Insertion loss of BAL-NRF01D3

[Jamie]

Hi,

I recently made a custom board using the nRF51822 and the LC network using values from the reference design. After fabrication and assembly, I tested the insertion loss of the resulting design and found it to be between 1-2dB.

Looking at the data sheet for the BAL-NRF01D3, I see it reports its own insertion loss to be 2.25dB. ST also claims higher performance of this part compared to the discrete LC network due to better harmonics rejection.

How to decide which method (LC or balun) can result in the best overall performance?

1. Using a pi-network, can I tune the LC network to have similar harmonics performance as the balun, and conversely, 2) can the balun be somehow tuned to have an insertion loss comparable to the 1-2dB I could achieve with the LC network?

Thanks, Jamie

[Jamie]

I just found the following at this link:

"The high Q (antenna performance boosting) thin-film technology employed in the ST BAL-NRF01D3 also significantly improves RF performance and characteristics over the use of standard passive component matching. This includes: a +0.9dB average output power improvement at 4dBm; +1dB better sensitivity at 1Mbs on-air transmit speeds; 10dB better LO (local oscillator) suppression; and -39.1dBm 2nd harmonic suppression. The improved harmonic and LO suppression levels also promote simpler mandatory FCC/ETSI compliance testing and approvals."

So I have another question: 3) Can the above be achieved with proper pi-network tuning of the discrete LC network?

Thanks, Jamie

[RENARD]

Hello Jamie,

The important parameter is not the "individual' IL of the balun elements but the global performance in the application (output power, sensitivity, harmonics...). Measuring the IL on balun depends on the conjugate impedance you are using to load the balun. Indeed, the optimized matching value to get the maximum power and the best harmonic filtering is not necessary the conjugate match of the RF IC.

This is in fact the case using teh BAL-NRF01D3 where the balun device integrates the optimized impedance to get maximum power and lower harmonics. In addition, using BAL-NRF01D3 instead of discrete LC, you "simplify the complexity" of the implementation.

Depending on which version of nRF51822 QFN you are using, you must use either BAL-NRF01D3 or BALF-NRF01D3. For nRF51822 WLCSP, you have to use BAL-NRF02D3.

If you have any questions about balun, do not hesitate to contact us.

Best Regards, Richard RENARD STMicroelectronics

AN4111.pdf

[Gus Gustafson]

What are the possibilities of getting a preliminary data sheet? Also are any engineering samples available?

Gus

[Jamie]

Thank you for your reply, Richard.

We will want to use the nRF51822-QFAA G0 version of the chip. Is that the version that goes with the BALF-NRF01D3 balun that you spoke of? If so, I cannot find any information on this balun anywhere online, including the ST website.

If I cannot get enough BALF baluns in time (beginning of February) can I pair the G0 chip with the BAL balun? I see that in the PAN, this combination fails Korean regulatory testing: "61. RADIO: Designs based on nRF51822 QFN packets using balun BAL-NRF01D3 are likely to fail Korean teleregulatory requirements."

What about FCC/IC and Bluetooth SIG tests? Should I just use a LC matching network to be safe even though I won't be testing in Korea?

Thanks, Jamie

[RENARD]

Hello,

Datasheet of BALF-NRF01D3 will be available very soon in the www.st.com internet site. In the meantime, see attached the preliminary datasheet

Best Regards Richard

BALF-NRF01D3.pdf