RAK4631 - VDD_NRF (no) connection

WisBlock
#1

Hello,
I’m trying to include the RAK4630 within our custom boards and I have a doubt regarding the power supplies.

According to the RAK4630’s datasheet, the supplies should follow these specifications:

image
image1373×519 36.5 KB

I’m using the RAK4631 schematic as a reference but I can’t see what’s the connection of the VDD_NRF signal from the RAK4630. Below is the schematic, in green I’ve marked the parts related to the VDD_NRF, R4 and R8 are supposed to connect VDD_NRF to a supply but both are marked as NC (not connected).
I also check these two resistors in a RAK4631 board and they aren’t soldered.

In the RAK4631 board, I also tested continuity between VDD_NRF and every contact in the WisConnector, but no connection seems to exist, so no voltage is being applied to VDD_NRF, is that correct?

image
image3491×2458 328 KB

Also, there’s a note in the RAK4630’s schematic:

image
image1669×714 210 KB

Since the RAK4630 has a metal shield that covers all components, how I can ensure that L10 is not soldered?

#2

L10 is always soldered and the RAK4630 stamp module as it is sold is not built for single 3.3V supply.
The power supply connections as shown on the RAK4631 schematics are the ones that should be applied when using the RAK4630 stamp module on a custom PCB.

image

#3

@beegee Thanks for your reply.
Can you confirm if VDD_NRF from RAK4630 is not connected in the RAK4631?

#4

Yes, I can confirm that. VDD_nRF is generated by the nRF52 itself by its DC-DC converter from VBat.

1 Like
#5

@beegee thanks for your reply.
After checking this other post, I’ve come to the conclusion that the information in the datasheets is confusing.

As I can see, the RAK4630 has the following power supplies according to each document:

image
image919×365 13.2 KB

I think that the Recommended Operating Conditions should follow the real operating conditions.

In my way of understanding why VBAT_NRF can’t be 3.3V I guess the nRF52840 is internally using the High Voltage mode and the output for REG0 (buck converter) is configured to 3.3V, so the input voltage (VDDH → VBAT_NRF) has to be higher than that.

image
image984×657 38.5 KB

So, the higher the voltage of VBAT_NRF, the higher the efficiency, and therefore less current consumption.
image
image1400×761 164 KB

Regarding VBUS voltage, it can be left unconnected if isn’t in use:

image
image1205×551 121 KB

In conclusion:

  • VBAT_NRF requires a voltage in the range 3.7V - 4.2V
  • VBUS can be left unconnected if USB is not being used.
  • Any other supply can be tied to 3.3V

Only for curiosity, If is it correct that the output for REG0 is configured for 3.3V, is there any reason that no lower voltage was chosen for that? So that a single 3.3V external voltage supply could have supplied everything in the RAK4630?

#6

Hello Nicolas,

I saw your post and though the same. I’m a little bit confused here, since I’ve planed my PCB with those connections:

image
image660×554 26.3 KB

But reading this post makes me think of removing R5 on my PCB. My custom PCB is manufacturing right now and I expect to come on a week.

I’m using LiFePo4 battery on V_BAT3V6 flag, and a step down regulator on 3V3_REG flag, probably my error was no email to RAK Wireless before manufacturing… let’s see… BTW I have also a footprint for RAK11720 on this PCB

#7

Hello Javier,
Yes, it seems that R5 should be removed since VDD_NRF is generated by the nRF52 itself.

1 Like
#8

Dear @nicolasech
Many thanks for the advice, I’ve received my custom PCB’s and tested removing VDD_NRF. At the moment the module works without problems. I’ll keep the forum updated if I find something weird.

Again, thanks!

1 Like