Spice does not find GND Vector. (Error: vector V(GND) not found!)

[fibergm]

Hi,
first post.
So i wanted to simulate a simple OPAmp circuit.

The simulation works, like i expect it to work, but it wont plot the curve for the ground signal.

I also tried with a different circuit just 1 resistor connected to GND but i wont find the vector.
So i wanted to ask what i did wrong and how i fix it.

BTW the netlist.
.title KiCad schematic
.include “C:/Users/finnr/Downloads/sloj067/TL072.301”
.save all
.probe alli
.probe p(R1)
.probe p(V1)
.probe p(V2)
.probe p(R2)
.probe p(V3)
.probe p(XU1)
.tran 100m 1 0
R1 Net-U1– Net-R1-Pad2 200k
V1 +15V GND DC 15
V2 -15V GND DC -15
R2 unconnected-R2-Pad1 Net-U1– 100k
V3 unconnected-V3-Pad1 GND DC 1
XU1 GND Net-U1– +15V -15V Net-R1-Pad2 TL072
.end

[JohnSG]

In Spice, GND is always zero, by definition. In fact, originally, you would use “0” to designate any ground node. This is built into the solver, and is why every node in the circuit needs a dc-conducting path (resistor, inductor, voltage source) to GND for the solver to function. Hence, no GND vector is ever generated.

I can’t think of a case where I ever needed a GND signal measurement from Spice and I have been using one flavor or another for nearly 40 years. You might ask yourself why you need it.

However, if you really need it for some reason, you might try connecting a voltage source set to 0 V to GND and measuring the other end. If that doesn’t work, add a resistor from the free terminal of the voltage source to GND and see if you can measure that.

Hope this helps.

John

[fibergm]

Hi John,

I was using the spice module to show something in one of my presentations.

In the presentation I wanted to show the virtual ground, between the positive and negative of the OPAMP.

For that I wanted to make a measurement between those two pins, like i would have done in LTspice (By clicking and dragging between the signals).

You see I’m kind of new to the KICAD spice and to the spice in general.

I will try to see if the 0 Volt source works.

Thank you very much for your advice.

(God I feel like talking about the book Dune)

[JohnSG]

Spice by itself is old and text-driven. It’s worth doing a couple simple circuits this way, because every different version of Spice, at it’s core, uses a text-based netlist that is sent to the simulator, which spits out a data file. The schematic entry and graphical output are extremely useful conveniences, but what they do is convert schematics to netlists, and datafiles to graphs. If you understand how this part works, it will be very useful if you want to get good at using Spice. A good and reasonably short reference for the core part of Spice is https://engineering.purdue.edu/~ee255/design_hub_files/References/SpiceManual.pdf. Note that various flavors of Spice will have some minor variations, but they all start from the given reference.

In your case, the op-amp input pin is identically equal to zero because it is tied to GND. So, you only need to look at the inverting input and show that it is close to zero. Any deviation from zero will be due to the finite gain of the opamp and any input offset (if you have an opamp model with non-zero offset). EDIT: finite numerical precision can also contribute to simulated error.

John