shireesha Post subject: IIP3 variation with input RF power in a
mixer Posted: Wed Jun 04, 2008 3:26 am Captain Joined:
Wed Jun 04, 2008 3:23 am Posts: 11 I am working on a single resistive
FET mixer. In the circuit, the input IP3 is decreasing by an amount
of around 5-10dBm when i vary the input RF power from 0 to -20dBm, eventhough
the input power for compression is greater than 15dBm. In the literature,
what i have gone through shows that IIP3 is invarient of the RF power.
So, can anyone please help me out in this regard. Does IIP3 changes
with RF power? If yes, at what power level should we carried out the
two tone analysis? Top IR Post subject: Posted:
Wed Jun 04, 2008 3:49 pm Site Admin Joined: Mon Jun
27, 2005 2:02 pm Posts: 373 Location: Germany Hello shireesha,
IP3 tests are performed in small signal. What can happen is
that you reach to a signal level in which the Mixer does not work anymore
in small signal region and experiencing some non-linearities.
Check also the following: 1. Do you provide the right LO
level? 2. Are the ports correctly matched? Top
FSomma Post subject: Posted: Wed Jun 04, 2008 6:17 pm Also
check that the intermodulation is not being generated on your test bench.
At a certain amount of power, IMD can be generated by the CW generators
(if you are using a combiner to feed the two carriers, the isolation
of the combiner is critical... you might need to add some circulators
to increase the isolation and perform a correct measurement) Also,
at a certain amount of power, your spectrum analyzer might be intermodulating,
try to lower the mixer level of your SA to check it.. You can
clear all the test bench doubts by replacing the active device with
a passive non intermodulating device (ex.. a resistive attenuator).
Take care Top shireesha Post subject:
Posted: Thu Jun 05, 2008 2:44 am Captain Joined: Wed
Jun 04, 2008 3:23 am Posts: 11 Hello Fsomma, Thanks for
your reply. Sorry, but i forgot to mention in the query that i am still
at the simulation level. Suggestions provided by you will be very useful
at the testing level. Can you suggest me what can i do in simulation.
Top shireesha Post subject: Posted: Thu Jun 05,
2008 2:57 am Captain Joined: Wed Jun 04, 2008 3:23 am
Posts: 11 Dear IR, Thank you for the reply. What i feel
is mixer will start experiencing non-linearities once the RF Power fed
is nearer to the compression point. In the present design, Input P1dB
is around +20dBm & the power what i am feeding is either -20dBm
or 0dBm. So, do you think that it will experience non linearities?
LO Power fed to the device is in the range of 12-17dBm, with a small
gate bias. All the ports are not perfectly matched, but the return losses
are around 10dB. Top FSomma Post subject: Posted:
Thu Jun 05, 2008 7:37 am Perhaps, because of some mismatches, your
Input P1dB is not being +20dBm... check that by driving the device until
having the IM3 tones at -30dBc from the fundamental tones, there, PEP
power will give you your real Input P1dB. Top shireesha
Post subject: Posted: Tue Jun 10, 2008 6:45 am Captain
Joined: Wed Jun 04, 2008 3:23 am Posts: 11 Hi FSomma,
I have checked the P1dB in the way what u have suggested. At around
21dBm RF power, IM3 tones are -30dBc down from the fundamental tones.
Top yendori Post subject: Posted: Wed Jun 11, 2008
9:00 am General Joined: Thu Sep 25, 2003 1:19 am
Posts: 50 Location: texarcana In its linear region a mixer's
IP3 "should" not be unaffected by RF power. Of course, when
RF power increases, the RF works against the LO bias the FETS or diodes.
So at RF compression levels the affective LO power will decrease greatly.
In the mixers linear region, the LO power applied to the mixer
may be much different than the power applied to the non-linear devices.
So if your circuit is not matched properly compression may occur to
soon. But, you said, Quote: 21dBm RF power, IM3 tones
are -30dBc down from the fundamental tones. So your IIP3
is 36dBm? Sounds good to me! Rod [/quote]
Top FSomma Post subject: Posted: Wed Jun 11, 2008
10:30 am Hello Shireesha, 1st: About the following: the
21dBm, are PEP, AVG, or it is the power of each tone? shireesha
wrote: At around 21dBm RF power, IM3 tones are -30dBc down from the
fundamental tones. 2nd: As I believe, while you are in the
linear zone, the IM3 tones should increase on a 3 to 1 ratio respect
to the change of input power, this, should keep your IP3 constant, I
understand that that´s the "linearity" you are not achieving on the
simulation and that I think you should... This is only to check, that
your question is regarding to the fact that on your simulation you don´t
have the 3 to 1 ratio.... Can you send an IM3(expresed on dBC)
vs. Input Power curve? bye! Top shireesha
Post subject: Posted: Sat Jun 14, 2008 3:22 am Captain
Joined: Wed Jun 04, 2008 3:23 am Posts: 11 Hi FSomma,
Regarding Input P1dB being 21dBm, i have checked the power of each
tone, but not PEP or Avg. At 20dBm of RF Power, power of the fundamental
signal was 13.175dBm and that in the IM3 signal is -17.83dBm. (diff=
-17.83-13.175 = -31.005). Sorry, i tried but could not upload the curve.
As you have said, in the linear zone, IM3 tones should increase
on a 3: 1 ratio respect to the change of input power. But my simulation
don't have 3:1 ratio eventhough the P1dB shown in simulation is >20dBm.
Can you please tell me how many harmonics should be included
for harmonic balance simulation of mixers, as i see that the number
of harmonics of the tones are also effecting the IP3 performance.
Thanks & Bye! Posted
11/12/2012
|