|
|
|
|
s-param mag./phase - RF Cafe Forums
|
boon
Post subject: s-param mag./phase Posted: Sat Jul 15, 2006 2:02
pm
Captain
Joined: Tue Feb 21, 2006 5:19 pm
Posts:
17
hello all
this could be very stupid question....as far
as s-param. related how is magnitude related to phase? meaning, if magnitude
changes does the phase also change? meaning if you see a shift in dB
curve of s11 does it necessarily mean the phase curve has to change
to somewhat?
thanx
-B
Top
Stephen
Post
subject: Posted: Sun Jul 16, 2006 1:27 am
Captain
Joined:
Wed Jun 21, 2006 8:33 pm
Posts: 21
Location: Queen Creek, Arizona
Boon,
Actually, S-paramters account for both magintude and phase
and can be described in complex numbers. Typically, you only see the
magnitude portion in terms of dB, however, every S-paramter should also
have an angular (phase) component reported with it, but typically is
not.
For example, a circuit may have an S11 of -26dB. But it
should also be reported that it has an angular component of -175 degrees
or an almost 180 degree phase shift (which you would expect from a reflected
compnent.).
Again, the problem is most of the time we read/discuss
about s-paramters, it is almost always in terms of the magnitude and
the phase is ignored. Just realize that S-params do account for phase
and it should also be reported with the dB's.
I hope this helps.
_________________
CMOS RF and Analog ESD Specialist!
www.srftechnologies.com
Top
Peter Raynald
Post subject: Posted: Tue Jul 18,
2006 1:28 pm
Captain
Joined: Tue Sep 07, 2004 3:09
pm
Posts: 11
Most of the time the phase will change, but there
are controlled situation where you can see either the mag or phase change
while the other one does not.
I think of a phase shifter circuit,
or a resitive pi attenuator on which you vary the resitances to change
the loss.
If you create missmatch using a resistor, in serie
or paralel, you will see the impedance change on the central axis of
the smith chart, therefore you have no phase shift, but a magnitude
change.
Top
boon
Post subject: Posted: Wed Aug
02, 2006 1:45 am
Captain
Joined: Tue Feb 21, 2006 5:19
pm
Posts: 17
hello all...
another question about phase.
if i have a common emitter (CE) amplifier then we expect the phase to
-180 degree ideally at the output. i have a case where the phase is
approx. -125 degree but is very constant and then increases as you compress.
in other words, its AM-PM curve is quite good. what can be said about
the linearity of this circuit looking at phase (assuming AM-AM is good
too)? is it good or bad? does not having -180 deg. mean/indicate bad
linearity even though AM-PM is good or you just look at AM-PM &
AM-AM curves?
thanx
-B
Top
IR
Post
subject: Posted: Wed Aug 02, 2006 2:25 pm
Site Admin
Joined: Mon Jun 27, 2005 2:02 pm
Posts: 373
Location: Germany
Hello boon,
Besides the 180ยบ shift which you should expect
at the output of CE amplifier, the AM-AM and AM-PM can yield almost
everything. The direction of the slope for these curves whether positive
or negative depends on the technology of the device. I have seen (At
the same device) both negative slopes for AM-AM and AM-PM and also positive
AM-PM and negative AM-AM. In general, you should expect that the slope
of the AM-AM and AM-PM will increase (To whatever direction it has)
when you reach compression.
_________________
Best regards,
- IR
Top
boon
Post subject: Posted: Wed Aug
02, 2006 2:42 pm
Captain
Joined: Tue Feb 21, 2006 5:19
pm
Posts: 17
hello IR,
thanks for replying. you are saying
even though phase might not be exactly 180 degree out of phase, if AM-AM
and AM-PM curves are good then it indicates linearity is good, regardless
of if phase is 180 degree out or not?
Top
IR
Post subject: Posted: Wed Aug 02, 2006 2:48 pm
Site Admin
Joined: Mon Jun 27, 2005 2:02 pm
Posts: 373
Location:
Germany
Hello again,
AM-AM and AM-PM curves are not an estimation
for good or bad linearity. You should consider linearity by measuring
the IM products. I was referring to the fact that the slopes of AM-AM
and AM-PM can have different slopes at the same time i.e. negative slope
and positive (Increasing) slope. The absolute value of the phase is
not relevant.
_________________
Best regards,
- IR
Top
ben00i
Post subject: Posted: Sat Mar 24, 2007
11:28 am
Lieutenant
Joined: Wed Mar 21, 2007 7:08
am
Posts: 1
I am measuring Avago's MGA 83563 medium power amplifier
at 2.45GHz. I got a negative AM-PM slope.
1) What does a positive
/ negative AM-PM slope tell me?
2) The absolute value as u mentioned
is important. I am confused as to what value of change would constitute
the PA device to having memory, quasi-memory or even memoryless. I think
this is a confusing point as there are papers out there that state AM-PM
measurement as a quasi-memoryless, memoryless or even a PA with memory.
Pls share. THanks.
Top
IR
Post subject:
Posted: Sat Mar 24, 2007 11:52 am
Site Admin
Joined:
Mon Jun 27, 2005 2:02 pm
Posts: 373
Location: Germany
The
negative or positive slope of the AM-PM is related to the transistor's
technology.
Memory effect can be caused (Most of) due to thermal
effects and to impedance variations at the frequency of the modulating
signal. The thermal effect causes the amplifier to act as a filter in
certain frequencies.
Here is a link that can be useful for you
which relates the causes and characteristics of memory effect based
on IM3 measurements.
http://herkules.oulu.fi/isbn9514265149/html/x1900.html
Hope this helps.
Posted
11/12/2012
|
|
|
|