joeydelp / josim

Superconductor Circuit Simulator

JoSIM

Superconductor Circuit Simulator

---

Full Documentation

Please see full documentation here: JoSIM Documentation

Overview

JoSIM was developed under IARPA contract SuperTools(via the U.S. Army Research Office grant W911NF-17-1-0120).

JoSIM is a SPICE syntax circuit simulator specifically created to handle superconducting elements such as the Josephson junction. It reads in a standard SPICE deck, creates an A matrix and solves the linear algebra problem Ax=b. The linear algebra package KLU by Tim Davis is used to solve the system of equations.

A two stage distribution platform is implemented in JoSIM with the first being a command line interface (CLI) binary that is always built as standard during compilation. This is accompanied by a C++ library (libjosim).

JoSIM allows output of results in various formats such as comma seperated value (CSV) or raw SPICE output. JoSIM has inherent support for .PARAM commands that allow components to have variable values as well as expression parsing. JoSIM implements the RCSJ model of the Josephson junction and only supports transient analysis at present.

JoSIM takes a .cir file as input and produces a .dat/.csv file as output.

Prebuilt binaries can be found here:

https://github.com/JoeyDelp/JoSIM/releases

Referencing:

JoSIM — Superconductor SPICE Simulator

---

Changelog

v2.6.9

• Fixed a bug resulting in bad optional access when shorting transmission line components.

v2.6.8

• Add a fix to warn of subcircuit redefenitions. Redefintions now overwrite the previous definition.
• Add a fix when parameters are used in subcircuits but defined within the global scope.

v2.6.7

• Added ability to parameterize time points specified for .tran and pwl commands.

v2.6.6

• Fixed a bug in noise source generation when noise step is less than or equal to the simulation step.
• Reduced aggressiveness of timestep scale down when close to transmission line step delay size. Now only scales down if TX delay is less than 1 step (previously 4)
• Added output dimensions option to josim-plot script when writing directly to PDF, SVG, EPS, etc.

v2.6.5

• Updated the CPR specification in the JJ model to allow multiple simulataneous harmonics.
• Added a Notepad++ UDL syntax highlighting script to the scripts folder.

v2.6.4

• Added the ability to specify CPR in the JJ model to enable non-sinusoidal CPRs needed for pi-junctions.

v2.6.3

• Fixed a bug where Windows and CentOS parameters do not coincide. Issue was due to expression substitution.

v2.6.2

• Fixed sp_generator.py having invalid tab spacing
• Fixed Area= and Ic= JJ parameters not producing the same result for the same critical current

v2.6.1

• Fixed an issue with custom waveform where it wasn't reading from the file properly
• Added spline method for cubic interpolation of custom waveform
• Moved entire codebase to 64bit integers to allow larger/longer simulations
• Fixed parameters
• Fixed transmission line integer step delay rounding issue that caused an error build up

v2.6

• Full Johnson-Nyquist temperature noise for Resistors and Resistive branch of RCSJ model
• Phi-junction model with verified Pi junction results
• Various bug fixes and code clean up

v2.5.8

• Updated KLU to a 64-bit version capable of handling non-zero count larger than the maximum value of a 32-bit integer.
• Included SuperLU 64-bit that can be activated using the -x 1 switch (default 0 for KLU).
• Bug fixes to pulse command.

v2.5.7

• Fixed an issue with comparison of double values for small numbers between various compilers.

v2.5.6

• Fixed a initial values when converting between phase and voltage when simulation output does not start at 0.

v2.5.5

• Fixed a bug in output conversion when plotting voltage in a phase simulation and vice versa.

v2.5.4

• Fixed comparison function for GCC interpretation of double values

v2.5.3

• Fixed an issue with the pulse command producing incorrect pulses
• Fixed an issue where timesteps smaller than 10fs were causing out-of-range issues
• Various other bug fixes

v2.5.2

• Fixed a major issue with relative paths not being correctly evaluated causing files to not be found even though the correct path was specified.

v2.5.1 - 18/05/2021

• Added .spread command. See documentation for more details.
• Added .file command. See documentation for more detauls.
• Added .iv command. See documentation for more details.
• Fixed a bug where .neb value was not being read properly
• Fixed a bug where pulse source type was outputting incorrect number of pulses.
• Fixed a bug where custom waveform was not allowing cubic interpolation method.
• Code cleanup.

v2.5 - 05/03/2021

• Phase mode is now the standard analysis mode. Produces the exact same results as voltage.
• Trapezoidal method has been replaced by superior 2nd order Gear method.
• JoSIM now internally determines the convention. (-c option removed)
• New progress indicators added. Spawns a seperate non blocking thread.
• New verbose modes. 1: Circuit statistics. 2: 1 + Parameters and evaluated values. 3: 2 + Complete expanded netlist.
• New DC source type. I.e. V01 1 0 DC 5V
• JJ has new IC= parameters that supersedes AREA= but ultimately performs similar model parameter scaling.
• Fixed NOISE source type to have syntax NOISE(AMP TD TSTEP)
• New '.temp' and '.neb' commands for noise thermal noise analysis
• Transient simulation command now allows PRSTART and PRSTEP commands. .tran TSTEP TSTOP PRSTART PSTEP
• Various memory improvements and overall speed-up
• Reduced output file size.
• Rewrote documentation: JoSIM Documentation
• Various bug fixes

v2.4.1 - 09/03/2020

• Fixed an issue where "GND" was not recognized as ground, creating a new node causing matrix singularity.
• Fixed an issue with multiport devices (TX, CCCS, CCVS, VCCS, VCVS) not adding the 3rd and 4th node to the node map.
• Fixed an issue with multiport devices not expanding properly if part of subcircuits.
• Various other bug fixes

v2.4 - 07/02/2020

• Complete rewrite of core internals
• Modularization of each component allowing easier implementation of integration methods
• Error reporting major improvment. Less random exits with no warnings. Better exits.
• Focus on object orientation for improved library access
• Speed up of matrix creation algorithm
• Improvements to output format
• Inclusion of '.include' control card
• Capability to read from standard in using '-i' command
• Various bug fixes

v2.3 - 10/05/2019

• Major changes to JoSIM internal engine
• Removal of plotting engines in favor of external tools such as Python
• Introduction of pyJoSIM and libjosim packages
• Immense speed-up of internals
• Various bug fixes

v2.2 - 12/02/2019

• Major rewrite of most of JoSIM code to accommodate library creation for use in C++ and Python
• Implementation of a phase source for use in phase-mode simulation

v2.1 - 16/01/2019

• Inclusion of the Haberkorn junction current which allows for user switching between ballistic and normal electron tunneling
• Various bug fixes

v2.0 - 10/10/2018

• New analysis method that allows Modified Nodal Phase Analysis (see ReadMe)
• Major memory improvement by not creating intermediate A matrix
• Various bug fixes

v1.3 - 28/07/2018

• Addition of mutual inductance with the netlist label K
• Addition of the sinusoidal source (Voltage and Current)
• Various bug fixes
• AQFP now works

v1.2 - 17/07/2018

• New Release brings RType=1 to the RCSJ model of the Josephson Junction.
• Various bugs have been fixed.
• Data structures have been majorly altered.
• Other changes and fixes

v1.1 - 04/06/2018

• Ability to change between plotting libraries (NONE, FLTK or Matplotlib [Default])
• Major bug fixes
• JoSIM now supports WRSpice syntax (mostly)
• Parameterization of component values with expression parsing
• Code cleanup and commenting

v1.0 - 03/05/2018

• Initial release of JoSIM
• Default FLTK plotting engine
• Replaces JSIM as default simulator

josim's Issues

JoSIM installation CentOS 7

Hello!

I did the installation according to your instructions. At the stage "cmake .." I had the first error:

At the stage "cmake --build . --config Release" I had the second error:

At the same time, I updated cmake and gcc:

The installation was performed on the CentOS 7 Minimal operating system.

Can you please tell me how this error can be fixed?

Issue with installation

Hi Joey,

I have tried installing the JoSIM, but the system displays the error in the stage of "Building CXX object CMakeFiles/josim.dir/src/Simulation.cpp.o".

I work on Ubuntu 22.04.1 LTS with gcc 11.3.0 and cmake 3.22.1. I also tried to install it with gcc 8 and specified that I use C++17: cmake .. -DCMAKE_CXX_FLAGS="--std=c++17"; but it didn't help.

Sincerely,
Emil

Python 3.6 DLL Files Missing

Dear Joey,

When I tried the Python 3.6 version, the below problem appeared:

Anaconda 3 Python Version 3.7 problem

Dear Joey,

I have a problem which is the Anaconda 3 Python 3.7. When I try to compile it is give me the below error.

Severity Code Description Project File Line Suppression State
Error LNK1181 cannot open input file 'python36.lib' JoSIM C:\JoSIM-master\src\josim_vs\LINK 1
Can you kindly advise how to fix this issue?

Thank you.

Best Regards
Hassan

Usage of custom wave source

I am having trouble using custom waveforms in v2.6 josim, please let me check if I am using it correctly.

Net list

R0    1    2   100
C0    2    0   100p
V0    1    0   cus(CUS.TXT 10p 1V 0)

.tran 0.1p 1000ps

.file out.csv
.print v(V0)
.print v(C0)
.print v(R0)

CUS.TXT

3 0 1 2 3 4 5 4 3 2 1 0 0 -1 -2 -3

I confirmed that it was working in v2.4. Was there a change in syntax?

Using josim has made my simulations 10x faster than jsim. thanks for making josim.

SQUID simulation

Is it possible to model a SQUID in JoSIM?

In particular how to introduce the externally applied magnetic field?

Windows: How to change the input path file?

Dear Joey,

Thank you for the tool. I want to know should we change the josim.vcxproj.user file to change the input path file? If yes, we should reopen the visual studio and reopen the solution to change the input file to the simulator?

Thank you.

Best Regards
Hassan

Parameter sweep simulations

Hi Joey,
A big thank you for creating JoSim!
I would like to ask you if it is possible in JoSim to perform simulations sweeping a parameter automatically.
In the documentation I see that with the .param command you can define a variable X and its value, but I don't understand if you can change the value of X in a range.

For example, I want to perform five transient analysis where I change the value of a resistor Rx each time in the range Rx = (1,2,3,4,5).
Can I do that using the .param command?
Best

Angelo

Question about AREA of junction.

Dear Joey,

Thank you for making josim. I don't quite understand what AREA means when using junction. Consider the following netlist statement.

.model jjmit jj(rtype=1, vg=2.8mV, cap=0.07pF, r0=160, rn=16, icrit=0.1mA)
B1 node1 node2 jjmit area=1.5

I know that AREA is a modifier of icrit, which means $Ic_{jj}=area \cdot icrit$. But is AREA just a modifier of icrit? what about cap, r0 and rn? Is AREA still a modifier of cap, r0 and rn? That is to say $Cap_{jj}=area \cdot cap, \quad r0_{jj}=area \cdot r0_{model}, \quad rn_{jj}=area \cdot rn_{model}$.

Thank you.

Best Regards

Mengfei Zhao

Error while building the JoSIM

Hi!

I am trying to install JoSIM on Linux mint but I have an error

In file included from /home/anatoly/Documents/JoSIM/include/JoSIM/BasicComponent.hpp:6:0, from /home/anatoly/Documents/JoSIM/include/JoSIM/Capacitor.hpp:6, from /home/anatoly/Documents/JoSIM/src/Capacitor.cpp:4: /home/anatoly/Documents/JoSIM/include/JoSIM/TypeDefines.hpp:8:20: fatal error: optional: No such file or directory compilation terminated. CMakeFiles/josim.dir/build.make:75: recipe for target 'CMakeFiles/josim.dir/src/Capacitor.cpp.o' failed make[2]: *** [CMakeFiles/josim.dir/src/Capacitor.cpp.o] Error 1 CMakeFiles/Makefile2:293: recipe for target 'CMakeFiles/josim.dir/all' failed make[1]: *** [CMakeFiles/josim.dir/all] Error 2 Makefile:145: recipe for target 'all' failed make: *** [all] Error 2

UPD: ok I see that the problem in c++17 libs. I changed everywhere optional to experimental\optional. However now I have an error AnalysisType.hpp:23:3: error: expression ‘<throw-expression>’ is not a constant-expression

Allow specifying the Stewart-McCumber parameter in JJ models

When working with overdamped Josephson junctions, it's often convenient to specify them in terms of their Stewart-McCumber parameter $\beta_c$ instead of their shunt resistance(s). Would it be a good idea to allow us to specify $\beta_c$ when declaring a model? $\beta_c$ is related to the other model parameters as:
$$\beta_c = \frac{2\pi C R^2 I_c}{\Phi_0}$$

The issues about ivcurve.py

Hello Joey,

When I use the ivcurve.py file, I need to supply it with two command-line arguments, but I don't know what those two arguments should be. As shown in the figure, what do argv[1] and argv[2] refer to? Did you tell me how to use this file correctly?

In addition to, I hope you can provide more examples for beginners and explain in more detail what each line in XXX.cir does or how it should be set up.

Thank you.

Best Regards

Peng Zhang

Parameter expression in PWL time value or .tran command?

Thanks for making JoSIM - it's great and has been super straight forward to get up to speed with!

However, I'm a bit confused on a "Cannot convert string to double" error I keep getting when using the .param command. I run into the issue when trying to define a .param expression for variables passed into PWL time points or .tran analysis tstep or tstop values. I've played around with the format that I'm using to define the parameter expression (i.e. "2m" vs "2e-3" vs "0.002" etc.), but nothing seems to work. Is it just not possible to define variables for these inputs?

Thanks!

Missing energy in a filter simulation

Hello,
I am trying to simulate frequency response of a filter with JoSIM by creating a Gaussian excitation and then doing Fourier analysis on the output. I saw that the S-parameters I calculate don't add up to 1 so I did the same time-domain based analysis using qucs-s and I can't get them matching. The results from qucs-s fit S-parameter simulation.

Here is a plot comparing the voltage at resistor P2:

And the power balance from JoSIM:

(where we expect a flat line at 1)

I feel I did double check my calculations but nonetheless can't find where could be a problem. Therefore it is not clear to me if I am using JoSIM correctly?

CC1 nodeinput midpoint 50f
LL1 nodeoutput midpoint 10n
LL2 nodeinput gnd 1n
LL3 nodeoutput gnd 1n
VIN nodeinput nodeinputR cus(waveform.txt 10p 1.0 2)
RP1 gnd nodeinputR  50
RP2 gnd nodeoutput  50
.tran 10p 25000p 
.print NODEV nodeinput
.print DEVI RP1
.print NODEV nodeoutput
.print DEVI RP2
.end

and the only CLI option I use in Python:
cli = CliOptions(["./josim_minimal.cir"])

The files I am using are packed here: josim_minimal_filter.zip

• JoSIM netlist
• Jupyter notebook with analysis
• waveform.txt that contains the gaussian pulse
• minimal_a1.csv, minimal_v2.csv results from qucs-s simulation (overlayed in the notebook)

Some issues about building this project

Environment: Win10+Visual Studio Community 2017 with the basic C++ for desktop installed

Here is my building result:
It says that it cannot find the file j_model.cpp.
Does this file exist? Or is this file user defined?

One more question, will the conversion warning have any influence on the usage?

Sorry for the disturbance.

1>------ Rebuild All started: Project: JoSIM, Configuration: DEBUG_None x64 ------
1>j_components.cpp
1>j_errors.cpp
1>j_input.cpp
1>josim.cpp
1>j_matrix.cpp
1>c:\users\tsyw\desktop\josim\josim-master\src\j_matrix.cpp(1353): warning C4267: '=': conversion from 'size_t' to 'int', possible loss of data
1>c:\users\tsyw\desktop\josim\josim-master\src\j_matrix.cpp(1354): warning C4267: '=': conversion from 'size_t' to 'int', possible loss of data
1>j_misc.cpp
1>c:\users\tsyw\desktop\josim\josim-master\src\j_misc.cpp(292): warning C4244: 'argument': conversion from 'double' to 'const unsigned __int64', possible loss of data
1>c:\users\tsyw\desktop\josim\josim-master\src\j_misc.cpp(334): warning C4244: 'argument': conversion from 'double' to '__int64', possible loss of data
1>c:\users\tsyw\desktop\josim\josim-master\src\j_misc.cpp(342): warning C4244: 'initializing': conversion from 'double' to 'int', possible loss of data
1>c:\users\tsyw\desktop\josim\josim-master\src\j_misc.cpp(376): warning C4244: 'initializing': conversion from 'double' to 'int', possible loss of data
1>c:\users\tsyw\desktop\josim\josim-master\src\j_misc.cpp(377): warning C4244: 'initializing': conversion from 'double' to 'int', possible loss of data
1>c:\users\tsyw\desktop\josim\josim-master\src\j_misc.cpp(411): warning C4244: 'initializing': conversion from 'double' to 'int', possible loss of data
1>c:\users\tsyw\desktop\josim\josim-master\src\j_misc.cpp(482): warning C4267: 'initializing': conversion from 'size_t' to 'int', possible loss of data
1>j_models.cpp
1>c1xx : fatal error C1083: Cannot open source file: '..\j_models.cpp': No such file or directory
1>j_output.cpp
1>j_parser.cpp
1>c:\users\tsyw\desktop\josim\josim-master\src\j_parser.cpp(27): warning C4267: '=': conversion from 'size_t' to 'int', possible loss of data
1>c:\users\tsyw\desktop\josim\josim-master\src\j_parser.cpp(32): warning C4267: '=': conversion from 'size_t' to 'int', possible loss of data
1>j_plot.cpp
1>c:\users\tsyw\desktop\josim\josim-master\src\j_plot.cpp(485): warning C4267: '=': conversion from 'size_t' to 'int', possible loss of data
1>c:\users\tsyw\desktop\josim\josim-master\src\j_plot.cpp(487): warning C4267: '=': conversion from 'size_t' to 'int', possible loss of data
1>c:\users\tsyw\desktop\josim\josim-master\src\j_plot.cpp(489): warning C4267: '=': conversion from 'size_t' to 'int', possible loss of data
1>j_simulation.cpp
1>c:\users\tsyw\desktop\josim\josim-master\src\j_simulation.cpp(78): warning C4244: 'initializing': conversion from 'double' to 'int', possible loss of data
1>c:\users\tsyw\desktop\josim\josim-master\src\j_simulation.cpp(235): warning C4244: '=': conversion from 'double' to 'int', possible loss of data
1>Generating Code...
1>Done building project "josim.vcxproj" -- FAILED.
========== Rebuild All: 0 succeeded, 1 failed, 0 skipped ==========

Can not apply two harmonic signals simultaneously

Dear Joey,

I am using JoSIM for JTWPA simulations, and for these purposes I need to apply two harmonic signals simultaneously. I am trying to do it the next way:

PSIG 0 1 sin (0 10 14G)
PPUM 0 1 sin (0 0.01 6.9G)

Here I tried to apply phase harmonic signals with different amplitudes with no biasing and frequencies at 14 and 6.9 GHz.

But unfortunately I am getting the next error:

Matrix is singular. Matrix will have no solution.
Please check the components in the netlist.
The program will abort.

I do not quite understand why it should cause a matrix with no solution, basically I just want to apply a sum of two harmonic signals. Can this problem be solved in the last version?

Also I would like to mention that it is not possible to do any math operation with outputs as far as I know. For example, if I want to look at the sum of currents through the dc SQUID, I can not just write

.print i(B01)+i(B02)

but I have to make the next output

.print i(B01)
.print i(B02)

and then sum it in OriginPro. It is not critical but causes light inconviniences.

Best regards,
Daniil

"cmake --build. --config Release" operation failed

Shorted transmission line causes bad_optional_access

Hello,
I've tried the following netlist:

CC1 nodeinput midpoint 50f
LL1 nodeoutput midpoint 10n
TLine1 nodeinput gnd gnd gnd TD=1n Z0=50
TLine2 nodeoutput gnd gnd gnd TD=1n Z0=50
VIN nodeinput nodeinputR cus(waveform.txt 10p 1.0 2)
RP1 gnd nodeinputR  50
RP2 gnd nodeoutput  50
.tran 10p 25000p 
.print NODEV nodeinput
.print DEVI RP1
.print NODEV nodeoutput
.print DEVI RP2
.end

where the transmission lines Line1 and Lin2 are shorted at one end. It leads to the following error:
RuntimeError: bad optional access
Adding a small resistor at the shorted node is a working hack, though I figured it would be good to let you know.

Simulation with noise

Hello

I am trying to use JoSIM to simulate Random Number Generator which relies on thermal noise.
I noticed that JoSIM use rand() to introduce Gaussian random number in noise source.
Does it mean that if I run simulation with a long period, it will show repeated behavior pattern in circuit since rand() gives a pseudo random number.

Besides, is it possible to run simulation with process variation in JoSIM?

Thanks

Subcircuit questions

Hi all, hope you're doing well. Was testing Josim (like the cmake build system!) and found some issues with subcircuits regarding input nodes labels and parameters.

1. Subcircuit input Node labels cannot be referenced in .print statements, e.g.:

X1 TEST 1 0
I1      0 1 1.0

.subckt TEST IN GND
    R1 IN GND 10.0
.ends

.tran 0.25p 5n 0 1.0p
.print v(IN.X1)
.end

yields

W: Controls
Unknown device/node IN|X1
Cannot store results for this device/node.
Ignoring this store request.

Expected behavior would be to look up the proper node label based on the input mapping. To be fair this is easy to work around, but is expected behavior for those coming from WRspice (IIRC).
2. Subcircuit parameters appear to be broken (unless I'm the broken one!), e.g.:

X1 TEST 1 0 RES=10.0
I1      0 1 1.0

.subckt TEST IN GND
    R1 IN GND RES
.ends

.tran 0.25p 5n 0 1.0p
.print v(1)
.end

yields

E: Parsing
The following variables were not found/defined.
Please ensure that these variables exist within the netlist.
Variables: 
RES

Another attempt, trying to interpret the manual's instructions for declarations:

X1 TEST 1 0 RES=5.0
I1      0 1 1.0

.subckt TEST IN GND RES=10
    R1 IN GND RES
.ends

.tran 0.25p 5n 0 1.0p
.print v(1)
.end

which yields

E: Input
The IO of line "X1" does not match the subcircuit IO.
Please check the line and ensure correct IO and that parameters do not contain spaces.

Expected behavior is to substitute the indicated LABEL=VALUE pairs into the subcircuit instantiation, as indicated in the manual and by custom. I'll try to dig through the source a bit, but wanted to bring this up. Thanks for pushing this project forwards!

Artifacting/error build-up on JJ in long simulations

Error build-up causing anomalous effects in the JJ after long simulations.

"cmake --build . --config Release" operation failed

Don't understand the component stamp

Hi Joey,

Thanks for making JoSIM.
In the paper, you wrote down the stamp for inductance like this,

The third row of the stamp is v=Ldi/dt. However, it seems the first two rows are I=0 and -I=0 which I don't know where they come from and what's the meaning of them.

Do I understand the matrix representation correctly?

add nonlinear inductor

Hi Joey

many thanks for your fantastic JoSIM. I am using it to simulate JTWPA.
It would be very nice to have the possibility to simulate nonlinear inductors of the type L(I) = Lo (1 + a I^2), to include kinetic inductance effects.
Do you think it is feasible?

many thanks

Sergio