TI based ISM Adapter - LNA on receive side? Filters?

[trevorth]

Hi,

I wondered if anyone could advise on how to go about calculating if I need a LNA on the receive side of a Bluetooth module.



I need to add an amplifier for range following testing of the TI device by itself. As Ive added RF switches which have an insertion losses, would I need, or would it be wise to add a LNA on the receive side? The RX LNA available seem to boost the signal much more than required; I presume this doesn't matter so long as its filtered and doesn't exceed the input specification of the TI RF SoC (all that would happen is the AGC would attenuate the signal)?

Would it be a good idea to add a filter just before the TX PA. Ive got a filter on the output of the TI RF SoC and have followed the PA specification but obviously this is currently no-mans land

Thanks,
Trevor

 

[biff44]

it all depends on the operating range. If you have to work at close range, you might overload the receiver and get bit errors. If you work mostly at extreme range AND there are not strong adjacent channel jammers, an LNA might make sense.

THe AGC range is pretty limited in these chips, adding a power amp and LNA eat into the AGC range, and the max/min ranges it will be happy at. The switch loss will only be around 1 dB per switch, btw.

These "system on a chip" ICs are carefully engineered based on an "optimum" set of trade-offs. If you try to push them too far (like adding a lot of system gain externally), you will have to sacrifice something else in performance. You can not take your sports car out on the road and try to load a ton of wood in the trunk....something is going to give! There is no access to the insides of the chip to modify the filtering, IF gains, demodulation method, etc....so you can only push it so far.

BTW, there is a chance you will need to add a lowpass filter at the power amplifier output to comply with ISM band requirements in your country.

 

[trevorth]

The devices may be right next to each other at one instance, than at the range extremes another so I think ill leave it as it looks like Id just be giving the AGC a hard time. And as you point out, the switches losses are only 0.4dB@2.4GHz. Im looking into the output filter. Thanks for the advice