Post Go back to editing

LTSpice 26 minor bug current source with a short circuit

Thread Summary

The user reported a minor bug in LTSpice 26 where a current source is automatically set to zero current when short-circuited with a wire, unlike a voltage source which can still deliver expected current with a non-zero series resistance. The support engineer confirmed this as a bug and noted it will be fixed. The user suggested adding a tiny series resistance or an error message to handle such situations.
AI Generated Content
Category: Software
Product Number: LTSpice 26
Software Version: LTSpice 26

Sorry!

This could be related to a previous post "LTSpice 26 minor bug when .NOISE points to wrong current source".  It seems that when any non-zero impedance (regardless of how small a value) is placed across a current or voltage source (both with .tran and .ac), the source delivers the expected current.  (The voltage source may add automatically a tiny series resistance to get DC convergence). 

However if the source has a short circuit drawn with a line, the two nodes of the source are recognised as identical, and the voltage/current value gets jammed automatically to zero.

This is logical when parsing a netlist, but not expected for a current source when looking at a schematic.  For the future, maybe LTSpice could recognise this situation with a current source in the netlist and automatically introduce another series node with a tiny series resistance.?  Or clearer, give an error message that shorted sources are not allowed?

Many thanks

Possible Bug Current source short circuited.asc

  • Hi  ,

    For the future, maybe LTSpice could recognise this situation with a current source in the netlist and automatically introduce another series node with a tiny series resistance.?  Or clearer, give an error message that shorted sources are not allowed?

    This might be a case of is it helpful? The schematic is merely a way to produce a netlist. By shorting a source, the user is eliminating the component. I do know this is a way of commenting out components. That said, would it be useful to have note of this in the log?

    mike

  • Thank you for your patience Mike.  This is a very minor detail and I am OK if you drop it without commentary.

    Best regards

    These two details were confusing me:

    1. In circuit theory it is OK to leave a voltage source open circuit since the current doesn't go to infinity, and a current source short circuited so that its voltage doesn't go to infinity.

    2. In LTSpice the behaviour with the short circuit using a wire connection is a little different with voltage and current sources.:

    voltage source with non-zero internal series resistance still manages to give the expected current; with zero internal resistance the voltage gets jammed to zero, with no comment.

    current source gets the current jammed to zero and comments this as a possible error in the netlist (which may not be read so often).

     

  • Hi  ,

    Thank you for your patience Mike.  This is a very minor detail and I am OK if you drop it without commentary.

    Turns out this is a bit of a bug; and it appears to be an old one. You are correct, these constructs should be supported, and not removed from the circuit or assigned zero current.

    Thank you for catching this and bringing it to our attention. It is admittedly a low priority fix, but will be fixed.

    mike