| T O P I C R E V I E W |
| Max |
Posted - Nov 08 2013 : 14:08:18
Hi Enrico,
I have just used FastCap for a couple of days and found it really super: Fast and Stable. Earlier I was used to simulating in other software to get the capacitance matrix. And now I think it would be more convenient to do the work in FastCap/FastHenry.
In order to proceed with the results delivered by FastCap/FastHenry I would like to import them in Matlab. Thereby I have two questions for FastCap and FastHenry:
1. Accidently I found that the result delivered by FastHenry was not coincide with the inductance matrix that dumped in Zc.mat. (In the example below)
2. I didn't find the result of capacitance matrix delivered by FastCap in any output file. Does FastCap also write the capacitance matrix to an output file?
***********************Example *********************************
.Unit mm
.freq fmin=100000000 fmax=100000000 ndec=1
.Default h=0.001000 sigma=5.97e7
Ncond1_N1 x=0.000000 y=1.000000 z=1.200000
Ncond1_N2 x=1.000000 y=0.000000 z=1.200000
Ncond1_1 x=0.500000 y=0.500000 z=1.000000
Ncond1_2 x=1.000000 y=0.000000 z=1.000000
Ncond1_3 x=0.000000 y=0.000000 z=1.000000
Ncond1_4 x=1.000000 y=1.000000 z=1.000000
Ncond1_5 x=0.000000 y=1.000000 z=1.000000
Ncond1_6 x=1.000000 y=0.000000 z=1.200000
Ncond1_7 x=1.000000 y=1.000000 z=1.200000
Ncond1_8 x=0.000000 y=1.000000 z=1.200000
Ncond1_9 x=0.000000 y=0.000000 z=1.200000
Ncond1_10 x=0.500000 y=0.500000 z=1.200000
********************************* Edges *****************************
.Default w=0.151057
Econd1_1 Ncond1_1 Ncond1_2
Econd1_2 Ncond1_1 Ncond1_3
Econd1_3 Ncond1_1 Ncond1_4
Econd1_4 Ncond1_1 Ncond1_5
Econd1_5 Ncond1_2 Ncond1_3
Econd1_6 Ncond1_2 Ncond1_4
Econd1_7 Ncond1_2 Ncond1_6
Econd1_8 Ncond1_2 Ncond1_7
Econd1_9 Ncond1_3 Ncond1_5
Econd1_10 Ncond1_3 Ncond1_6
Econd1_11 Ncond1_3 Ncond1_9
Econd1_12 Ncond1_4 Ncond1_5
Econd1_13 Ncond1_4 Ncond1_7
Econd1_14 Ncond1_4 Ncond1_8
Econd1_15 Ncond1_5 Ncond1_8
Econd1_16 Ncond1_5 Ncond1_9
Econd1_17 Ncond1_6 Ncond1_7
Econd1_18 Ncond1_6 Ncond1_9
Econd1_19 Ncond1_6 Ncond1_10
Econd1_20 Ncond1_7 Ncond1_8
Econd1_21 Ncond1_7 Ncond1_10
Econd1_22 Ncond1_8 Ncond1_9
Econd1_23 Ncond1_8 Ncond1_10
Econd1_24 Ncond1_9 Ncond1_10
.Equiv Ncond1_N1 Ncond1_8
.Equiv Ncond1_N2 Ncond1_6
.external Ncond1_8 Ncond1_6
.end
It deliveres the result:
Row 0: ncond1_8 to ncond1_6
Freq = 1e+008
Row 0: 3.56655e-005+5.19667e-010j
_________________________________________________________
But in Zc.mat the result is:
Row 1: ncond1_8 to ncond1_6
Impedance matrix for frequency = 1e+008 1 x 1
3.56655e-005 +0.326517j
It's weird that the real part is correct but the imaginary part not.
Best regards,
Max |
| 2 L A T E S T R E P L I E S (Newest First) |
| Max |
Posted - Nov 18 2013 : 17:51:50
quote:
Originally posted by Enrico
1. the answer to your question can be found in FastHenry2's embedded help, talking about the result shown in FastHenry2's output window:
quote:
This is the actual impedance of the square loop, for each frequency. You can see that the result is in the R+jL form, that is, the real part is the resistance and the imaginary part is the inductance.
The results are also dumped to the file Zc.mat. Note however that in this file the results are in the R+j#969;L form, where #969; is the angular frequency, that is, 2#960;f and f is the frequency.
Please note that the original FastHenry for Unix did not dump the impedance matrix, but just created the Zc.mat file, that was read using the ReadOutput utility (available also for Windows, to be run in a DOS shell), ultimately providing the impedance in R+jL format, i.e. the same output format that today FastHenry2 directly provides.
2. No. However for both FastCap, FasterCap and FastHenry you can import the result matrices in another program running under Windows using Automation, see the embedded online help for each of the three tools for the supported methods and properites, and the Automation samples under the Samples directory for easy-to-use scripts for retrieving the data, as well as for running simulations from another application.
Just for reference, there is also a paper I happened to read recently, using exactly this methodology. I don't know if it matches your application, but could be interesting, see paragraph IV: P. Jacqmaer, J. Zwysen, R. Gelagaev, J. Driesen, Accurately modelling of parasitics in power electronics circuits using an easy RLC-extraction method, IEEE International Instrumentation and Measurement Technology Conference (I2MTC), 2012
Best Regards,
Enrico
Thanks very much for your help! This paper seems very interesting and It can help me a lot :) |
| Enrico |
Posted - Nov 11 2013 : 16:42:12
1. the answer to your question can be found in FastHenry2's embedded help, talking about the result shown in FastHenry2's output window:
quote:
This is the actual impedance of the square loop, for each frequency. You can see that the result is in the R+jL form, that is, the real part is the resistance and the imaginary part is the inductance.
The results are also dumped to the file Zc.mat. Note however that in this file the results are in the R+j#969;L form, where #969; is the angular frequency, that is, 2#960;f and f is the frequency.
Please note that the original FastHenry for Unix did not dump the impedance matrix, but just created the Zc.mat file, that was read using the ReadOutput utility (available also for Windows, to be run in a DOS shell), ultimately providing the impedance in R+jL format, i.e. the same output format that today FastHenry2 directly provides.
2. No. However for both FastCap, FasterCap and FastHenry you can import the result matrices in another program running under Windows using Automation, see the embedded online help for each of the three tools for the supported methods and properites, and the Automation samples under the Samples directory for easy-to-use scripts for retrieving the data, as well as for running simulations from another application.
Just for reference, there is also a paper I happened to read recently, using exactly this methodology. I don't know if it matches your application, but could be interesting, see paragraph IV: P. Jacqmaer, J. Zwysen, R. Gelagaev, J. Driesen, Accurately modelling of parasitics in power electronics circuits using an easy RLC-extraction method, IEEE International Instrumentation and Measurement Technology Conference (I2MTC), 2012
Best Regards,
Enrico
|
|
|
