Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronic Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

microwave power combiner/divider

Satellitmph2

Newbie level 4
Joined
Sep 16, 2019
Messages
7
Helped
0
Reputation
0
Reaction score
0
Trophy points
1
Activity points
57
Dear all
Hi,
I have simulated a two way power combiner, two inputs port1 and port 2 are excited by two excitations with the same amplitude and combined in outport port3, but the resuts of s21 and s31 hqve a positive values (+ 2dB).
My question is, can have a positive values for s21 and s31 in dB?

I check several times my promgram, I do not know where is the problem.
I have used the fdtd method for simulation.
i have extract s21 by a ratio of the incident waveform at port 1 and the transmitted waveform at port 2 ( Fourier transforms).

i will be very grateful for your help and advice.

Best regards
 

BigBoss

Advanced Member level 5
Joined
Nov 17, 2001
Messages
4,745
Helped
1,439
Reputation
2,876
Reaction score
1,301
Trophy points
1,393
Location
Turkey
Activity points
28,664
Combiners/Dividers are passive elements so they have losses, not gain..
How did you simulate his combiner?? Which EM simulator have you used ?? What is your port and substrate configuration ??
 

Satellitmph2

Newbie level 4
Joined
Sep 16, 2019
Messages
7
Helped
0
Reputation
0
Reaction score
0
Trophy points
1
Activity points
57
We have used fdtd
configuration
inputs: port1, port2
output: port3
RT 5880

I want to know if it's right or wrong: for the calculation of s21, the two ports port1 and port2 will be excited at the same time by two excitations and we calculate the fields?
 

BigBoss

Advanced Member level 5
Joined
Nov 17, 2001
Messages
4,745
Helped
1,439
Reputation
2,876
Reaction score
1,301
Trophy points
1,393
Location
Turkey
Activity points
28,664
We have used fdtd
configuration
inputs: port1, port2
output: port3
RT 5880
I want to know if it's right or wrong: for the calculation of s21, the two ports port1 and port2 will be excited at the same time by two excitations and we calculate the fields?
Ports are not excited simultaneously by definition of s-parameters.The simulator does not do too..
If this structure is a "Planar Wilkinson Power Divider", FDTD is a wrong choice for these type of problems.I mean, it's not wrong but it's more complicated than ought to be..
MoM or FEM Methods are very convenient and easy to configure.
 

Satellitmph2

Newbie level 4
Joined
Sep 16, 2019
Messages
7
Helped
0
Reputation
0
Reaction score
0
Trophy points
1
Activity points
57
Ports are not excited simultaneously by definition of s-parameters
what do you mean about this?

We have excited both ports port1 and port2, they are combined in port 3, the problem is how to extract scattering parameters?

Best regards.
 

Satellitmph2

Newbie level 4
Joined
Sep 16, 2019
Messages
7
Helped
0
Reputation
0
Reaction score
0
Trophy points
1
Activity points
57
Big Boss thank you for your reply,

we have already finished our program with the fdtd method, it's part of my project.
yes power comb/div are passives components, but our results of s21 varies between + 2dB and -3 dB
s21 can have a positives and negatives values (between +3 dB an -3 dB)?

Best regards
 

BigBoss

Advanced Member level 5
Joined
Nov 17, 2001
Messages
4,745
Helped
1,439
Reputation
2,876
Reaction score
1,301
Trophy points
1,393
Location
Turkey
Activity points
28,664
Big Boss thank you for your reply,

we have already finished our program with the fdtd method, it's part of my project.
yes power comb/div are passives components, but our results of s21 varies between + 2dB and -3 dB
s21 can have a positives and negatives values (between +3 dB an -3 dB)?

Best regards
The results are wrong..1/2 Power Divider must have at least -3dB Insertion Loss..
 

Satellitmph2

Newbie level 4
Joined
Sep 16, 2019
Messages
7
Helped
0
Reputation
0
Reaction score
0
Trophy points
1
Activity points
57
Thank you BigBoss for your reply
For the power divider we have less -3dB as you say, but for power combiner (two input port1 and port2 ,one output port3) we have s21 and s31 varies between -3 dB and +2 dB, we have problem with power combiner, when we excite two input ports.

Best regards
 

cariban

Full Member level 1
Joined
Aug 3, 2018
Messages
97
Helped
28
Reputation
56
Reaction score
28
Trophy points
18
Activity points
721
Depends on the correlation coefficient of your two inputs, you may get S31/S32 higher than 0dB. For example, if the combiner is ideal and you have exactly the same signal x(t) at port 1 and port 2, then port 3 outputs sqrt(2)*x(t). In this case, you will get
S32=S31=3dB.

In general, the output power is:
Pout = (P1 + P2)/2 + r(x1(t),x2(t)) * sqrt(P1*P2), where r(x1(t),x2(t)) is the correlation coefficient.

If x1(t) = -x2(t), r(x1(t),x2(t)) = -1, Pout=0, S31=S32=0, all the power is lost in the combiner.
 
Last edited:

volker@muehlhaus

Advanced Member level 5
Joined
Apr 11, 2014
Messages
2,483
Helped
976
Reputation
1,954
Reaction score
957
Trophy points
113
Activity points
14,708
We have excited both ports port1 and port2, they are combined in port 3, the problem is how to extract scattering parameters?
Your method is wrong. For S-parameters, you need to do individual port excitation. FDTD excitation at two ports simulatenously is wrong if you want S-parameters.
 

cariban

Full Member level 1
Joined
Aug 3, 2018
Messages
97
Helped
28
Reputation
56
Reaction score
28
Trophy points
18
Activity points
721
Your method is wrong. For S-parameters, you need to do individual port excitation. FDTD excitation at two ports simulatenously is wrong if you want S-parameters.
Agree with this. The unused port should be terminated.
 

Satellitmph2

Newbie level 4
Joined
Sep 16, 2019
Messages
7
Helped
0
Reputation
0
Reaction score
0
Trophy points
1
Activity points
57
At this case, if I do individual port excitation, we will not have a power combiner, the input port will be divided in two other ports (power divider).

Thank you very much for your reply.

Best Regards

- - - Updated - - -

At this case, if I do individual port excitation, we will not have a power combiner, the input port will be divided in two other ports (power divider). how to exctract scattering parametters for power combiner for two inputs ports and one output port.

Thank you very much for your reply.

Best Regards
 

volker@muehlhaus

Advanced Member level 5
Joined
Apr 11, 2014
Messages
2,483
Helped
976
Reputation
1,954
Reaction score
957
Trophy points
113
Activity points
14,708
At this case, if I do individual port excitation, we will not have a power combiner, the input port will be divided in two other ports (power divider).
For full 3-port S-parameters, it is always a "black box" 3-port device, and you always excite one port after another.

It seems that you are looking at one special excitation only, for the power combiner mode, so that you get 2-port S-parameters only. That is possible, but your passive device must NEVER have >0dB transmission. It seems your source definition is wrong, and your total input power is 2 * 100% which leads to the wrong +3dB transmission.
 

cariban

Full Member level 1
Joined
Aug 3, 2018
Messages
97
Helped
28
Reputation
56
Reaction score
28
Trophy points
18
Activity points
721
At this case, if I do individual port excitation, we will not have a power combiner, the input port will be divided in two other ports (power divider).

Thank you very much for your reply.

Best Regards

- - - Updated - - -

At this case, if I do individual port excitation, we will not have a power combiner, the input port will be divided in two other ports (power divider). how to exctract scattering parametters for power combiner for two inputs ports and one output port.

Thank you very much for your reply.

Best Regards
The definition of s-parameter is based on excitation at each port. Of course you can excite two ports and the s-parameter based on linear model is still correct, but you have to consider how they are summed together at the output port.
 

stark21

Newbie level 2
Joined
Jul 11, 2017
Messages
2
Helped
0
Reputation
0
Reaction score
0
Trophy points
1
Activity points
21
Dear Dr. Volker,

Thank you for your inputs. For simulating such passive transmission line combiners/ splitters at frequencies greater than 100 GHz, which EM software do you recommend? How has your experience been with ADS momentum? I have heard mixed reviews about it.

Which mesh settings (cells/wavelength) in Momentum would you recommend for such structures? Which other setting should I worry about to get accurate results? How would RF Momentum fare as compared to Microwave Momentum at these frequencies? I know that Microwave Momentum simulator is more accurate but it takes reasonably longer simulation time as compared to former case. Since my passive structure with the ground plane is 200um X 200um with no radiation, can I rely on the results provided by RF Momentum above 100 Ghz?

Looking forward to your valuable inputs.

Thanks.
 

Toggle Sidebar

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Top