XTR117: Failing Surge Immunity test where TI's TIPD126 design passes.

Hi Yash,

the SMBJ40CA might be too weak to withstand Surge. Try the 1.5KE36A.

Kai

Thank you for the suggestion. Can you help us with some resource to calculate the right wattage for the TVS diode?

Also how SMBJ40CA worked in TI's testing.

Hi Yash,

Your schematic is a little different from the reference design so I will look into this. Is there any way you can send you PCB board layout with us to review as well? 

Katlynne Jones said:

Hi Yash,

Your schematic is a little different from the reference design so I will look into this. Is there any way you can send you PCB board layout with us to review as well? 

Hey Katlynne,

I can share the PCB layout files. Can you confirm the format you need it in? Also, where should I share it?

Hi Yash,

these two documents explain a lot:

6835.4_Transient_immunity_testing.pdf

SA3-37.pdf

Read them very carefully :-)

Kai

Hi Yash,

Kai's recommendations for a higher rated TVS diode should help, but I can still take a look at your layout to see if we can figure out why the TIPD126 part selections did not work for you. You can share your Gerber files or a screenshot like this:

You can share it here or send me a connect request in e2e so you can share them privately. 

Though I get the layout of the board is very different but I am not sure how that effects the surge protection test. And more importantly I find just doing a basic calculation on the selection of the TVS diode the SMBJ40CA doesn't quite makes sense. I have following issues.

1. SMBJ40CA has a clamping voltage of 64.5V and XTR117 has absolute maximum rating of 50V, even after considering the meager 0.7V drop due to reverse protection diode the voltage is just too high for XTR117.

2. According to IEC 61000-4-5 we the EUT is subjected to 1kV peak voltage with 42 Ohm (40 Ohm external + 2 Ohm power supply impedance) effective output impedance for non power line devices. Here we get a maximum peak current of 1000 V / 42 Ohm = 23.8 A where as the selected diode SMBJ40CA has a maximum peak current 9.3 A.

I find this method of calculation in accordance to our test, as 500V were applied to single SMBJ40CA it didn't survive because 500V/42Ohm = 11.9A which is higher than diodes peak current but two diodes survived the 500V test as combined pair they had approximately double the current capacity of 18.6A.