2018-08-17 05:13 AM
The SPI works well. I can asleep the sensor but when I try to asleep the EnOcean module, I see that the sensor consume near 500µA. I have put two pull down on SDI/SDO and SCL and two pull up on the CS_ACC and CS_MAG.
2018-08-20 12:00 AM
Any other idea?
2018-08-20 12:48 AM
Considering the high current consumption is only when the EnOcean module is in sleep mode and the GPIO are probably in high-Z, the problem could be cause by undefined level on the CS pins. The pull-up 10M are very week. You can try to decrease the value. But it is just my idea.
Btw. are you sure the high current is drawn by LSM303AGR?
2018-08-23 05:53 AM
Hi,
I have decreased the value of the pull-up resistors but nothing happened.
I'm sure that high current is drawn by LSM303AGR because I have put one ammeter on the VCC of the EnOcean module and another of VCC of LSM303
2018-08-24 04:26 AM
I don't know which value of the pull-up you used, but please try following:
Put two 10k pull down resistors on SDI/SDO and SCL and two 10k pull up resistors on the CS_ACC and CS_MAG. Input pins must always be driven high/low (no floating).
2018-08-24 07:41 AM
Hi,
I have the same behavior...
2018-09-25 01:51 AM
Hi,
@Jérémy Da Silva , did you solve your problem?
I'm having what I guess is a similar problem.
When I put the accelerometer in power-down mode it starts consuming way too much power(~650 uA).
If I put it in low power mode it goes down to ~10uA(the whole system).
I'm figuring it has something to do with pin states as you've discussed.
Is it okay to have CS_X pins tied to VDD?
I've tried that in combination to pulling I2C pins high/low without success.
EDIT: I just solved my problem, the solution was to put the INT_MAG signal to a pullup.
2018-09-26 02:14 AM
OK, this is what I recommended in the thread.