nrf52840 ANT impedance clarification

RE: nrf52840 ANT impedance clarification

Kaja — Thu, 16 Mar 2023 09:03:15 GMT

Hi,

the measurement is done using the radio test example, using the radio in TX mode, with a constant unmodulated carrier.

Best regards,
Kaja

RE: nrf52840 ANT impedance clarification

Lucas — Wed, 15 Mar 2023 19:23:36 GMT

Hi Kaja,

After reading the matching network pdfs I noticed that you have measurements for different TX power levels (0-8dbm). Do I conduct these measurements in advertising mode?

RE: nrf52840 ANT impedance clarification

Kaja — Wed, 08 Jun 2022 11:18:41 GMT

I don't have this data saved.

Best regards,
Kaja

RE: nrf52840 ANT impedance clarification

JouniP — Tue, 07 Jun 2022 12:56:16 GMT

Hi,

Thanks for quick reply.

And yes I have pictures of impedances in PDF file, but S-parameter files with real data are much more useful.

Best Regards

Jouni

RE: nrf52840 ANT impedance clarification

Kaja — Tue, 07 Jun 2022 12:47:54 GMT

Hi,

I already sent you this one:

[quote user="Kaja Soerbotten"]s. nRf52840 QIAA matching network (1).pdf[/quote]

Best regards,
Kaja

RE: nrf52840 ANT impedance clarification

JouniP — Tue, 07 Jun 2022 12:38:56 GMT

Hi Kaja

I simulated our layout and matching component values for IC-ANT to 50 Ohm point. I have used S2P-parameter file found from HERE as reference.

With slight component tuning I was able to get quite nice correlation with your measured S-parameter file and my simulation model. We can now proceed to our first prototype build with these questimated/simulated values:

We have also 2nd design with your three component reference matching circuitry. Would it be possible to have S-parameter file for this reference design matching circuit?

Kind Regards

Jouni

RE: nrf52840 ANT impedance clarification

Kaja — Thu, 21 Apr 2022 12:49:15 GMT

Hi,

1. Yes, the radio is the same.
2. Yes it varies with output power, so we do a tuning of the components values with TX in channel 2, 40 and 80, and adjust to get low harmonics and high output power. Only matching based on impedance would not work, as harmonics will vary from channel to channel. You can have a look at our RF performance test guidelines
3. When it's adjusted for TX, it's usually good for RX as well, it very rarely/never that I need to adjust the components for RX, when it's tuned properly for TX,
4. Yes, it's a measurements of the matching network:

The two ports are the ports on the VNA.

Note: We offer to do RF tuning and antenna tuning for you, this is a free service we provide for our costumers.

Best regards,
Kaja

RE: nrf52840 ANT impedance clarification

Philippe — Thu, 21 Apr 2022 07:51:56 GMT

Thanks for the reply and the document. I have some questions about the documents and its measurements.

Measurement was done with HW revision C0. Is it for HW revision F0 still the same?
The measurements shows that the impedance vary with the output power, so the matching should be done according the selected output power?
If I do matching for max output power, what happens with the RX path? Is max output power matching equal perfect RX matching?
The 2 port measurements I don't understand. Is this the measurement of the matching network. If yes, how does this matching network look like.

RE: nrf52840 ANT impedance clarification

Kaja — Wed, 20 Apr 2022 12:58:54 GMT

Hi,

the matching network with four components are better at filtering harmonics, so if your planning to use 8 dBm as TX power, we recommend the one with four matching components.

We do recommend that you use two matching networks, one for the radio and one for the antenna. This is to make sure that you can adjust the values, for your specific design, making sure that you are able to pass the certification criteria's.

Since the PC is not designed to see a 50 Ohm load, so we can't do the measurements on the antenna pin directly. But we have done measurements on the matching networks. devzone.nordicsemi.com/.../nRf52840-QIAA-matching-network-_2800_1_2900_.pdf devzone.nordicsemi.com/.../nRF52833-QIAA-matching-network.pdf

Again, I recommend that you have two matching network, so that you have a 50 Ohm reference point, to do conducted measurements.

Best regards,
Kaja