Is IP3 such high as 74dBm really needed ?

[artiste]

Hi everybody !

I've found that switch from Hittite having a IP3 of 74dBm. It's a T/R switch for Cellular/3G, TD-SCDMA, PMR,...
I do understand the necessity of having very high IP3.
But such a value of 74dBm is really needed ? For an amplifier of the same manufacturer, the highest IP3 I've found is 50dBm for a PA with a gain of 21.5dB.
From cascade IP3 calculation, a switch with an IP3 of 60dBm would give a cascaded IP3 of 48.7dBm while a switch with an IP3 of 74dBm give a cascaded IP3 of 49dBm.

I'm not an expert in system design but, it might not worth to use the switch with the IP3 of 74dBm....

Is there really applications that need such high IP3 ? If yes, please tell me !

Thanks in advance for your answers..

 

[biff44]

I am not up on the latest cell phone basestation designs, but you could pretty easily see where you might need such a thing. Lets say you have an amplifier, and a really tight channel filter to reject adjacent channel interference. After the bandpass filter, you would have what looks like a very high ip3 point for the whole system. If you then went thru a switch of somesort for beam forming, etc, and it screwed up the IP3 that you faught hard to achieve, then you are screwed again.

I have a cell channel filter somwhere in the lab. It is huge! 4 resonators, and each one is the size of a office trash bucket!

 

[Abhishekabs]

May be they have designed for specific customers.
Like said by Biff44, there is possibility for reduction in IP3 duto to use of filters.
Also there are different regulations in all ccountries.

So to suppress the harmonics & unwanted signal you will use filters after your PA.

So it wiill degrade IP3

I think, this will be used only when you have one system ready & you need to use it somewhere else without much change in system to reduce the cost.

I am right? I am also new in this systems.

 

[vfone]

This ~70dBm IP3 is related to other very important specification of high power RF switches (semiconductor not mechanical) which is P0.1dB (not P1dB).
This is the point where the insertion loss of the switch increase by 0.1dB vs input power.
To get a high level for P0.1dB the IP3 should be higher as possible.

 

[artiste]

Hi,
Thanks all for your fast answer!
The P0.1dB is interesting. Vfone, you wrote it is a "very important specification of high power switches".
Can you tell more about this spec ? The why and the how ? It's pretty hard to find litterature about this point.

Thanks in advance,

 

[BigBoss]

Hi,
Thanks all for your fast answer!
The P0.1dB is interesting. Vfone, you wrote it is a "very important specification of high power switches".
Can you tell more about this spec ? The why and the how ? It's pretty hard to find litterature about this point.

Thanks in advance,

Because RF switches are made with generally FET devices so these are naturally nonlinear behaving components under hard pressure levels. And in all cases, these devices may create nonlinear behaviour under very high signal levels.Therefore OIP3 is figure of merit of a switch even tough the switches don't contribute into nonlinearity of the system below their OIP3 level.