Break off the +12V_battery, after shutdown

Hello,

My project use a Jetson TK1, connected to a 10" Universal LCD with HDMI Interface and capacitive multi-touch. These are powered from Li-PO 2200mAh battery. To protect the battery life (display and his back-light have an appreciable consumption!) I want to interrupt the connection from battery of 11.7V, after that the system is shutdown (enter in deep mode). I think is possible to use the signal SATA_PWR__EN_T124 (this breaks off the +12V from SATA POWER connector)to do my intention. My question is: where can I find the all power signal diagrams in shutdown moment? I find some of these diagrams (SATA_PWR_EN_T124 isn’t through them), but only as Power-Up Sequence (see Tegra K1_Embedded_DG_v03.pdf)

I will appreciate any useful answer and piece of advice.
Thank you very much,
JB

I don’t know if you’ve seen the schematics, but they are available here (may require login):
https://developer.nvidia.com/embedded/downloads

I’m unsure of the particular diagrams, but they would be a result of some combination of the Tegra K1 TRM and/or the Jetson schematics.

To clarify, are you wanting to use the SATA drive power function to power the touchscreen instead, such that shutting down or sleep mode would remove power from the display in the same way as power would be removed from the SATA drive?

Thank you dear LinuxDev. Yes, I’ve seen schematics and I can establish the SATA_PWR_EN level (taking in account the specifications of MOSFET Q3E2 - RJU003N03 etc).
Yes I already powered the LCD from +5V of SATA header and with this solution (very good, actually) I saved about 6 hours from the discharge time of LiPO 11.1V 2.2Ah, 20C. The DISPLAY current consumption is about 0.52A. Without this consumption, if I consider an IDLE consumption = 2.5W [i.e. about 0.225A, how is in mentioned in different documentations], the discharge time will be: 2.2Ah/0.225A = 9.78 hours. This means to re-charge the battery in every day.
This is the reason of my intension: to break off just the connection (a Reed Relay)from battery (11-12V). Sure, this discussion is away of others solutions to power save on Jetson (to reduce the CPU and GPU speed) and heatsink instead of Fan. I replaced the fan with one heatsink.
Finally, I don’t wish to affect the TK1 processor through adding one more power switch circuit on U3C1.pin M31 (SATA_PWR_EN_T124)- i.e. one supplementary load on this outpout pin.

Sure, one superfluos question could be: why is not possible to get the detailed specification of TK1 processor, how there are for many other processors?!?!

Thank you again for your advice and promptitude.
JB

It sounds like you wish to use M31 pin directly (SATA_PWR_EN_T124), avoiding circuits which indirectly connect to SATA connector 12V…is this correct? Would you be removing R3E3 (0 ohm)?

I don’t see the voltage level of M31 either. The easiest thing might be to just measure it with an oscilloscope. If no oscilloscope, and if R3E3 is removed, you may wish to use a current limiting resistor (perhaps 10k since this is already used across M31) and a small capacitor, e.g., 0.1uF, and measure directly with a volt meter after a few seconds of run time.

Does anyone have documentation on actual SATA_PWR_EN_T124 pin M31 level?

I am curious, if the SATA +5V works, how would the SATA_PWR_EN_T124 be used? Is the +5V solution lacking in some way which the SATA_PWR_EN_T124 would solve?

The level of signal SATA_PWR_EN_T124, at M31, is about 1,8V in Hi state.

I drew a diagram with my suggestion but, unfortunately, I cannot attash it in message.
In words, the description of diagram is as follows: will be connected at M31, one circuit that modify the signal SATA_PWR_EN_T124, so that to get from this a signal in Hi state(about 1.8V) until apears the falling front of SATA_PWR_EN_T124, when the signal goes in Low state (about 0V). One second power switch identical with the tandem MOSFETs (Q3E1+Q3E2) used for break off +12V from SATA header, will be interposed between battery and 12V Jack (J5C1). This will be commanded of the signal got from the circuit above mentioned. This way, the second power switch is closed when is powered Jetson and is switched to OFF only when Jetson enter in Idle state.
One variant could be using of a Reed Relay instead of the second power switch between battery and Jack, that to be commanded from a timer activated with SATA_PWR_EN_T124.
I hope to works. If it is a stupidity, please tell me to avoid a failure. Thank you for your help.

It sounds like it might work. The only complication I see is if there is a case where SATA drive software changes behavior depending on detection of a real SATA drive or not. Perhaps SATA hard drive software will complicate things, but easiest thing would be simply to test it and find out.

I may be missing the point of this discussion because i stopped reading after specific things were mentioned, but the +5v/+12v sata plug on the board should always be powered when the board is on (takes a few sec at boot).
Not sure about suspend states though. Probably turns off.

I run the fan from the 5v and a powered usb hub from the 12v and they both function as i would like them to. (maybe the hub is 5v, it would make sense, but its been some time since ive even looked at it, so i have forgotten)

If you really wanted to get fancy, you could use one of the GPIOs with a ‘transistor switch’. Not sure the power requirements of the screen and backlight, so you might have to even do a transistor switch coupled with a relay to do it.

Thank you to all. I solved problem in different way.