TPS22810: Expected part lifetime

OK after looking through the related posts I found this one:

https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/1090487/tps22810-number-of-switches-in-lifetime-before-failure

and by that this information on MTBF:

https://www.ti.com/quality/docs/estimator.tsp?opn=TPS22810TDBVRQ1&cpn=&partNumber=TPS22810#resultstable

However, no further information on this MTBF of 1 * 10^10 hours. It is given independent of the number of switching cycles.

Is there any information on the applied testing method, i.e. how many cycles were done during the 1000h testing period...?

Best,

Michael

Michael, 

I am pinging our backend team to get some clarity here on exactly how the MTBF number is calculated. Related, but if we provided a functional safety document similar to the following:

https://www.ti.com/document-viewer/lit/html/slvaek2

Would this help at all?

Best Regards,
Tim

Michael,

The following document describes the MTBF number a bit more in detail (this is for oscillators, however the information is relevant to our devices too):

https://www.ti.com/lit/an/snaa391/snaa391.pdf?ts=1716908963447&ref_url=https%253A%252F%252Fwww.ti.com%252Fsitesearch%252Fen-us%252Fdocs%252Funiversalsearch.tsp%253FlangPref%253Den-US

Specifically:

Specifically, the following JEDEC standard is followed:

https://www.jedec.org/standards-documents/docs/jesd-85

Our exact method for HTOL (as described above) is considered proprietary, however the design of the device is examined and HTOL is run in a way that stresses the device in its most extreme operating conditions.

Best Regards,
Tim

Dear Timothy,

thank you for your answer and the supporting information, which is very helpful!

I understand that parts of the test method are proprietary and cannot be fully disclosed.

I would like to ask for one detail, however: I understand that HTOL and FITs are determined during operation of the component. For an oscillator, this operating state is probably the permanent oscillation at the specified frequency. For a switch IC, however, it is not clear what influence the number of switching cycles performed has on the lifetime of the part. Therefore, an indication of the switching frequency, which characterizes the operating state during HTOL-testing, would be very informative.

Is it possible for you to provide further information on the number of switching cycles performed during the performed HTOL-test of the TPS22810?
If you cannot give exact numbers, the information whether there were at least 200 * 10^6 cycles would be already sufficient.

Many thanks and regards,
Michael

Michael,

I am currently working with our product engineering team to find the information we can provide. I will provide an update shortly.

Best Regards,
Tim 

Dear Timothy,

any news here?

Best,

Michael

Michael, 

I asked our design team to see if switching the FET would have any impact on lifetime calculations. I will ping them again and let you know when I hear back from them.

Best Regards,
Tim

Dear Timothy,

any news from the design team on your end?

Thank you and with best regards,

Michael

Michael,

I apologize about the delay. Talking with the design team, they said that the lifetime calculations will be load dependent. In the context of the MTBF number they do a worst case analysis, however the design team did offer to investigate further based off your specific load conditions. Would you be able to provide what the loading conditions of your applications are?

Best Regards,
Tim 

Dear Timohy,

In my first question above, I described the expected temperature conditions, number of switching cycles and expected load in our application.

Here again in brief:

- The ambient temperature conditions are between -15 °C and + 70 °C. Our device is potted and therefore the TPS22810 is protected from moisture.
- The active time of the TPS22810 is about 120 seconds every 10 minutes, with a switching frequency of 8 Hz and a duty cycle of about 50%.
- The expected are 60* 10^6 switching cycles per year  (365 * 24 * 6 * 120 * 8 = 50.5 * 10^6 + 20% safety margin)
- The minimum required lifetime is 3 years (i.e. 180 * 10^6 switching cycles), but  5 years would be better (300 * 10^6 switching cycles).
- The logic level on the EN pin is 3.3 V. The switched input voltages are 3.3 V or 5.3 V. The loads are usually very low and are usually less than 50 mA.

So the information I would require is if under above described operation conditions the lifetime of the TPS22810 is expected to exeed 5 years (or at least 3 years)?

Thank you for your assistance!

Best regards,

Michael

Dear Timothy,

any news?

Best regards,

Michael

Michael,

Thank you for your patience here. I am talking with the design team to get this calculated and confirmed for your exact loading conditions, however do you mind if I contact you via your myTI email address to take this offline as it is a bit application specific?

Best Regards,
Tim

Dear Tim,

any news yet?

Regarding your suggestion to continue this communication on e-mail:

Well, I would still think that the reply could be interesting for other members of the forum here...

But of course if you prefer to continue this communication on e-mail it is also fine form me...

Best Regards,

Michael

Hi Michael,

I am taking over this device responsibility from Tim. I would need couple of days time to go through the requirement and work with internal teams.

Appreciate your patience.

Thanks

Rakesh

Perfect,

thank you very much!

Best regards,

Michael