[SOLVED] losses in microstrip line

[Manjuladevi]

HI all, I'm a researcher, i'm getting an error, when i was simulate 100 mm,er-2.55,-50 ohm transmission line from (500 Mhz to 6 GHz), im getting an insertion loss for 2.5 GHz will be -0.1 dB but practically i haven't getting that result. I got -0.35 db. what is the loss behind in it, and also i have cross verified calibration and SMA soldering.
please help me!

 

[vfone]

100mm (10cm) is a long microstrip circuit working at 2.5GHz. Small variations (from the simulated ones) of dielectric constant, dissipation factor, substrate height, trace width, strip discontinuities, would affect the final instertion loss.

I don't know what substrate you are using, but if is about Arlon AD255, this one should be a good quality substrate, with very small variations of parameters from low frequencies up to 10GHz.
At 500MHz and at 6GHz, how is the insertion loss compared to simulation?

 

[volker@muehlhaus]

Is that really loss (dissipation) or just reflected?
What is your S11?

 

[BigBoss]

Another additional error mechanism is -seen frequently- Manufacturing Errors.

 

[biff44]

Is that really loss (dissipation) or just reflected?
What is your S11?

exactly, what are s11 and s22 in your test?

why don't you measure 15 cm of line instead, where the loss will not be masked by other minor perturbations

 

[Manjuladevi]

100mm (10cm) is a long microstrip circuit working at 2.5GHz. Small variations (from the simulated ones) of dielectric constant, dissipation factor, substrate height, trace width, strip discontinuities, would affect the final instertion loss.

I don't know what substrate you are using, but if is about Arlon AD255, this one should be a good quality substrate, with very small variations of parameters from low frequencies up to 10GHz.
At 500MHz and at 6GHz, how is the insertion loss compared to simulation?

I have compared S21 from simulated and measured from VNA

 

[ktr]

To check for manufacturing errors, you may need to inspect the board with a camera system. Check that the traces have the correct width as intended; sometimes when traces are long; width of the line may change.

 

[Manjuladevi]

To check for manufacturing errors, you may need to inspect the board with a camera system. Check that the traces have the correct width as intended; sometimes when traces are long; width of the line may change.

how traces related with width..width is mainly depend on impedance but traces are not like that

 

[Manjuladevi]

100mm (10cm) is a long microstrip circuit working at 2.5GHz. Small variations (from the simulated ones) of dielectric constant, dissipation factor, substrate height, trace width, strip discontinuities, would affect the final instertion loss.

I don't know what substrate you are using, but if is about Arlon AD255, this one should be a good quality substrate, with very small variations of parameters from low frequencies up to 10GHz.
At 500MHz and at 6GHz, how is the insertion loss compared to simulation?

Directly i'm taking s2p plots mapped to xcel and make it as a comparison from 500M to 2.5G