Lossy Transmission Line Model (LTRA) |
The uniform RLC/RC/LC/RG transmission line model (referred to as the LTRA model henceforth) models a uniform constant-parameter distributed transmission line. The RC and LC cases may also be modeled using the URC and TRA models; however, the newer LTRA model is usually faster and more accurate than the others. The operation of the LTRA model is based on the convolution of the transmission line's impulse responses with its inputs.
.MODEL <model name> LTRA [model parameters] |
The LTRA model takes a number of parameters, some of which must be given and others that are optional. In the model must be specified the length of the line and the values of R, L, G and C per unit length.
Parameter | Description | Units | Default | Example |
R | resistance/length | ohms/m | 0 | 0.2 |
L | inductance/length | henrys/m | 0 | 9.13e-9 |
G | conductance/length | mhos/m | 0 | 0.0 |
C | capacitance/length | farads/m | 0 | 3.65e-12 |
LEN | length of line | m | 1.0 |
The following parameters are optional and affect the accuracy of the results and the speed of the simulation.
Parameter | Description | Units | Default | Example |
REL | breakpoint control | 1 | 0.5 | |
ABS | breakpoint control | 1 | 5 | |
NOSTEPLIMIT | Don't limit timestep to less than line delay | Flag | not set | |
NOCONTROL | Don't do complex timestep control | Flag | not set | |
LININTERP | Use lineair interpolation | Flag | not set | |
MIXEDINTERP | Use lineair when quadratic seems bad | Flag | not set | |
COMPACTREL | Special reltol for history compaction | RELTOL | 1.0e-3 | |
COMPACTABS | Special abstol for history compaction | ABSTOL | 1.0e-9 | |
TRUNCNR | Use Newton-Raphson method for timestep control | Flag | not set | |
TRUNCDONTCUT | Don't limit timestep to keep impulse-response errors low | Flag | not set |
The following types of lines have been implemented so far:-
Any other combination will yield erroneous results and should not be tried. The length LEN of the line must be specified
NOSTEPLIMIT is a flag that will remove the default restriction of limiting time-steps to less than the line delay in the RLC case.
NOCONTROL is a flag that prevents the default limiting of the time-step based on convolution error criteria in the RLC and RC cases. This speeds up simulation but may in some cases reduce the accuracy of results.
LININTERP is a flag that, when specified, will use linear interpolation instead of the default quadratic interpolation for calculating delayed signals.
MIXEDINTERP is a flag that, when specified, uses a metric for judging whether quadratic interpolation is not applicable and if so uses linear interpolation; otherwise it uses the default quadratic interpolation.
TRUNCDONTCUT is a flag that removes the default cutting of the time-step to limit errors in the actual calculation of impulse-response related quantities.
COMPACTREL and COMPACTABS are quantities that control the compaction of the past history of values stored for convolution. Larger values of these lower accuracy but usually increase simulation speed. These are to be used with the TRYTOCOMPACT option, described in the .OPTIONS section.
TRUNCNR is a flag that turns on the use of Newton-Raphson iterations to determine an appropriate timestep in the timestep control routines. The default is a trial and error procedure by cutting the previous timestep in half.
REL and ABS are quantities that control the setting of breakpoints.
The option most worth experimenting with for increasing the speed of simulation is REL. The default value of 1 is usually safe from the point of view of accuracy but occasionally increases computation time. A value greater than 2 eliminates all breakpoints and may be worth trying depending on the nature of the rest of the circuit, keeping in mind that it might not be safe from the viewpoint of accuracy. Breakpoints may usually be entirely eliminated if it is expected the circuit will not display sharp discontinuities. Values between 0 and 1 are usually not required but may be used for setting many breakpoints.
COMPACTREL may also be experimented with when the option TRYTOCOMPACT is specified in a .OPTIONS card. The legal range is between 0 and 1. Larger values usually decrease the accuracy of the simulation but in some cases improve speed. If TRYTOCOMPACT is not specified on a .OPTIONS card, history compaction is not attempted and accuracy is high. NOCONTROL, TRUNCDONTCUT and NOSTEPLIMIT also tend to increase speed at the expense of accuracy.
.MODEL LOSSY LTRA R=0.99 G=0 L=13.56E-9 C=1.323E-12 LEN=1