VNA Test Connection to Reference Design

[chalecampb]

Hi all,

I need to be able to check the matching networks on a design using the nRF51 series chip. My understanding of the procedure is,

1. Set up the board, enclosure, etc. and the cabling
2. Calibrate the VNA to account for the cabling impedance
3. Perform VSWR measurements on the board under test

My confusion is with part 2. What I have found is that

• SWF is not usually adaptable from an N-type standard VNA connection
• SWF Male connection is not usually provided on calibration loads
• There are no adapters from SWF to N at all, or SWF to SMA with cables

Do people usually swap J1 in the reference design with an SMA connector? If not, my test setup is going to look something like

• N-Type to SMA Cable -> SMA to SWF Cable -> SWF Male to MM8130-2600
• N-Type to SMA Cable -> SMA to N Adapter (no cable) -> Calibration Load

My concern is that the calibration load is not going to account for the SMA to SWF cable, and I do not understand why this is not an issue. I am also concerned that the SMA to N adapter will differ in characteristics from the SMA to SWF cable. What kind of error would this introduce? Are we talking 1% or 30%? For reference, the SMA to SWF probe cables associated with the MM8130-2600 female SWF connector is 12" long.

Is SWF referred to by any other connector series name?

I also checked Nordic's recommended test setup white paper, and they listed overall requirements and suggestions, but didn't provide cables or part numbers for the cables.

What does everyone typically use for the VNA connections?

Thanks, -Alex

[ketiljo]

Hi Alex

You shouldn't use a VNA for this. It will only tell you the impedance when you look into the matching network. It will not give you any information about the actual output power, the level of the harmonics or spurious in RX mode. It's far better to use a spectrum analyzer and measure all these parameters.

Setup a constant carrier and measure output power and harmonics at the same time. For this, you can use the SWF cable if you have the same connector on your design as on the nRF51-DK.

A VNA is usefull for antenna measurements. You calibrate the coax cable with your cal kit as usual. Consult the operating manual for the procedure. Then you need a short test cable with a suitable connector in one end. The other end should be open. Setup the VNA to measure impedance and add electrical length (in the calibration menu) until you measure open, or near open, impedance. Then, solder the cable to the board. The center conductor should be as short as possible, 1 mm or so. You need to break the connection between the antenna and the matching network and solder the cable to the antenna side. Make sure the outer conductor is firmly connected to the ground plane.

[chalecampb]

Hi,

I am sorry I might not have been clear, but this is to explicitly perform antenna matching. We do have a spectrum analyzer, but that cannot provide the VSWR measurements that I need to select the components required for manufacture.

Thanks, -Alex

[ketiljo]

If you want to measure and match the antenna, this must be done after the connector, not looking into the matching network. You have a defined 50 ohm plane. Measure the radio part (matching network) with the spectrum analyzer and the antenna part with the VNA.

Note that the matching network components can not be used to tune the antenna, you need additional components for that.