how to define IIP3 of this system?

[fxxjssc]

Hi all,
I am designing a direct-conversion receiver for UHF RFID reader but
with some confused question about how to define IIP3 of such a system.
This application defines 250KHz for each channel while a large block
signal(say -15dBm) centered at the center. Two desired signal spaced
40KHz or 80KHz from the block with -80dBm system sensitivity required.
With these information, how can we get IIP3 of the system?
the picture of what i just said is as following:
Thanks all
fxxjssc

 

[Mazz]

Do you mean that in a single channel can be toghether the blocker and the useful signal?
how can it be possible? Where the blocker comes from?

Mazz

 

[jallem]

If IIP3 comes from non-linear devices then it looks
to me that you need to work with the stronger
signal. IIP3 will give you the signal level that
eventually saturate the receiver. I do not think
it has any thing to do with -15 dBm and less with
a signal 55 dB below that. It has to deal with the
stronger signal the receiver can receive before
act as a non-linear device.

 

[milkdragon]

I agree with jallem that there is nth to do with the blocker when characterize IP3. The blocker simply says that the max interference in the channel, and you gotta design your circuit such that with this power of blocker, you can still senes your "original" signal (i.e, the blocker not saturate your circuits). So, IIP3, if it is inband, should be measured with the two tone test in same channel that the IM3 product falls inside the band, or it can be measured with two out of band signals that produce IM3 which falls in you channel

 

[fxxjssc]

Thanks all,
So I make my increase LNA's input P-1dB to
0dBm while 10dBm for mixer and 10dB power gain
for LNA.
what's your opinion?
By the way, I prefered passive mixer for my
application. Is that OK?
Best wishes

 

[fanshuo]

well, I think it has something to do with the blocker.

The ideal is the 3rd order product of the strongest blocker signal or interference signal (using two tone test) should be smaller than the wanted signal with certain margin, say 10dB.

You can refer to Razavi's "RF microelectronics"