Hi all,

for the IIP3 simulation, we usually take the inteferrer as very close to the main signal. However, if the adjacent or alternate channel is quite far away, then this simulation method might not apply, especially after the channel select filter. My question is

A) in general can we ignore IIP3 after channel selection?

The channel select filter is also quite difficult to design with very high linearity and low noise/ low power. I'm aiming for IRN <50 nV/sqrt Hz and IIP3 > 1V while power consumption is around 500 uA for a 3rd order BP filter at 2 MHz with around 0 dB gain. I've come up with an architecture which barely meets this, but i'm thinking that since filtering is taking place before the second and third stage of the filter, the IIP3 requirement is not so strict.

B) Is there a way to simulate IIP3 with the interferrer at the adjacent/alternate channel? My frequency is 2 MHz with channel spacing of 5 MHz. I was thinking i could put in 1 tone at 7 MHz, one at 2.01 MHz (fundamental from pac analysis) and one at -3 MHz (if that's possible) and measure at 1.99 MHz...is that right?

thanks,
Aaron

Hi Aaron

Most often the interferer for your IP3 are the adjacent channels. In general I would not say you can not ignore IP3 after the channel selection - it really depends on the system/filter specs. How low does the distortion in band have to be relative to your desired signal? Depending on the filter roll off, the two adjacent interferers may still be present (although reduced) and can still beat together to produce a tone on your desired. The filter will definitely reduce your linearity requirements but then some other specification may determine the overall linearity requirement that can translate into an equivalent IP3 spec.

The simulate the IP3 for you setup, I would recommend placing your interferers at 7MHz and 12MHz, they will produce a beat tone at 2MHz (and 19MHz).

Bill

Thanks for the replies

12 and 7 MHz is also a possibility, the only thing i'm worried about is that the interferrer levels i'm dealing with are different for adjacent and alternate channels leading to different IIP3 levels for each. Maybe i should try all cases and take the worst case... better check my overall IIP3 with the stages after the CSF too

My system is operating in the ISM band and i'm working on the IEEE 802.15.4 standard (2.4 GHz). So far i've only been trying to meet the requirements of the standard without making any assumptions of other bluetooth/WLAN/others operating nearby. I think the idea is that you have to assume that there aren't too many other devices in the area otherwise the specs you must meet become very tough.

Also from my understanding, if a narrowband interferrer is in your channel then it is treated as noise and if its large enough then you must switch channels. Is that right?

thanks,
Aaron