nRF52805 eddystone beacon issue

RE: nRF52805 eddystone beacon issue

JONATHAN LL — Wed, 21 Apr 2021 12:35:41 GMT

Sorry for late reply,

X2 is fine, X1 is a bit high, we recommend 12pF as CL. Have you been able to test with those values, 21pF and 36pF?

[quote user="matt-lionembedded"]Is your recommendation to test the load caps on X1, X2, or both?[/quote]

Both, The LF(X2) is used to time when to turn on and of the radio so using a X2 you will have better accuracy on the timing. The HF(X1) is used for the transmitting frequency so if this is off then it also effects the result, but it is the CW that is effects instead of transmitting at the intended frequency there will be a offset. So X1 is probably the culprit in this case.

Regards,
Jonathan

RE: nRF52805 eddystone beacon issue

matt-lionembedded — Wed, 14 Apr 2021 15:42:33 GMT

For X1: Load capacitance for the 403C11E32M00000 is 20pF
For X2: Load capacitance for the FC-135 32.7680KA-A3 is 12.5pF

In both cases, I can see in the reference design that the X1 CL recommendation is 8pF and for X2 CL recommendation is 9pF (w/ 12pF external caps for both).

Is your recommendation to test the load caps on X1, X2, or both? Is one more likely to cause BT related issues? I would think the 32 KHz (X2) would be the more likely culprit, but am unsure.

Furthermore, if i understand the link you posted about crystals and load caps correctly, give the following:

C1 = C2 = 2Cl - C_pcb - C_pin

Cl: Load capacitance of the crystal (found in the crystal datasheet)

C_pcb + C_pin is approximately 4 pF.

For X1, with a Cl of 20pF, C1 = C2 = 2*20 - 4 = 36pF. Does this mean C1/C2 should be closer to 36pF for the crystal we are using for X1?

For X2, with a Cl of 12.5pF, C9 = C10 = 2*12.5 - 4 = 21pF. Does this mean C9/C10 should be closer to 21pF for the crystal we are using for X2?

I will work with our hardware team to address some of these.

Thank you for all your help,

Matt

RE: nRF52805 eddystone beacon issue

JONATHAN LL — Wed, 14 Apr 2021 15:18:30 GMT

Can you confirm what Load capaticant you have for X1 and X2 ?

This issue is most likely a offset caused by the crystal loading not being good enough. So one way to test this is to go up or down on the load capacitors and see of the devices(cheep phones) are able to see the 805 HW. You can test by going up first, try phones then if it doesn't work go down and se again.

If you are able to measure the the frequency you can check if not you can send a device to us and we can test.
We do also have a blog on the 52 devices that has a section on Crystals and load capacitors.

Regards,
Jonathan

RE: nRF52805 eddystone beacon issue

matt-lionembedded — Wed, 14 Apr 2021 13:53:30 GMT

Hi Jonathan,

See below the 805 chip-down design schematic (proprietary content obfuscated).

devzone.nordicsemi.com/.../805_5F00_chipdown_5F00_schematic_5F00_20210414.pdf

For X1 (32MHz crystal):

X1: 403C11E32M00000
C1/C2: Murata GRM1555C1H120JA01D

For X2 (32kHz crystal):

X2: FC-135 32.7680KA-A3
C9/C10: Murata GRM1555C1H120JA01D

Thank you for your help,

Matt

RE: nRF52805 eddystone beacon issue

JONATHAN LL — Wed, 14 Apr 2021 11:22:33 GMT

Hi,

What crystal and load capacitors are you using.
Can you share the schematic for the 805 HW you have and include the datasheet for the crystals?

Regards,
Jonathan