This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

AM3517 Minimum Power

Other Parts Discussed in Thread: AM3517, AM3715, INA219

I am working with an AM3517 SOM-M2 and am attempting to configure it to use as little power as possible while in sleep mode.  Following the instructions in: http://processors.wiki.ti.com/index.php/UserGuidePowerMgmtAM35xx I was able to obtain a current draw of 35 mA while in sleep.  I have attempted to disable some Kernel components, but so far the only significant impact I have seen is a 14mA drop with USB support disabled.  Adjusting the MPU clock rate does reduce the normal operating current, but does not appear to lower the sleep current draw.  My goal is to get the module running below 10mA while in sleep.  Can anyone give me some advice on further Kernel and/or hardware pin configurations which may help me further reduce the current draw?

Thanks!

  • Hello,

    How are you measuring the power ? Are you using current shunt registers on the SOM ?

    What's your VDDSHV voltage setting ?

  • I am measuring current draw of the entire SOM by connecting it to a voltage generator that has a current readout.  On the SOM I am supplying regulated 5V DC to the MAIN_BATT_IN lines.  All other pins (except the GND lines and UART3 tx/rx) are not connected.  I am aware of the current draw from the RS-232 chip I have connected and have subtracted it from the values above.

    I have not modified the current shunt registers.  Would it be possible for you to advise me on this or direct me to the necessary documentation?  Any additional advice on reducing sleep mode power consumption would be greatly appreciated.

    Thanks!

  • There are four current shunt registers to measure the power for separate power supplies on the SOM. If you could measure voltage drop across the current shunt register, you could see how much current is drawn in each power supplies. This measurements will help to see what could be optimized further.

    R178 : MAIN_BATT_IN

    R174: 1.2V

    R175: 3.3V_or_1.8V

    R176: 1.8V

     The schematic and the assembly drawing are available at Logic PD website (http://support.logicpd.com/auth/login.php). The assembly drawing shows the location of the current shunt registers.

  • Measuring the voltage drop across the following resistors(while in sleep), I got these values:

    R178: 1.2mV

    R174: 0V

    R175: 1.0mV

    R176: 0V

    When the schematic says 0.04 for these resistors, does it mean 40 ohms(0.04k) or 0.04 ohms?

    I also noticed that these resistors are connected i2c addresses.  Is it possible to obtain these values internally while in sleep mode?  I would imagine those would give very accurate readings and may be simpler than manually probing.

    Edit: I checked and VDDSHV is set to 3.3V.

  • Your 3.3 V power supply shows higher current than expected. The 3.3 V also supplies an external USBPHY and Ethernet PHY. You may want to make sure that these devices are also in suspend status. To quickly see if these devices are active or not in suspend, you could assert reset signal to the devices before going into suspend and see if the current is further reduced or not. Ethernet PHY reset signal is connected to GPIO30 and USB PHY reset signal is connected to GPIO_57.

    The register value should be 0.04 ohms. The i2c interface is for reading power from a power management daughter card. You can still measure the power in better resolution if your multimeter supports 0.01 mV resolution. Can you set your multimeter to read the voltage in 0.01 mV resolution ?

  • My apologies, I did not notice you had replied to this older thread of mine.

    I have switched to a newer SOM module and used a more precise volt meter to get these results:

    R178: 0.35mV

    R176: 0.04mV

    R175: 0.25mV

    R174: 0.06mV

    The current draw on this module is better, but still not near the 4.7mA minimum I was quoted in my latest thread.

    As for asserting the reset signals, unfortunately it did not improve the current draw on sleep.  I received the following results: Eth high: no change, Eth low: +12mA. USB high: +30mA, USB low: no change.

  • I think the powers measured by R176 and R174 are almost optimized. I can't tell if the power measured by R175 is minimum or not since the 3.3V_or_1.8V power supply powers not only AM3715 VDDSHV IO but also other components on the SOM. The VDDSHV standby current in 3.3 V mode measured on the development board was 1.7 mA when the pad configurations were optimized for the board, so you might be able to lower the 3.3V_or_1.8V supply current 6.25 mA by optimizing the pad conf for your setup.

    Here are current calculations from your measured voltage drop. I used voltage values to calculate the current  from my measurement.

    Current Sense Resistor SOM Signal Name AM3517 Power Supply Resistor Value [ohm] Voltage [V] Voltage Drop [mV] Current [mA] Power [mW]
    R178 MAIN_BATT_IN   0.04 4.15 0.35 8.75 36.31
                   
    R174 1.2V VDD_CORE 0.04 1.192 0.06 1.50 1.79
    R175 3.3V_or_1.8V VDDSHV 0.04 3.296 0.25 6.25 20.60
    R176 1.8V VDDS, VDDS_SRAM_MPU, VDDS_SRAM_CORE_BG, VDDS_DPLL_MPU_USBHOST, VDDS_DPLL_PER_CORE, VDDS_DSI, VDDSOSC 0.04 1.789 0.04 1.00 1.79







    24.18

     

  • Hi,

    Logic PD release their new Wattson software for the AM3517 EVM today which will help you see actual power consumption across all power rails on the CPU and SOM design.  There is an adapter required to connect to the PC available from this website: http://createnewstuff.webs.com/breakoutboardwattson.htm

    The Logic PD software is available from here: http://www.logicpd.com/wattson

    -A

  • Hi Justin,

    I am also working with AM3517 SOM-M2 and trying to lower the consumptions, but i didn't find a way to get your results. I'm using AM3517SDK 05_02_00_00.

    My voltage drops in suspend mode are:

    R178 => 2.1mV

    R174 => 0.163mV

    R175 => 0.846mV

    R176 => 0.907mV

    I check this voltage drops with multimeter and reading INA219 power monitors registers. This is an order of magnitude over your results! Should I only follow the instruction in http://processors.wiki.ti.com/index.php/UserGuidePowerMgmtAM35xx, or should I also make major changes in kernel?

    Thanks

    Matteo