cycle over-current limit protection working".

    In my opinion the current should achieve the over current limitation threshold at MOSFET on period then the SN6505A can turn off the MOSFET accordingly.

    But for SN6505A the current limit level is about 1.7A. And the current at the point which MOSFET starts to turn off is less than 400mA. 

    Why does SN6505A protect so early?

    Would you please help to explain?

    Thanks. 

    Best regards.

    Kylin

  • 0 in reply to kylin z
    TI__Guru 52080 points

    Hi Kylin,

    Please see my inputs below, thanks.

    kylin z said:
    In my opinion the current should achieve the over current limitation threshold at MOSFET on period then the SN6505A can turn off the MOSFET accordingly.

    I do agree and expect the same as well. Could you please confirm again that the monitored voltage is D1 and the current is the current going into D1?
    Could you please share the schematic and indicate where these two in the schematic are measured and what were used to measure?
    Please also show an image of the test setup showing voltage / current monitoring connections to D1.

    I would also like to mention that when transformer is saturated, the behavior is not predictable.

    kylin z said:
    But for SN6505A the current limit level is about 1.7A. And the current at the point which MOSFET starts to turn off is less than 400mA. 

    Please note that over-current limit typically doesn't trigger exactly at the moment when current limit threshold is hit. To avoid false triggering, over-current is monitored for a small duration to confirm this is not a false noise spike and is actual current. This duration seems to be 1µs in your waveform.


    Regards,
    Koteshwar Rao

  • 0 in reply to Koteshwar Rao
    Prodigy 10 points

    Hi Koteshwar,

     About the voltage and current measurement I use KEYSIGHT oscilloscope DSOX4104A, voltage probe N2873A and current probe 1147B.

    And the position shows below: So actually the voltage is transformer pin1.

      

    schematic:

    Best regards.

    Kylin

  • 0 in reply to kylin z
    TI__Guru 52080 points

    Hi Kylin,

    Thank you for sharing the test setup image and confirming that the current monitored is current going into D2 from pin 1 of transformer. I am also assuming that the voltage monitored was D1 which aligns with the expected operation / waveform.

    Let me know if you have any other questions, thanks.


    Regards,
    Koteshwar Rao