LTSpice anomaly. Same wire, different voltages.

[Mooly]

The two pictures show part of the same circuit. When I run the ".op" command to simulate DC conditions I find the opamp output line can show different voltages depending where on the line is clicked. Hovering the cursor shows the correct voltage (around 7 volts). If I then click below the opamp to tag that line with the voltage it is correct. If I do the same again but tag the line above the opamp then an incorrect voltage is placed.

Is this a glitch or am I doing something fundamentally wrong ?

[Rundmaus]

Actually, this is not supposed to happen!

Spice does *not* know about different 'places' along a wire - all circuit points connected by wires/lines belong to the same so called 'node', they are regarded as one point without dimensions. All calculations are made for these nodes (listed in the netlist file), and everything belonging to a node shares the same properties. Maybe it is a problem with the graphical user interface.

If you're inclined, check the raw data output for the values assigned to that node.

Greetings,
Andreas

[jan.didden]

It seems that in the right pic, you're reading a different wire. Look at the small tag handle below the '6'.

jan

[Rundmaus]

Thats no tag handle, but the 'minus' sign in that crappy font!

EDIT: And the other wire is connected to the 'ground' node, which is always zero potential in Spice - so the reading is definitely *not* from the horizontally crossing wire...

[jan.didden]

Can we see the whole circuit?

[EssB]

I've had this sort of thing happen when moving components, they appear to drop in place but don't always connect. Maybe the opamp output isn't really connected? I end up just deleting and remaking the wire connections.

..err that's not right, but there must be a break in that wire that just isn't visible --> delete and redo

[dchisholm]

Originally Posted by EssB

I've had this sort of thing happen when moving components, they appear to drop in place but don't always connect. . . .

I haven't noticed that problem with LTSpice, but it seems to be (too) common in Multisim.

Whenever a program - not just circuit simulators, but ANY program - has a "mouse over to see <some kind of information> " feature, it's not always easy to determine WHICH object is being referenced. Sometimes it seems to be the object at the point of the cursor, other times it's an object under the cursor, and occasionally it seems to be some random object in the general vicinity of the cursor.

With LTSpice, as you move the cursor around you can get the message in the lower ribbon bar, e.g., "DC operating point V(pre_out) = 12.34 mV". At least that tells you which node LTSpice thinks you're interested in, though the default node assignments ( "V(n0008)", for example) are pretty cryptic. That encourages me to select the handful of nets that I'm especially interested in and give them explicit names.

Dale

[jcx]

you can grab, drag/move lines/parts slightly to see where/if the connections are made with the mouse button held, then drop back in place or use undo if your movement messed up the connection

you can also highlight a net

[bill_a]

Mooly,
If you really think there's a problem with LTSpice, report a bug (click on the Help, About ltSpice). If you have a real point, and aren't just blowing smoke, then Mike Englehardt will be very happy to respond - I've written him several times and gotten very cordial responses.
Bill

[Mooly]

Originally Posted by Rundmaus

..........If you're inclined, check the raw data output for the values assigned to that node.
Greetings,
Andreas

Thanks... that seems to show a problem. The node is V (no32). When I tag the node above the opamp and then hover over it the node has changed to V (no23)

The only difference bewtween "Test1" and "Test2" is the tagging of the line. The curcsor was hovering over the line each time the screen grab was taken.

Originally Posted by jan.didden

Can we see the whole circuit?

Here's the exact file that's causing the problem.