I am using THz time-domain spectroscopy to characterize my waveguide, I am getting absorption and tranmisttance values from the THz setup PC. I am getting S21 by taking the log of transmittance. I am confused about S11. if I am using the R=1-T-A formula, and taking the log of R, then I am getting very poor S11. Can anyone guide me in this case.
What do you mean by poor S11? Ideally the log of S11 is minus infinity - this means no reflection.
How are you separating loss due to S11 from the absorbtion measurement?
Malcolm White Thank you for your response.
I have taken the Transmittance (T) and Absorption (A) in linear form (0 to 1) from THz time-domain spectroscopy.
I want to extract the S-parameters from the above parameters.
For S21, I am using 10*log(T), so I am getting a similar response as in the simulated one.
For S11, I am using the formula R=1-T-A (linear) and then I am using 10*log(R) to get S11 in dB but in this case, S11 is too different from the simulated one.
Simulated S11 is almost below -80dB, while from the above calculation S11 obtained is around -12 dB.
can you guide me on how to get S11 and S21 from Transmittance (T) and Absorption (A) or it is not possible to get S11 from THz time-domain spectroscopy data?
Thanks
It would be very surprising to get S11 of -80dB. This is usually only possible if you can fine tune a system. To get -80 dB you need your measurement of T+A to be accurate to 1 part in 100 million. I would expect them to be accurate to 1% or worse. If you really do expect excellent S11 then the result will be entirely dominated by the error or uncertainty in the measurement of T and A. -12 dB seems much more likely.
Malcolm White
Thanks for your guidance,
I am using a circular metallic waveguide and simulating it in CST MW Studio.
and simulated S11 is around -80 dB.
while in the measurement process (THz TD Spectroscopy), I got T and A. I want S11, therefore, I used the formula.
R=1-T-A
and for getting S11, I am taking log(R).
My main question is whether S11 around -12 dB is considerable or my waveguide is too lossy. is there too much difference in simulated and measured S11 or I am making any mistake in the calculation?
Thanks
How accurate is your measurement of A and T? Do you really think you can measure these to 1 part in 100000000?
If you have errors in these then they add together in your estimate of S11.
Malcolm White
Thanks.
The measured A was not stable, so I focused on only T which is between 0 and 1 as it should be.
I have measured the transmittance (T) of a waveguide between 100 and 500 GHz. I am getting values of T between 0 and 0.7 throughout this frequency range. I want to calculate the reflection coefficient (S11) and insertion loss (S21) of the waveguide.
I think that I can calculate S21 simply by taking 10*log(T). However, to calculate S11, I think that I need to know the absorption (A) and reflectance (R) of the waveguide.
To calculate A, I used the following formula:
A = log(1/T)
However, when I calculate A for values of T less than or equal to 0.1, I get values of A greater than 1. This is not possible, because A cannot be greater than 1.
I am also using the following formula to calculate R:
R = 1-T-A
However, I cannot calculate R using this formula, as A is above 1 at certain frequency points.
So, my question is:
- Am I using the correct method and equations to calculate S11 and S21?
- Why am I getting values of A greater than 1 for values of T less than or equal to 0.1?
If R=1-T-A this is using conservation of energy. All the numbers should be in the range 0 to 1. They are not logs, they are proportions.
The reflected power added to the absorbed power added to the transmitted power should equal the incident power.
You need to measure the absorption. You cannot calculate it if you don't know S11 (R). If You use the formula A=1-T this assumes R = 0.
Don't use logs unless you want ratios. The formula R=1-T-A doesn't work for logs. If 10% is reflected and 10% absorbed then 80% will be transmitted, 10% =100%-80%-10%. If you log this it turns into -1=0-(-.1)-(-1) which is totally wrong.
You have calculated/simulated/predicted that S11=R=10-8. If that is right then very nearly A=1-T. You only know A to the accuracy with which you know S11 and T. If T is 0.1 then A is 0.9-10-8
Arslan Ahmed Sohoo Can you elaborate how does your experimental setup measures absorption? Are you using a VNA?
Mentes Jose, I am using THz time-domain spectroscopy to get the absorption.
- Badges
- Science topic
- Similar topics
- Instrumentation Engineering
- Instruments