微波EDA网,见证研发工程师的成长!
首页 > 研发问答 > 微波和射频技术 > 天线设计和射频技术 > [ASK] Bandpass filter Frequency shift after layout and schematic co-simulation in ADS

[ASK] Bandpass filter Frequency shift after layout and schematic co-simulation in ADS

时间:04-07 整理:3721RD 点击:
Hello.

I've made an bandpass filter in schematic and the frequency responce is good (left picture in attachment pic) Spec = f0=145.95 MHz, BW = 20 MHz, Type:butterworth.

But when i simulate the transmission line as shown in the pic in right side with fr4, er=4.2, thickness=1.6mm, dan tand=0.02. I get frequency shifting in ADS simulation. Why is this happen ? what can i do to minimize the transmission line effect in ADS ?

can anyone help me ?

thanks.

Keep your lines as short as you can, or include them in your design.

The 3nH inductor LP3 is very sensitive to extra length in that path. I would implement that as a piece of line (a few mm will give 3nH). Do not forget that the via to ground is also part of that inductor.

Another important aspect are the parasitics in your SMD elements. From what I have seen, you have simulated with ideal L/C so far. Have a look at the equivalent circuit models and/or *.s2p models from the component manufacturer.

In the posting https://www.edaboard.com/thread211699.html, a coupled resonator filter was suggested instead. This design is more robust against inductor parasitics.

Some of your components (like 3.7 nH coil) are a bit unrealistic at 145 MHz. The unload Q will be bad. The PCB short transmission lines at 145 MHz are pretty much just contributing additional stray capacitance to ground, the most damaging being the 0.7 pF or so PCB stray to ground between your series 1.8 pF/644 nH coupling resonators. If you design for higher termination impedance you will get more realistic component values.

14% bandwidth is achievable with just capacitance top coupled parallel tanks. It is near the limit but do-able.

Your model does not include effects of Q of components, primarily the coils, which will change the design quite a bit. A tank with 3.7 nH and 325 pF resonanting capacitance will have a very low unloaded Q, probably in range of 20 to 25. Probably below the minimum necessary to make the five pole filter design.

Your layout is completely wrong..

上一篇:Sinusoidal Inductor on PCB
下一篇:最后一页

Copyright © 2017-2020 微波EDA网 版权所有

网站地图

Top