Hi,

Thank you for using power integrations product.
To better understand what is causing the problem
Can you provide is the following:

* PI Expert design file (.uds) - If you go to PI Expert File Manager there is a share column
you can provide me the link

* Trasformer Design

* PCB Files
* Picture cleary showing the top and bottom components
* BOM
* Magnetics design (core type, turns ratio, etc)
* Switching waveforms

Question:
* on the image you provided , is there a short on the optocoupler (Pins 1 and 2)?
* is this CV and CC operation or just CV operation?
* Before you dispatched the units what sort of testing you were able to perform? Short Test?
* after the blow up, are there any other components that were damage? is it on primary side or secondary side?
* When the part blew up, can you describe the actual testing operation? is it 265VAC? Full load (1.17A?)

few recommendations
* I think you can start by checking the components rating
there might be components that are over stressed

Hello,
Please find attached uds file.
Could you please firstly inform me about my question please? Is this PI-Expert created schematic correct or not? I just download new version of PI Expert and created same project but it added to schematic a varistor, 100nF/250VAC to input and shifted 6mH filter before bridge and used 1 pcs 33uF/400V capacitor instead of 10+22uF/400V capacitors in new schematic.

My question is do you have another reference design or I must rely on this newly created PI Expert schematic to prevent such a problem? Please consider I used all declared quality/values components and transformer constructed as PIExpert defined in transformer construction.

Answers:
* on the image you provided , is there a short on the optocoupler (Pins 1 and 2)?
Interesting question, if there is shortcut how this circuit will work? It is 1K resistor as defined. I wrote this project worked in the field for a several weeks and then it happened.

* is this CV and CC operation or just CV operation?
CV Operation

* Before you dispatched the units what sort of testing you were able to perform?
Tst before dispatch 5sec full load run.

Short Test?
Yes

* after the blow up, are there any other components that were damage?
No

* When the part blew up, can you describe the actual testing operation? is it 265VAC? Full load (1.17A?)
It happened on field, it was not testing operation.

* I think you can start by checking the components rating
there might be components that are over stressed
All component has enough Up limits ratings.

Hi ,

the file you attached is .csv file. I'm actually looking for .uds file.
The varistor is mainly used to protect the circuit during line surge events by
clamping the input voltage.

The 100nF/250VAC to input and the inductor 6mH serves as filter for reducing EMI.

Please send me the PI expert uds file so I can help you better

please find attached file

Hi, 

Can you also send me the PCB files.

By the way is the blow-up reproducible?

Can you enumerate what is country A or Country B, some countries are 115VAC some are 230VAC.

Did you follow all the BOM recommendations?

Hi,

All countries 230VAC.

Blow-up not reproducible

I cant send you PCB files.

Lets go step by step, how about firstly send me your last recommended reference design schematic?

If you read my messages you could see that I wrote I did follow all BOM recommendations and take this issue as all components are OK.



I want from PI will assign to this issue another expert! So long time passed, so late replies and not concentrated to the job.

Hi ilkintacan,

I have seen the UDS file and I am not seeing anything wrong with the design. Bulk of the changes introduced on the newer PIExpert version are for EMI only as mentioned by PI-Phlox. However, the additional varistor may matter. What I can deduce is Country B might have more instability in their line voltages i.e. higher and/or more frequent line surge occurrences.

Considering that only the TNY290 (and fuse?) was damaged for the PSU, it could be that the breakdown voltage rating was exceeded. This may damage the IC and impair its operation which might increase output voltage and damage the load. If differential surge is the issue, then a varistor would definitely help. The likelihood of this occurring (damaging the load) however is very low since the output voltage usually ramps down immediately when the device is damaged due to differential surge.

It would then be most likely that the issue is related to common mode surge/noise and is associated with the PSU/load setup i.e. grounding, heatsinking, shielding, etc. plus the presence of input line instabilities.

Thus I would recommend testing both differential and common mode surge in accordance to ANSI/IEC standards for the PSU to have more immunity since it seems that line instabilities and/or setup related issues is the culprit and the issue is only present in one country.

Regards,

Thank you for your enlightening reply.

Best Regards