External Power Supply Problem:

RE: External Power Supply Problem:

Burnie Barren — Fri, 10 Sep 2021 10:59:17 GMT

You are a legend Stian Røed Hafskjold! I did what you said and I am able to advertise with a current on the order of uA from the supply. May no sorrow come to your door. <3

RE: External Power Supply Problem:

Stian Røed Hafskjold — Fri, 10 Sep 2021 09:55:44 GMT

Hi, I read through your other posts, and as I understand you are trying to power the whole board with a piezo harvester. The problem is that the external supply header supplies the segger debugger/UART, USB circuitry, level converters, analog switches, ++ in addition to the nRF52 itself. If you want to power the nRF52 only, you will have to do some modifications to the board. Please follow my steps here:

Cut SB9
Cut SB26 (bottom side) (RESET)
Cut SB20 (bottom side) (SWDIO)
Cut SB21 (bottom side) (SWDCLK)
Power the board directly on VDD_nRF as shown in the picture below:
Modify the ble_app_beacon application to disconnect all pins (by default the 4 LED pins and the UART TX pins are configured as outputs). Otherwise the rest of the board will be powered through the GPIOs.
- Set NRF_LOG_ENABLE to 0 in pca10040/s132/config/sdk_config.h
- Remove the leds_init() call from the main function.

When flashing the board you will no longer be able to use the on board debugger as normal since the SWD wires are cut. You have two choices (unless you want to re-solder SB20 and SB21 every time you flash the chip):

Use an external debugger:
1. Power the board on the External supply header (3V)
2. Short P22 with a jumper
3. Connect the debugger to the Debug in header
Use the on board debugger:
1. Connect a USB cable
2. Short P22 with a jumper
3. Connect a SWD 10 pin cable between the Debug Out and Debug In headers

RE: External Power Supply Problem:

Burnie Barren — Fri, 10 Sep 2021 06:43:05 GMT

Hello Stian, thank you for your answer.

I am using nRF52 DK, on it it says 'nRF6827 2.1.0 2020.39' if it helps. I also tried to eliminate JLink debugger by openning SB29 but the chip did not function. As with your suggestion given below I checked the clock parameters. Application uses external LF clock X2 on the board having 12 pF capacitors as bypasses. They are all intact on board. I have actually posted a ticket about powering and another ticket about some peculiarities elaborating further on respective issues. I would be glad if you could help me out with my issue.

Best regards.

RE: External Power Supply Problem:

Stian Røed Hafskjold — Thu, 09 Sep 2021 19:06:23 GMT

Hi chinmaydixit,

If the chip is running when the debugger is running, but not when the debugger is disconnected, it usually means that the reference design has not been followed. It could be wrong decoupling caps or a missing LF/HF crystal. Can you please post the board schematics?

RE: External Power Supply Problem:

Stian Røed Hafskjold — Thu, 09 Sep 2021 19:04:06 GMT

[quote user="Burnie Barren"]Initially I supplied with a 80% DC square wave[/quote]

So are you powering the board with a function generator? I doubt that it is able to deliver the current needed by the DK.

[quote user="Burnie Barren"]I increased the current limit up to 80 mA because the input could not rise again and current seemed to be the limiting issue.[/quote]

Not sure what's wrong here. Which DK version are you using? I've used the external power connector on lots and lots of DKs over the years here in Nordic and never had any problems with that. I agree that 80mA is way too much, and that 3mA is normal. But keep in mind that the ble_app_beacon demo running on the chip has a consumption in the µA range. So the 3mA you are seeing is mostly drawn by the JLink debugger chip.

RE: External Power Supply Problem:

Burnie Barren — Tue, 07 Sep 2021 12:50:55 GMT

I connected a large capacitance as suggested in this post and the board executes nicely drawing 3 mA from the external pins. Still I find the behaviour to be pretty peculiar.

RE: External Power Supply Problem:

Burnie Barren — Tue, 07 Sep 2021 12:08:47 GMT

I have the same problem so instead of opening up a new ticket, I wanted to reply this one. I am using nRF52 DK and running ble_app_beacon demo. With USB and battery I verified the functionality. When I wanted to check with external power supply I encountered some weird behavior as listed below:

> Initially I supplied with a 80% DC square wave, f = 10 mHz, 1.9 Vpp and 2.6 V_offset. I did like this because I did not have a DC supply so I tried to converge to it as much as possible. When input tries to reach HIGH, it can't and oscillates frantically between 3 and 2.5 V. When I stopped the the graph on the oscilloscope screen, after a few minutes, demo DID actually run, i.e. LED 1 started blinking. But did not continue to do so when I ran the oscilloscope live. I could recreate this queerness another time but not a third time.

> After this I changed the power supply and found one that can actually give constant 3.3 V with an ability to measure current. With that, I increased the current limit up to 80 mA because the input could not rise again and current seemed to be the limiting issue. As the supply reached 80 mA the board resets itself such that there are ground spikes in the input voltage waveform.

During these experiments only SB12 is shorted on the default configuration of the board. I think even reaching 80 mA is questionable with a simple beacon demo. Diode check yielded no shorting between the external supply pins.

I would be glad if you could help me out with this.

RE: External Power Supply Problem:

Mttrinh — Fri, 03 Sep 2021 14:19:54 GMT

Hi,

Can you give me more details about your setup?

What device are you using? Are you using a DK or custom board?

What SDK and Softdevice are you running?