|
dstraight Post subject: ADS using Harmonic Balance with mixers Posted:
Sun Oct 02, 2005 6:40 pm Captain Joined: Sun May 09,
2004 1:23 am Posts: 10 Location: Morgan Hill, CA (Silicon Valley,
Bay Area) Hello all, I am having trouble simulating a mixer
with ADS's Harmonic Balance. I would like to, very much, sweep the frequency
source feeding the RF port of the mixer while using a single frequency
source (non-swept) at the LO port. I desire to measure the IF output
over the swept frequency (or, quite possibly, frequencies higher than
the swept frequency depending upon which sideband I want to view at
the IF port). Although I have not successfully performed a sweep with
a second CW source using the Harmonic Balance engine I have been able
to simulate two CW sources in the RF and IF port of the mixer. Also,
I can sweep one source with the harmonic balance simulator, but it appears
that I cannot use a second source while doing so. For whatever reason,
things fall apart for me when I try to use two source with one being
swept. Any help would be much appreciated. I documented my issue
with ADS here (once you visit the link click on each icon to make the
picture larger and readable): https://www.pbase.com/dstraight/harmonic
... and_mixers I will also try to embed the same information
in this message (but if it doesn't work then use the link supplied above
- you should see four pages in this posting). If none of my links work
and you're interested then email me and I'll either post again or email
you back. Many thanks, David <img
src="https://www.pbase.com/dstraight/image/50162072/original.jpg" border="0"
/> <img src="https://www.pbase.com/dstraight/image/50162073/original.jpg"
border="0" /> <img src="https://www.pbase.com/dstraight/image/50162074/original.jpg"
border="0" /> <img src="https://www.pbase.com/dstraight/image/50162075/original.jpg"
border="0" /> Top dstraight Post subject:
Posted: Sun Oct 02, 2005 6:52 pm Captain Joined: Sun
May 09, 2004 1:23 am Posts: 10 Location: Morgan Hill, CA (Silicon
Valley, Bay Area) Let's try embedding those pictures again....
Here they are... Top pallavi_d
Post subject: Harmonic balance and ADSPosted: Thu Oct 06, 2005 4:58
am Captain Joined: Thu Oct 06, 2005 3:45 am Posts:
7 Location: India Hi David, If you want to sweep RF frequency
and keep LO Freq constant, you can just use the template provided in
ADS. There is an example of Gilbert Cell Mixer in ADS, you can just
use the sweptrf.dsn file and modify it according to your freqency range.
It should work , i have used it . But still if you ahve further problems
, do let me know, i will try to solve it. regards, Pallavi
Top dstraight Post subject: Re: Harmonic balance
and ADSPosted: Thu Oct 06, 2005 12:47 pm Captain Joined:
Sun May 09, 2004 1:23 am Posts: 10 Location: Morgan Hill, CA (Silicon
Valley, Bay Area) pallavi_d wrote: Hi David, If you want
to sweep RF frequency and keep LO Freq constant, you can just use the
template provided in ADS. There is an example of Gilbert Cell Mixer
in ADS, you can just use the sweptrf.dsn file and modify it according
to your freqency range. It should work , i have used it . But
still if you ahve further problems , do let me know, i will try to solve
it. regards, Pallavi Hello Pallavi,
Thanks for offering a helping hand. I use ADS a lot for s-parameter
simulations and occasionally delve into harmonic balance. However, I
find that I do not use harmonic balance very much because I am not so
familiar with how variables are used and how to set up many seemingly
useful features in the harmonic balance engine. Sometimes I look at
examples and sometimes that helps. I am late for work right now,
but I did look at the project (sweptrf.dsn). I can see that the project
runs correctly so then copied it into another directory and only added
a net called VLO at the port of the LO so that I could view signals
at this port. I ran the project again and then attempted to plot spectral
output of the LO and RF port of the mixer. What I found was inconsistent
power levels at these ports. The plots that I produced power (in dBm)
over swept frequencies and for each the LO and RF the levels were -50dBm
and below. I believe the LO should be -5dBm and the RF was something
higher than this. I didn't change much other than adding the
label (net) "VLO" yet the project seems to work correctly (aside from
the plots I tried to generate). Can you try what I did and let me know
if you see the same? I am a bit rushed right now so maybe I've missed
something. As I've mentioned I'm late for work and will not be
able to look at this more until about 10pm tonight. At that time if
I have the energy I'll look at this some more and perhaps post more
information using nice pictures. Many thanks, David
Top dstraight Post subject: Posted: Thu Oct 06,
2005 4:27 pm Captain Joined: Sun May 09, 2004 1:23 am
Posts: 10 Location: Morgan Hill, CA (Silicon Valley, Bay Area)
Okay, I decided to drive home for lunch so that I could gather a
few screen captures (I am lucky in that I live and work in the same
town). Below is a schematic of the example design (SweptRF.dsn)
from the Mixers_prj project and below this we see graphs of the resultant
simulation. As stated in my last post I've only added one net
at the LO port of the mixer and named it "VLO" then I produced a plot
of the VLO and Vrf (verses swept RF frequency, Frf) as shown in the
graphs below. The third screen capture shows the most detail.
As can be seen from these graphs the project appears to be running
correctly, but when I try to look at the signals going into the mixer
I do not see anything that would make sense. In the graphs RF power
should say -20dBm at 1GHz and LO power should read -5dBm at 1GHz-70MHz=
930MHz, but as the two graphs show LO power is below -50dBm and RF power
is below -60dBm. Perhaps the graphs that I have drawn take frequency
samples that do not sample exactly at 1GHz and 930MHz thus explaining
absence of power. I do not know.
Top pallavi_d Post subject: Harmonic balance for Mixers
in ADSPosted: Fri Oct 07, 2005 8:21 am Captain Joined:
Thu Oct 06, 2005 3:45 am Posts: 7 Location: India Hello David,
If you want to see the IF spectrum , just use the template Convgain.dsn
(again an example, but it helps) modify it according to the IF freq
you want and see the results. To sweep the RF freq , as i have
already told use sweptrf.dsn I also worked on it today(on sweptrf.dsn)
and if you want to see the spectrum at RF and LO port , only naming
the nodes as Vlo and Vrf won't work. In the data display window
write an equation: RF_spectrum=dBm(HB.Vrf) , similarly for LO
spectrum and you will see the plots over harmonic index. When
i plotted them over RF freq , i get to see the power level at RF port
at particular frequencies(it is constant as i haven't swept the rf power
) You can use all the example designs for mixer and modify it
according to the plots you want to see. I dont know if you are
satisfied with my answer, but still let me know your feedback.
regards, Pallavi Top dstraight Post subject:
Posted: Fri Oct 07, 2005 11:42 am Captain Joined: Sun
May 09, 2004 1:23 am Posts: 10 Location: Morgan Hill, CA (Silicon
Valley, Bay Area) Hello Pallavi, Many thanks taking the time
to help me with this. It is much appreciated. I am hoping to learn harmonic
balance well enough so that I can use it for work as there are many
features that seem very useful (such as NoiseCons for mixers). But,
one thing at a time. By suggestion I created two variables each
for the RF and LO and also created plots showing power verse harmonic
index. I can see that the power levels correspond well with set values
on the original schematic. This is probably a stupid question,
but how do I relate harmonic index to frequency? Also, I tried
placing a marker on some of these frequencies and I did notice that
the frequency of the swept RF "Frf" was displayed. As can be seen in
the graph multiple frequencies were lined up on top of each other (probably
due to sweeping, number of harmonics of each source and maximum order
to solve). I did notice that the power level was slightly higher
than expected. For example, I've placed the cursor showing -19.5dBm
(Frf=3GHz) at harmonic index 6. Likewise, -3.8dBm (Frf=3GHz) at harmonic
index= 5. I would have expected to see -20dBm and -5dBm respectively.
Do you have any idea why the power level is higher than what the
source says it should be? David Top
pallavi_d Post subject: Harmonic Balance in ADSPosted: Sat
Oct 08, 2005 2:56 am Captain Joined: Thu Oct 06, 2005
3:45 am Posts: 7 Location: India Hi David, I have no
problem spending time on your query as i am myself working on MMIC mixer
in ADS,it is also helpful for me. Ok , the relation between frequency
and harmonic index is: Freq= c1*f1+c2*f2+c3*f3+............+cn*fn
where c1,c2........cn are harmonic index I am also giving you
a link which gives u the pdf . maybe u have read it , if not just
go through it , it will be helpful xanadu.ece.ucsb.edu/~long/
ece145a/ADS_Harmonic_Balance_w03.pdf Regarding the power level
its showing , it again has to do with harmonic index.This article can
provide the answer. regards, Pallavi
Posted 11/12/2012
|