AM5726: POH (power on operating hours) details

Ioan,

The variables you mention here: 

- ARM CPU at 1.5GHz uses only USB 2.0 and PRU (GMAC and PRU ETH)

- running at lower speed (somewhere between 1.5 and 1.176) what would be the TJ - POH relation (this assume VDD_MPU is OPP_HIGH)

don't significantly impact the estimated POH.

The significant contributor is Tj which is already shown in the table.

Regards,

Kyle

Ioan Marusca said:

The cooling conditions for the application are creating a TJ that by the table 5-3 gives a too low POH number.

Keep in mind that these numbers represent the expected lifetime when operating 100% of the time at the stated voltage (OPP) and temperature.  If most of the time you operate at a lower temperature, and only have excursions up to some of these high temps, your lifetime will increase.

If you expect to operate most of the time at the temperature you're looking at, the easiest way to dramatically increase the lifetime is to eliminate OPP_HIGH from your system.  In your case you're not using HDMI/PCIe/USB3_SS, so the 105C expected lifetime is 50k hours at OPP_HIGH, but doubles to 100k hours if you operate at OPP_OD (1176 MHz).

I understand your comment but running the CPU at highest speed is needed for the project to achieve it's targets.

This is a multidimensional problem (frequency, voltage, Tj, operating hours) to solve and I would need more details from TI.

I will try to insert here (hopefully this web tool permits) an excel file where with green I entered data provided in the datasheet and with blue I entered some numbers that I tried to come up as an "educated guess".

I would appreciate TI taking a closer look at this example and help replacing the numbers in blue with actual numbers based on reliability/process testing.

Ioan,

I'm looking into this and will get back to you.

Regards,

Kyle