What is the proper power down sequence?

[mscottwvc]

I have an FT813 and am trying to figure out what the proper power down sequence is. Initially I tried:

EVE_memWrite( REG_PWM_DUTY, 0 );
Then setting PD_N pin low

The memwrite turned the backlight off, when I sent the PD_N low the backlight turns back on, so clearly I missed some step or steps, because when I startup my board the backlight is off.

[mscottwvc]

I tried some more stuff and got the power usage to go pretty low, but I dont understand why I can't use the PD_N pin to turn things off, even with the new sequence below as soon as I pull that pin low the power usage goes up and the backlight turns on. Guessing a lot of these steps are needed, but this what I have now:

EVE_memWrite8(REG_PWM_DUTY, 0x0U); //turn off backlight
uint8_t g = EVE_memRead8(REG_GPIO);
g = g & ~0x80;//clear bit 7
EVE_memWrite8(REG_GPIO, g);
usleep(10000);
EVE_memWrite8(EVE_PCLK, 0U);
usleep(10000);
EVE_memWrite8(EVE_PWRDOWN, 0U);
usleep(50000);
sleep(2); //for debug
gpioWrite( GPIO_N_SCR_PD, 0); // here the power usage goes up and the backlight turns on

[RudolphRiedel]

I have not actually tried to go into low power.
But I know that the PD pin works as a reset.

I have an Arduino UNO and a RVT50H up and running right now, so BT815 actually.
The display is supplied with a bench PSU, so the current it shows is only from the display and my interface board with a step-down regulator.
The PSU supplies 12V.
The current drawn running is 124mA.

First test, just switching off the backlight when pressing a button.
EVE_memWrite8( REG_PWM_DUTY, 0 );
while(1);

-> the current drops to 58mA

Second test, set PD to low:
EVE_pdn_set();
while(1);

-> the current drops to 24mA

Third test, doing both:
EVE_memWrite8( REG_PWM_DUTY, 0 );
EVE_pdn_set();
while(1);

-> the current drops to 24mA

Not actually low current, but there are components still powered on the display.
That would be the touch controller and the flash chip, at least.

I disconnected the display now from my interface board -> 0mA
When I press the reset button on the Arduino I am getting 24mA, my interface board has a pulldown on PD.
When I press the reset button on the Arduino and do a power-reset with the PSU, the current drawn is only 2mA.

Ah, ok, just pulling the rug with the PD pin is likely not the best idea.
And I believe now that I have seen a power-down sequence somewhere, I only can not remember where.

Edit: tried
EVE_cmdWrite(EVE_PWRDOWN,0U);
but that makes no difference.

Setting REG_CPURESET to 7 also makes no diference.
Setting the low-power mode bit in REG_TOUCH_CONFIG also makes no difference.

Hmm, I am not getting below 24mA.

[mscottwvc]

@RudolphRiedel Thanks for trying that. At least your power consumption goes down when you do EVE_pdn_set(). Mine goes up and the backlight turns on, which is odd, Ill have to probe some stuff later and try to understand why that is happening.

Also curious what power do you get if you have the display connected and then power on your system from off, but dont make any writes to the ft815 chip, is it still 24mA or something else?

[RudolphRiedel]

Yes, it really should not go up and the backlight should definately not turn on.
The current drawn before initialization is only 2mA for the display module.

Which microcontroller are you using?

[mscottwvc]

Thanks for measurments. I have been desinging two products that share the same screen type. Microcontrollers are NRF5340 and raspberry pi 4s. This issue happens with both, I measured the PD pin and its going to 0 volts. I'm currently using this screen https://newhavendisplay.com/7-0-inch-ips-eve-tft-lcd-module-with-capacitive-touchscreen/ , which has the circuitry as part of it. So that circuitry must be the problem. I'll look into it a little more and reach out to new haven, but seems like the safest solution may be to just put a fet type switch between the screen power and backlight power to completely cut off power. Then as long as all my IO are pulled low should be good to go, what a pain though.

It's odd that even with your setup a fair amount of power is still being drawn after shutdown.

[RudolphRiedel]

Ok, I have neither of these to play with, but that should not really make a difference.
I found the NRF5xxx boards to be a bit expensive to buy one for no other reason than to play with another micro. :-)
And raspberry pi is just way out of scope for me.

I switched to a EVE2-43G now, so FT813.
But the general behaviour does not change.
Baseline is 44mA with the backlight setting.
17mA when switching off the backlight.
The same when also using an empty display list before switching off the backlight.
16mA when sending the host command for power-down.
9mA when using EVE_pdn_set() instead.

Oh, that one has a red LED on the back, so 3mA to 4mA are just for that.
That leaves power-down current of the FT813, the touch controller and the panel itself.
The backlight driver should be in shutdown.

And before init this display module draws 5mA.

At least from 5mA to 9mA is less of a gap and the LED could be removed.