LNK623DG IC Burst Issue - Require Technical Support to root cause the issue
Greeting from Silvan Innovation Labs. We have been using the LNK623DG IC in many of our Home Automation products to convert 230V AC (input) to DC 12V 300mA ( 3W output). We have prototyped the design and results are satisfactory which leads to production. we have been facing LNK623DG IC Failure in production boards in which the IC's are behaved differently and have the following issues
1. The Output of the SMPS generated to 50V which leads to damage of DC side components as they are operating at 12V.
2. The SMPS burst with no trace from the PCB after few days of continuous operation in our client place.
Once the IC' are replaced with new ones, the boards are working back to normal.
As we are going for next production slot, We need your Quick response - Technical support & guidance to root cause the issues.kindly Note that all failed ICs belong to batch 1430.
Attached Schematic and Transformer winding details for Reference.
Comments
Can you share the Screen shots of your design verification(DCON Violation), which will help me to understand the solftware tool completely.
Could you please let me know the calculation for finding the conduction time for the diode. We are using the part: B1100-13-F (Datasheet Attached for your reference). kindly let me know whether the implemented part need to be changed or suggest us alternate SMD diode - SMA package
I have a doubts related to the schematics attached.
---->We have not used the snubber components for the output diode - Will it cause any issue in operation?
----> We have not implemented the EMC filter at the input and output stages - Will it create any issues.
Kindly let me know if anything i need to do debug the Design.
| Attachment | Size |
|---|---|
| diodes inc b1100 13 f.pdf (154.16 KB) | 154.16 KB |
Hello Hemaraj T,
the spreadsheet screenshoot is attached. The calculation you will find it in the application note of LNK-CV. the link is below:
https://ac-dc.power.com/sites/default/files/product-docs/an45.pdf
I think you should adjust your design based on the suggestion of the PIxls. Verify your design in the PIxls to see where is teh violation, below is link for PIxls:
https://piexpertonline.power.com/site/login#
---->We have not used the snubber components for the output diode - Will it cause any issue in operation?
If you don't have snubber circuit for the output diode, the spike during the turning off of the diode will be high and it will create more loss on the diode.
You should check your shottkey diode waveform to make sure the selection of your diode can stand the big spike. You also need to check the efficiecy and thermal performanc of the output diode to see if you can accept it.
----> We have not implemented the EMC filter at the input and output stages - Will it create any issues.
You should check your EMC performance. if it meets your requirement, then it should be OK. If not, you should try with those filters.
Best Regards
| Attachment | Size |
|---|---|
| Spreadsheet_LNK-CV.png (368.29 KB) | 368.29 KB |
Hello Hemaraj T,
Thank you for considering using Power Integrations Products.
I run your design in our PIxls tool (below is the link, it is free), the best way to verify the design to see if anything voilates. i found the DCON is violated and it suggested a bigger device.
https://piexpertonline.power.com/site/login#
Output Diode Conduction Time, DCON (ms)
This is the calculated diode conduction time of the output diode. This value should always be greater than 3.1 ms, the point at which the voltage on the FEEDBACK pin of LinkSwitch-CV is sampled
The power table in the front page of datasheet is based on 5V design. if you want to use for 12V design, the best and simplest way is to use the PIxls to confirm the design to see if there is any violations.
Best Regards