1

SOLVED nRF52810 module high current, wakeup every 26ms

nemik gravatar image

asked 2018-01-12 19:42:42 +0100

updated 2018-01-13 16:47:28 +0100

Hello,

I'm using a module based on the nRF52810, the Fanstel BT832A. I'm running the basic "ble_app_beacon" firmware from the "nRF5_SDK_14.2.0_17b948a" SDK with a few changes:

  1. I have no external low frequency clock so using the RC with CTIV on 16 and TMP on 2 as recommended.
  2. It's advertising every 2 seconds
  3. I created a 'CUSTOM' board for it with most pins set to 0
  4. In the Makefile I commented out "-DDEVELOP_IN_NRF52832", "-DCONFIG_GPIO_AS_PINRESET", and "-DNRF52_PAN_74". I'm not sure what the last 2 are but just tried them.
  5. I based it all off of the "pca10040e" which is supposed to be the one for the 810 and uses the right SoftDevice.
  6. I removed all references to logging and Segger RTT to try to reduce the current.

Everything is working well and it's broadcasting, but when I'm measuring the current usage from my scope, it's much higher and noiser than the nRF51822 projects I did before.

Here's an image from my scope:

image description

I expect it to be very low current between the 2 second advertisement as I circled in green, but really it's mostly like the red. So the average current is about 180 uA which is pretty high I think for a simple non-connectable advertisement every 2 seconds.

Any suggestions would be greatly appreciated, thank you!

EDIT:

Digging in further, I'm seeing current spikes about every 26 milliseconds. Here's a graph: image description

It's a clean signal and I can't really see why this would be happening. I've tried to disable everything I could in the config.h and still keep the thing working. I don't understand what could be causing it to wake up so often.

EDIT 2: After tracking this down further, it appears to be some noise in my setup. The current spikes were very sharp and dropped off in a downward slope, so they did not look like digital signals. When I replaced my power supply with a 3V lithium battery, things improved and the spikes were gone. Wiring and connectors could have also played a role. So the problem was my setup.

Here's an image zoomed in of one of those little spikes: image description

The example code, module, and soft-device appear to be working fine!

edit retag flag offensive close delete report spam

1 answer

Sort by » oldest newest most voted
1
J.Kwon gravatar image

answered 2018-01-13 08:00:53 +0100

updated 2018-01-13 08:03:11 +0100

I encountered this issue but did not receive any answer from nordic..

This problem may be related with using radio.

I think second screenshot is captured when your device is advertising.

The more radio you use, the more spikes occur.

Maybe if you do not use radio, these spikes will not occur.

Since the size of the spike is insignificant, it will not have a significant effect on the circuit unless the amplifier is used.

But I also care about this problem... I was planning to use the bead on the power supply to test it, but I did not have time.

edit flag offensive delete publish link more

Comments

1

The second screenshot is actually between advertisements. The advertisements themselves are very clear larger peaks that go up to 13 mA which makes sense for a 0 db transmission and are only present in the first screenshot.

Those smaller spikes every 26ms don't look like transmissions, their peaks are only 600 uA and they slope downwards instead of a sharp drop like a transmission, and don't even last for long enough. I updated the post with a new screenshot zooming in on one of the spikes.

If you see similar problems, try to shorten cable runs, tighten up connectors, and use a battery instead of a bench power supply. All of those helped me out, but I also have a stupid error in my PCB that I think can produce stray capacitance and make this problem worse.

nemik ( 2018-01-13 16:44:58 +0100 )editconvert to answer
1

Just want to also add that a decoupling capacitor close to the Vdd pin is no joke, it's extremely important as everyone says and my fault for not placing one. I put in a 0603 22uF cap there and the signal is so nice and smooth now, only the Tx power spikes can be seen and the noise floor is very low as it should be.

nemik ( 2018-01-13 19:45:02 +0100 )editconvert to answer

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer. Do not ask a new question or reply to an answer here.

[hide preview]

User menu

    or sign up

Recent questions

Question Tools

1 follower

Stats

Asked: 2018-01-12 19:42:42 +0100

Seen: 65 times

Last updated: jan. 13