nRF8001 Antenna Matching Network

RE: nRF8001 Antenna Matching Network

Dino — Fri, 10 Mar 2017 17:34:15 GMT

Thank you,

RE: nRF8001 Antenna Matching Network

AmbystomaLabs — Fri, 10 Mar 2017 17:32:14 GMT

It is on page 3. www.johansontechnology.com/.../Johanson_nRF8001_App_Note_v2.pdf

RE: nRF8001 Antenna Matching Network

Dino — Fri, 10 Mar 2017 17:28:57 GMT

you motioned 2mm distance from the nRF8001 to the BALUN, do you have the documentation from Johanson or the link to it. Thank you

RE: nRF8001 Antenna Matching Network

Dino — Fri, 10 Mar 2017 14:48:44 GMT

I'll work on that Thank you for your help I really appreciate that.

RE: nRF8001 Antenna Matching Network

FormerMember — Fri, 10 Mar 2017 14:18:56 GMT

The balun will match the chip to 50 Ohm. To achieve minimal loss, the antenna should also be matched to 50 Ohm as well. All chip antennas will need to be tuned when placed on a board that is different from the manufacturers board, because the antenna impedance depends on it's close surroundings.

RE: nRF8001 Antenna Matching Network

AmbystomaLabs — Fri, 10 Mar 2017 12:16:09 GMT

That is not true, Dino specified that they are using a Johanson chip antenna which is spec'd as 9.5dB return loss without matching though a pi network is recommended to clean it up if needed.

RE: nRF8001 Antenna Matching Network

FormerMember — Fri, 10 Mar 2017 08:32:22 GMT

In addition to a balun, an antenna matching network is needed, to match the antenna to 50 Ohm. This tutorial, section "Matching networks" explains why it is necessary to match the antenna to 50 Ohm. The tutorial is written for nRF1, but the theory in the "Matching network" section is valid for nRF8001 as well.

RE: nRF8001 Antenna Matching Network

AmbystomaLabs — Thu, 09 Mar 2017 23:13:20 GMT

And here is datasheet/app note for the antenna. Again, the antenna placement, feed and grounding are detailed here: www.johansontechnology.com/.../2450AT18A100.pdf

RE: nRF8001 Antenna Matching Network

AmbystomaLabs — Thu, 09 Mar 2017 23:11:03 GMT

Here is the link to the Johanson app note for your balun. Note in their design it is 50ohm to an SMA connector. Only difference for you is you go 50ohm from balun through pi network to chip antenna. www.johansontechnology.com/.../Johanson_nRF8001_App_Note_v2.pdf

RE: nRF8001 Antenna Matching Network

AmbystomaLabs — Thu, 09 Mar 2017 23:00:53 GMT

The important factors about the balun location is it needs to be very close to the radio IC or the match it was designed for is no longer valid. Johanson specifies less than 2mm in their app note for the balun. Also, input and output feed lines need to be characterized. Input is per Johanson spec output is 50ohm. For standard 2 layer, 0.062" board most people do this as a coplanar wave guide with ground plane. It's the only style of stripline that will be sufficiently narrow to actually hook up to these parts.

RE: nRF8001 Antenna Matching Network

AmbystomaLabs — Thu, 09 Mar 2017 23:00:37 GMT

I'm not aware of any reference designs by Nordic that have used the Johanson balun. Most have a discrete balun/match. Only the 51822 beacon reference design used an ST balun. But I would not recommend the ST balun to someone new to the field. The WLCSP ST balun is tricky to place.

You need to follow the Johanson design criteria which are detailed in their datasheets. Also, as I noted in my answer there was a design done using a very similar Johanson balun for the nRF51822 that was correctly done in the end. You could look to that design for guidance.

RE: nRF8001 Antenna Matching Network

Dino — Thu, 09 Mar 2017 22:34:37 GMT

Thank you for your prompt replay,

is Nordic has any layout reference for nRF8001 that we can look at and try to match.

Thank you Dino

RE: nRF8001 Antenna Matching Network

AmbystomaLabs — Thu, 09 Mar 2017 18:47:12 GMT

Without seeing the actual layout it is impossible to say you did it correctly.

But in terms of the schematic, yes what you did is correct. The balun replaces the external match/balun circuit for the nRF8001 and the antenna is spec'd to be 9.5dB return loss without a match.

However, I should stress than any deviation from Johanson's design guidelines will cause you to have a mismatch condition and an antenna that doesn't broadcast. So all the trace lengths, trace widths and ground planes need to be precisely correct.

Also Johanson does highly encourage one to place a pi network ahead of the antenna to fix any matching problems that might show up on the antenna side of the problem. I would encourage you to do this as well. Also, the pi network can help out with harmonic filtering if you ever decide to go through certification.

A few weeks back I walked someone through all the layout issues with a similar design using the 51822. Here is the final design which was done correctly: http://imgur.com/a/33xSx

Here is a link to the original post: devzone.nordicsemi.com/.../