Considerations for adding an LNA to the EM250/EM260

Is it worth it to add an LNA to my EM250/EM260 design? What are the tradoffs? How can I evaluate if my design will benefit from adding an LNA?

The EM250 receiver had a nominal noise figure design target of 10dB (including balun losses). Thermal noise floor is at -174dBm/Hz, so total noise referred to the receiver input is at -164dBm/Hz. The noise bandwidth of the matched filter is about 1.2MHz, so receiver input noise in this bandwidth is 10*log(1.2e6) bigger, i.e. approx. -103dBm/Hz. We need about 3dB SNR to achieve the 1% PER, so signals at -100dBm should be detectable.

Adding an LNA could improve the sensitivity. It's very possible to achieve under 2dB noise figure in a bipolar based external LNA. If its gain is much greater than the 10dB receiver noise figure, say 15dB, it will dominate the overall noise figure. An overall figure of 3dB should be possible. This will extend the sensitivity achievable to -107dBm.

In order to add an external LNA, a customer will need to add an external SP2T Switch. In this instance, the loss of the switch would add directly to the noise figure, e.g. a 2dB loss would increase the overall receiver noise figure by 2dB. There's a bit of an assumption here that we're talking about a constant source and load impedance, but I'm sure this will be pretty accurate. As to what might be acceptable, that's rather subjective. Clearly, if this loss approaches around 7dB, all of the advantage will have been lost.

You can use the Friis formula for cascade noise factor to see the overall effect:

Ftotal = F1 + ((F2 - 1)/G1) + ((F3 - 1)/(G1*G2)) + ...

Where Fn & Gn are stage n noise factors and gains respectively. Use 10*log of these numbers to get noise figure & dB gain respectively. I'd recommend designers plug all the numbers in, assuming an existing Rx noise figure of 10dB and see what they get, because all designs will be slightly different.

There are side-effects though. The level of signals hitting the receiver will be louder by the gain of the external LNA. The receiver has been designed as a complete system to meet a set of sensitivity and interference rejection requirements that are mutually conflicting with respect to receiver gain. The net effect will be that the receiver's ability to cope with loud WiFi signals will be reduced by the LNA gain.

In addition, to take advantage of the increased sensitivity, background noise to non-thermal effects needs to be that much lower, e.g. mush from the digits. Results achieved in the laboratory do not always translate to range testing with real antennas.

All designs and applications are unique. This would need to be carefully tested to make sure that the desired results are achieved and the LNA achieves a worthwhile improvement in sensitivity.