BCM20737 power consumption during Normal advertising

Tip / Sign in to post questions, reply, level up, and achieve exciting badges. Know more

cross mob
Anonymous
Not applicable

Dear BCM supports,

We have a customized board ack like below:

1. Wake up for advertising for 20 seconds.

2. After 20 seconds timeout, go into deep sleep for a while.

The deep sleep current is about 80uA, which is pretty good, but the wake up current is alway base on 3mA, the device is just advertising. Please check the below pic:

PIR_on_PIRHW.png

Our local FAE told us to download mybeacon project as a reference, here is the result, you can see the base current is still 3mA:

mybeacon_on_PIRHW.png

Our question is:

1. Is it possible to make the current lower during advertising? 3mA is still huge since we are battery based device.

2. Is it the limitation of BCM20737 for 3mA operation current? We have tried a lot of method to make it lower, such as Turn OFF ADC, disable all timer, turn off UART function etc, neither of them works.

3. We know the normal sleep functionality is handled by ROM code, but it seems not work with it very well, and hint to fine tune our application to let the ROM code work with sleep properly?

Thanks.

0 Likes
1 Solution
Anonymous
Not applicable

Hi all,

After remove the resisters on the SPI pin, our current issue has been solved. Thanks.

View solution in original post

0 Likes
5 Replies
Anonymous
Not applicable

Sorry for a typo for last question:
ANY hint to fine tune our application to let the ROM code work with sleep properly?

0 Likes
Anonymous
Not applicable

Hello Henry,

These are a few modifications you can make to your hardware to reduce your advertising current:

1. If you have a ferrite bead connected to your battery or your USB to reduce noise, removing that might decrease the power consumption

2. Add a LDO regulator to your board.

Thanks,

Arsh Buch

0 Likes
Anonymous
Not applicable

Hi Arsh,

I have check with our HW engineer, below is the comment from him:

Here is our schematic for BCM20737 RF circuit. The ferrite bead we designed on our board is with below spec. I don't think it would affect our current consumption that much (your reference board is 600uA, but ours test result is 3000uA). Also, adding a LDO dedicated for these 1.2V power rails (VDDIF, VDDVCO, VDDFE and VDDPLL) should be the wrong direction to go because the LDO does introduce extra power losses. Any other idea to bring the power consumption down?

FERRITE BEAD, 120Ohm@100MHz, 680mA, DCR 0.13 Ohm, 0402

BCM20737 schematic for RF circuit.png

0 Likes
Anonymous
Not applicable

Did you initialize the GPIO's to a low power state?  What are your advertising parameters?  I would not expect to see much sleep during advertising if your intervals are small.  The system will not sleep if it knows there is work to be done in the near future.  I agree there is not enough information (hints) on how to optimize sleep, but I doubt sleep is your problem.  Our custom board with a 200 MS fine timer and connection interval roughly the same draws about 1 mA.  In theory it should go lower if it was sleeping properly but that's as low as I can get it.   An early design of ours was flawed where a voltage regulator was drawing 4 mA because one of the pins needed to be grounded to put the regulator into low power mode.  Once that was solved the 4 mA drain was gone.

Anonymous
Not applicable

Hi all,

After remove the resisters on the SPI pin, our current issue has been solved. Thanks.

0 Likes