VL53L1X - i2c consumption

[Alskandar]

Hello,

How do I deactivate/activate a SDA/SCL line?

I perform a test on an Icarus board with a VL53L1X sensor, everything works fine.
However, in idle/sleep mode, whether I turn off the sensor or not (by XSHUT, GPIO or physically), it consumes an additional 10uA. 
Even without code to initialize the sensor, and only plugging in the SDA & SCL ports, I have a 10uA boost. 
I don't understand where this is coming from or how to fix it. 

Does someone have an idea ?

Thanks !

[ivnn]

Hi,
try to configure gpio pins to tri-state, which could be done with nrf_gpio_cfg_default(pin)

[Alskandar]

Hi,
I'm with an nRF9160.
I tried like this, without success.

That is to say that the ports are at 0V but no reduction in consumption. 
Worse when I reactivate the port, it rises above 1.5mA.

pins_vl53L1X = device_get_binding("GPIO_0");
gpio_pin_configure(pins_vl53L1X, 26, GPIO_OUTPUT);
gpio_pin_configure(pins_vl53L1X, 27, GPIO_OUTPUT);
gpio_pin_set(pins_vl53L1X, 26, 0);
gpio_pin_set(pins_vl53L1X, 27, 0);

[Alskandar]

The Otii Arc power the board in 3.6V (on Vbat) at the same times it takes the measurement.
For UART, in my prj.conf i make CONGIG_SERIAL=n, same for spm.conf.
Otherwise the consumption increases by 600uA.

[ivnn]

I meant you should desolder it

[mike_actinius]

Hi Alskandar,

Are you using the VL53L1X with a breakout board? If so, which one and what is on it?

PS the FTDI is completely unrelated to this and it anyway does not draw any current when the USB is unconnected. You schematic Alskandar seems fine and VBAT is indeed the right pin to connect the Otii.

BR,

Mike

[Alskandar]

Hi mike_actinius

No, i don't use a breakout board.

Not in the sense with additional features.

This one:

It seems, as Øyvind assumed, that it comes from the leakage current at in off mode or 0V (or just i2c plugged in without 3V3 and GND).

I am looking to see if there is a simple solution to put in place to "scrape" these uAs and lower consumption.

[mike_actinius]

Hi Alskandar,

In general the I2C should never be connected without the IC being properly connected on VCC and GND. Some (most?) times the ICs get backpowered from the I2C/UART/etc lines when not properly connected to VCC+GND.

From what I see there are no pull-ups on the breakout board, so that takes away the possibility of leakage from them.

Here is what I would do to pinpoint the source:

- Connect GND + VCC, tie XSHUT TO VCC, leave I2C and GPIO1 unconnected -> measure the current

- Tie the XSHUT line to GND -> measure the current

- Connect the I2C lines with short wires -> measure the current

That should give us where the current difference is coming from.

The datasheet specifies that hardware shutdown current draw can be up to 7uA, and that's for 2.8V VDD, so in my opinion most of what you see is just that. But anyway, the above should give more info.

BR,

Mike

[Alskandar]

Hi Mike,

Yes i agree for i2c, it's just an information like what to connect or not VCC / GND, as long as i2c is connected, there are these 10uA.

I did the tests:

• always 45uA (for VCC + XSHUT without or with i2c)
• 340uA (for GND + XSHUT without or with i2c)

if it is the hardware shutdown current, this consumption should disappear once the operation is carried out, right ?

Even after 30sec, still nothing.

[mike_actinius]

Hi Alskandar,

In the case of XSHUT connected to GND, you measured 340uA? Is that correct or is it a typo?

BR,

Mike

[Alskandar]

Hi Mike, yes that's it.

I made a bridge between XSHUT and GND.

I think this is the same as if i directly connect XSHUT to 0V, right ?

And i get 340uA.

[mike_actinius]

Hi Alskandar,

Since you see this regardless whether I2C is connected perhaps something in your firmware "misbehaves" when the I2C to the sensor is lost (when XSHUT is low I2C is disabled on the sensor anyway). As a sanity check I would check the sensor separately, without any other board connected and that should yield 5-7 uA. If that is indeed the case, then you should check your firmware on how it handles sleep mode when losing connection to the sensor.

Are you absolutely positive that there are no components on the breakout board? A ~10k pull-up resistor on XSHUT could easily justify this amount of current drawn when setting XSHUT to low

BR,

Mike

[Alskandar]

Hi Mike,

I tested sensor by itself (juste VCC & GND connected).

At 3,3V, it gives 12,5uA

At 2,8V, it gives 10,5uA

So i have a little more than in datasheet (5uA at 2,8V).

Unfortunately , i don't have the design plans for this sensor.

Here is a photo, there seems to be resistors (350 Ohms, i think, on GPIO1 & XSHUT) but i don't know what the others are.