完善资料让更多小伙伴认识你,还能领取20积分哦, 立即完善>
您好,我正在尝试使用FDTD模拟来模拟微带环谐振器。
为了达到理想的结果(我有电路的实际测量),我建立了一个密集的网格,包括电路的所有细节,以尽可能接近真实的测量。 问题是通过模拟获得的S参数图具有非常少的点。 谐振器频率是正确的(2 GHz和4 GHz),但图表看起来很可怕,只有这么几点。 我试图更改所有FTDT模拟选项但是没有办法设置将出现在图表上的点数,但是,我意识到点数与时间步长参数有关系。 当时间步长减小时,点之间的频率范围增加。 时间步长参数取决于网格中最小单元格的大小,因此我发现增加图形点数的唯一方法是减小单元格大小并减少一点精度。 我正在评估的频率范围是1GHz到5GHz,最小的单元尺寸是0.1x 0.1 x0.017 mm。 是否可以在不改变单元尺寸的情况下降低点之间的频率范围? 谢谢大家! 塞尔吉 以上来自于谷歌翻译 以下为原文 Hello, I'm trying to simulate a microstrip ring resonator using an FDTD simulation. In order to achieve the desired result (I have the real measures of the circuit) I've build a dense grid to include all the details of the circuit to approach as much as possible to the real measure. The problem is that the S-parameter graph obtained with the simulation has very few points. The resonator frequencies are correct (2 GHz and 4 GHz) but the graph looks horrible with such few points. I tried to change all FTDT simulation options but there is no way to set the number of points that will appear on the graph, however, I realised that the number of points has a relationship with the timestep parameter. When the timestep decreases, the frequency range between points increases. The timestep parameter depends on the size of the smallest cell of the grid, so the only way I've found to increase the number of points of the graph is decreasing the cell size and loosing a little bit of precision. The frequency range I'm evaluating is from 1GHz to 5GHz and the smallest cell size is 0.1x 0.1 x0.017 mm. Is it possible to decrease the frequency range between points without changing the cell size? Thanks all! Sergi |
|
相关推荐
2个回答
|
|
您好Sergi,当您从结果页面创建图表时,您将在设置对话框中看到的其中一个选项是“FFT大小”。
这可以控制正在转换为频域数据的时域数据的采样。 尝试增加此值,直到获得所需的频率分辨率。 请注意,您不希望将其设置得太高,因为创建绘图需要更长时间,并且您可能会过度采样。 我相信默认值是15(意味着2 ^ 15个样本)。 尝试将此值增加到18并查看是否可以改善您的情节。 希望这会有所帮助,马克 以上来自于谷歌翻译 以下为原文 Hi Sergi, When you are creating a graph from the Results page, one of the options you'll see in the setup dialog box is "FFT size". This controls the sampling of the time domain data that is being converted to frequency domain data. Try increasing this until you get the frequency resolution you want. Note that you don't want to set it too high because it will take longer to create the plot, and you may get oversampling. I believe the default is 15 (meaning 2^15 samples). Try increasing this to 18 and see if that improves your plot. Hope this helps, Marc |
|
|
|
非常感谢!
有效! 我不知道这个选项。 您诚挚的Sergi 以上来自于谷歌翻译 以下为原文 Thanks a lot! It worked! I was unaware of this option. Yours sincerely Sergi |
|
|
|
只有小组成员才能发言,加入小组>>
1291 浏览 0 评论
2375 浏览 1 评论
2194 浏览 1 评论
2065 浏览 5 评论
2952 浏览 3 评论
1123浏览 1评论
关于Keysight x1149 Boundary Scan Analyzer
758浏览 0评论
N5230C用“CALC:MARK:BWID?”获取Bwid,Cent,Q,Loss失败,请问大佬们怎么解决呀
934浏览 0评论
1294浏览 0评论
小黑屋| 手机版| Archiver| 电子发烧友 ( 湘ICP备2023018690号 )
GMT+8, 2024-12-26 05:27 , Processed in 1.387296 second(s), Total 81, Slave 64 queries .
Powered by 电子发烧友网
© 2015 bbs.elecfans.com
关注我们的微信
下载发烧友APP
电子发烧友观察
版权所有 © 湖南华秋数字科技有限公司
电子发烧友 (电路图) 湘公网安备 43011202000918 号 电信与信息服务业务经营许可证:合字B2-20210191 工商网监 湘ICP备2023018690号