Phantom current - where's it going?

[angust_uk]

Hi all

I'm struggling to figure out where I'm losing ~11mA in the very simple design below - it's an STM32F103 and a TLC5947 LED driver (driving 24 white LEDs). The design will be battery powered, so in standby I am putting the STM32 into sleep mode (which I've checked works ok with current consumption reducing as it should), and switching the driver's VCC off with a P-channel MOSFET. However even in this state I'm seeing 11mA into VCC and I can't figure out where it's going!

I have noticed that when off, the MOSFET's drain is at 1.5V (around half of VCC), so I guess I'm seeing leakage from somewhere else, but I can't figure out where!

Any ideas appreciated!


(post edited as apparently I managed to forget the schematic!)

[Siwastaja]

Assuming Vcc=3V, 33mW dissipated in a small spot on a PCB should cause small but well noticeable heating, which you should be able to see using, for example, a cheap IR no-contact thermometer. In similar cases, I remember seeing heating of around 1-2 degC over the surroundings.

Put it on a table, let it run for 15-20 minutes without touching it anywhere and measure.

[jhpadjustable]

Wild guess, but are you disabling output drive on the STM's SPI pins when going to sleep?

[angust_uk]

Ok, it looks like the issue may well be my lack of understanding of the STM's various sleep modes! After more poking around at the registers I've put it into a deeper sleep mode which reduces consumption to around 1mA. I'm not using an external 32kHz crystal which I guess may limit my ability to go much lower, but for the moment this is significantly better at least!

Many thanks jhpadjustable and Siwastaja for the pointers!