OMAP3503+TPS65950 power consumption

This is posted in the wrong forum.  This forum is for the Stellaris parts (LM3Sxxxx).

MIK Monnireth said:

- what is the difference between Off and Off/0v?

In the Off, the power domain in chip is turned off. In the Off/0V, the internal power domain is turned off and external power supply to the processor or core is turned off, so there could be additional power savings by putting the DCDC in the PMIC to low power state.

MIK Monnireth said:

-could i consider that "OFF" state is one of standby mode?

Yes, it is one of standby mode. You could still do suspend/resume to/from DDR w/o cold boot.

MIK Monnireth said:

-for our budget of 20uA typical power consumption in Standby mode, which standby mode should i adopt?

The OFF mode is the lowest power consumption standby mode but can't achieve the 20 uA target. The lowest current measured in the mode is around 100 uA but can vary depend on device process and temperature.

Dear,

Thanks for replying.

We will use OMAP3503 CUS package @600Mhz (423 pins - PBGA) with TPS65950 under -40 to +71 degree C.

With the lowest current of 100uA (according to your answer), could we find amongst this reference the best device (OMAP3505 CUS) from the TI's best process & temperature characterization to get less power comsumption.

Regards

M.MIK

Dear,

Could we have an idea in terms of wake-up latency (approximately) per standby mode.

The standby 3 is the matched one vis-à-vis to our consumption budget.

I would like to know the difference in wkup latency between standby 2 and 3 for OMAP3503 + TPS65950.

Kind regards

M.MIK 

The wake-up latency would be ~ 4 ms for standby 1 & 2, ~ 5 ms for standby 3 &4 and ~ 45 ms for the OFF mode.

Dear,

Thanks for quick reply.

I get back to you concerning the power estimation tool (we have 10mw in consumption budget.

I wondered in entering input information into the spreadsheet.

See atteched file input spreadsheet for power consumption in standby mode.

Coud you have a look in those files? are there coherent: input information and output  results?

To conclude what I obtained from these trials.

-standby 1 => 12mw

-standby 2 => 8 mw

-standby 3 => 3mw

Another question about active state power consumption estimation (see attached files ARM 600MHz at OPP5, Core at OPP3).

With some peripheral use in our application, I got 62 mw. It's so small I think!

Did I miss some thing?

Kind regards4061.omap35x_pet_input_adv_standby1_bis.xls

M.MIK

 1106.omap35x_pet_input_adv_standby2.xls6278.omap35x_pet_input_adv_standby3.xls7382.report-omap35x_pet_input_adv_standby1_bis.xls6886.report-omap35x_pet_input_adv_standby2.xls2604.report-omap35x_pet_input_adv_standby3.xls3582.omap35x_pet_input_adv_active.xls3125.report-omap35x_pet_input_adv_active.xls

The standby power numbers look reasonable. The most part of power is from leakage in standby mode and the leakage current varies depends on temperature so it would be lower at room temp (25 C) than your results at 71 C. The reason why the power is so small in your active power use case is due to your low module utilization. You are setting only 1 % for the module utilization in all enabled modules. Is that what your use case looks like ?