DWM1001 issue with 32.768KHz xtal for nrf52832 RTC

Wireless Connectivity Ultra-Wideband
1

Hello everyone,

I am currently developing a custom application on a DMW1001-Dev board.

I need to keep track of time for a synchronization between multiple devices.

I checked on Nordic forums and I implemented a RTC counter for this (depending on the 32.768KHz xtal), following all their recommendations to get the best precision. However, I am now blocked with a 100-120ppm precision that I don’t understand…

I checked the Schematics of the DWM1001 and I confirmed that it uses a 32K crystal (Commutek). I tryed to find this KCBB crystal (as specified in the BOM) and I found this one : TF-SMD (http://www.commutek.com/?p=6&cat2=2&id=30&s=&lang=en).

My problem is that this xtal seems to need a capacitance load of 12.5pF whereas the nrf52832 has a 9pF reference for the 2 capacitors of 12pF on the PCB.

Is it possible that the capacitors needed by DWM1001 modules for getting a +/-10ppm precision has wrong value, 12pF instead of 21pF maybe ? Is it that the xtal that I found is not the right one. Have someone any other idea which could cause the drift of my RTC ?

I hope you will give me some insights, thanks, Fabien.

2

Hi
You can usually calculate C1 = C2 = 2 * CL-4pF (Input capacitance on XL1 and XL2 pads) according to the specification of the crystal CL = 12.5
C1 = C2 = (2 * 12.5) -4 = 21pF

according to the nRF52 Load capacitance = 12.5pF data sheet. in theory that crystal would not be compatible. You can try a 21pf capacitor and see what happens or look for another crystal.

regards

3

Hi,

Thank you very much for your response. It confirms what I was afraid of…

If this is really this 12.5pF quartz in DWM1001, I don’t know how I can change the capacitors or the quartz since they are encapsulated in the module. I don’t want to change to DW1000 because the DWM1001 is already certified for EMC. So it seems like a dead end for me, maybe i will have to use an external RTC IC…

Regards

4

Hi,

Just checking if maybe you had found any solution to this. I seem to be in the same issue where I am also seeing 100ppm of drift on the crystal.

Thanks