Hi, First!!! Why U need to plot V(OUT) -I(V1) , they are in diffrent unti, maybe useless, am I right?

In the simple example schematic, I wanted to plot the current in L1 but I did not want to add a current probe just to do this. So I plot the current in the power supply, V1. But if you plot `I(V1)`, the current is the wrong polarity (upside down), so I want to plot `-I(V1)`. I *also* want to plot `V(OUT)` so I have two traces: one for `V(OUT)` and one for `-I(V1)`. The problem is that if I write the probe statement like this: `probe V(out) -I(V1)` It does *not* plot two traces: one for `V(OUT)` and one for `-i(V1)`. What it plots is one trace of: `V(OUT)-I(V1)` which is not what I want. To answer your question: "Why U need to plot V(OUT) -I(V1) , they are in diffrent unti, maybe useless, am I right?" Other way round: I do *not* want to plot `V(OUT) -I(V1)` What I want to plot is two traces: one for `V(OUT)` and one for `-i(V1)`. But Waveform (or is it ngspice?) will not let me! I can work around the problem is different ways: `let -I(V1)=IL1` `probe V(OUT) I(IL1)` or `probe V(OUT) (-I(V1))` or `probe -I(V1) V(OUT)` all give me the two traces I want but it is hard for the newbie to learn all these things. Why should people have to know that: `probe -I(V1) V(OUT)` is OK but `probe V(OUT) -I(V1)` is not? (It also adds a lot of words to trying to explain how to use `probe` ... ) :)

Ahhhhh. OK. Maybe I misunderstand how the syntax of the probe command works. I think you can ignore this as a bug report: I need to describe this carefully in a tutorial somewhere .... If you are interested, below, I have gone through step-by-step what I think happens. Suppose I have a circuit with V(a), V(b), V(c). If I want to probe V(a) and V(b) then I can do: `probe V(a) V(b)` That will show me two traces; `V(a)` and `V(b)`. That's correct. If I want to probe the difference between V(c) and V(b) I could do: `probe V(c) -V(b)` or `probe (V(c) -V(b))` or `probe V(c,b)` All 3 options would show me V(c,b) (i.e. V(c)-V(b)). What If I want to show V(a) and V(c,b)? If I write: `probe V(a) -V(b) V(c)` then I will get the wrong traces showing V(a,b) and V(c) but, if I write it like this: `probe V(a) V(c) -V(b)` or: `probe V(a) (V(c) -V(b))` or: `probe V(a) V(c,b)` then I will get the correct traces showing V(a) and V(c,b). Now what if I want to show 3 traces: V(a), -V(b) and V(c)? If I write: `probe V(a) -V(b) V(c)` then I will get the wrong traces showing V(a,b) and V(c) but if I write: `probe V(a) (-V(b)) V(c)` then I will get the correct traces showing V(a), -V(b) and V(c). So what this means is that to be sure that a minus sign is always treated correctly in a probe command, it must always be inside a pair of brackets. Even if it is just being used to negate a trace and not as part of an expression to create a difference. So, in my original post, the correct way to show two traces, one for V(OUT) and one for -I(V1) is to write the probe command like this: `probe V(OUT) (-I(V1))` The arithmetic of the trace display comes out correct: it shows two traces representing V(OUT) and -I(V1). The only problem then is that the -I(V1) trace is actually labelled as: `I((-I(V1)))` instead of simply `-I(V1)`. but that is already the subject of my Feature Request: