50 ohm line simulation

Hi,
when i do the cpw antenna structure simulation i got the returnloss as -45db. but if we calculate the char. impedance using the basic cpw formulae it shows 64 ohm. will i face any problem when i go for fabrication?

with thanks,

kbmani

Hi, Kbmani:

It may or may not. Certainly, if you connect your CPW to a 50-ohm line, there might be some difference in return loss. The difference should be small normally. However, it does not exclude the possibility with bigger difference.

I do suggest you to optimize the Zc of your feed line to the antenna to the designed Zin if you are going to connect the antenna to the feed network with Zc = Zin. In this way, it will reduce loss of accuacy due to uncertainty in the connection.

Regards,

hi jian,

thanks for ur immediate reply. for simulation we are giving 50 ohms wave port only right. but in returnloss if we get -45db, then there won't be a problem?


should i remodify my circuit with 50 ohms impedance.? i mean the width and length should be readjusted?


with thanks,

kbmani

Added after 9 minutes:

hi jian,
if i change the metal thickness the char impedance is coming exactly 50 ohms. is the metal thickness play a main role in grounded cpw?


with thanks,

kbmani

Hi, kbmanick:

If you use 50-ohm normalized extension port to solve your antenna and you get return loss of -45 dB, it means that your antenna is very good matched to 50-ohms.

(1) If you attach an additional line of the same cross-section (Zc=50-ohms), the return loss should not change theoretically. In practical applications, the return loss will change very little. The above fact is true for other ohms if you match the Zin and the Zc.

(2) If you attach an additonal line of different cross-section but the same Zc, the return loss can change because of the discontinuities in the connection due to the difference in the cross-section. Normally, the change should be small.


(3) If you use 64-ohm Zc at the feed of the antenna and you get return loss of -45 dB (normalized to 50-ohms), and you attach a 50-ohm feed line to it, there will be some change in the return loss too. Again, it is due to the discontinuities in the conneciton due to the difference in the cross-section. The change can be small unless in special case.

What I would suggest you to do is the following:

(1) You first optimize your feed line to 50-ohms.
(2) You optimize your antnena to Zin = 50-ohms using this feed line.

In such a way, you should be able to preserve the best accuracy. You should try to use the feed line with the Zc = Zin and the Zin is the input impedance of the antenna you desire.

Regards.

Respected Sir,

there is an option in the toolbox.---50 ohms for waves---if i use that i could not give negative port.


with thanks,

kbmani.

Hi, Kbmanick:

It is an old style port implemented in IE3D. Plese do no tuse it. The best one is the Advanced Exension port. It is a the newest scheme and it has been working excellently in the last a few years.

Best regards,

hi jian,

normally the prts are normalized to 50 ohms right? then i think there will not be any problem. give your feed back.

Hi, kbmanick:

Yes. Normally, every extension port normalized to 50-ohms should give you good results. However, each scheme is implemented differently. It may yield slightly differently results at some extremes (high or low frequency). For the wave scheme, it is using wave concept. It normally is good at higher frequency. However, it may not be very stable at low frequency. Also, for highly lossy transmission lines such as those in RFIC, waveguide heory may not be precise (Please read the Appendix on complex Zc of IE3D User's Manual). wave scheme is also not as good as the Advanced Exension scheme on IE3D.

On IE3D, there are two more extension schemes (Advanced Extension and Extension for MMIC). Both of them are very flexible. However, the Extension for MMIC scheme or the 1st implementation will degrade at high frequency when dispersion is high. The Advanced Extension scheme is the latest addition a few years ago. It has been yielding excellent results from low to high frequency. Both schemes are using voltage and current. The only difference is at how we calculate the voltage, which can be different at high frequency for dispersive systems. The scheme implemented into Advanced Extension scheme yields better results from our experience of the last a few years.

Best regards,