I want to simulate an RF balun response to a DC offseted RF signal.
Started with a pure balun, made from 3 inductors and mature inductance among them.
Very odd.
What happens if you short-circuit R1?
What will happen if you put a constant voltage onto an inductor, which does not have any series resistance? Did you measure the current?
What will happen when adding a resistor into the circuit loop?
Spice does not accept 0 Ohm resistors, as this would result in an infinite conductance in its circuit matrix. in ngspice they are replaced by a 1m Ohm resistor.
If I short R1, it come back to the same simulation result as the version do not have R1.
The reason I put R1 there is trying to be the source resistance of the V1, and try to see the impedance ratio of R3 convert back to the prime side.
@holger Thank you for the help.
I checked the current and believe that is the problem as the current all going really high.
If I add a 1 ohm series resistor to the inductor, it start to make some sense.
But the the question come is how should I simulate a differential 100 ohm trace on PCB at the DAC output?
I tried a simple VSIN followed by a TLINE, and configure the TLINE to 100 ohm, 1ns as the following:
If R1 and R2 are 50 ohms, TLINE should also be 50 ohms.
KiCad should not crash, even with wrong input.
Ground connections are missing. Spice always needs at least one GND, here two (on both sides of the TM).
A delay of 1ns is absolutely not visible with a 1kHz input frequency, but will lead to 10m/1n=10Million data points, a little too much.
Please run the attached file with a somewhat corrected circuit and tell me which ngspice version you have.
tm.7z (4.6 KB)
You might raise the frequency (a DAC output is 100MHz, not 1kHz), and use a pulse input. Then a 1ns delay becomes reasonable.
