Hello,
When using Arbitrary Behavioral Current Source with parallel resistance, but not external resistor, just device parameter set to Rpar=1, Qspice return correct voltage waveform, but current amplitude is super small, while it should be 1A.
This problem does not occur if I use external resistor. Then both current and voltage waveforms have correct amplitude.
Am I doing something wrong here?
Well, are you having another resistive path or gshunt? This example returns 0A for B1 instead of pA. For the B-source, its current reading is what can be output from its terminals. Right now entire current is circulate within B1 and no current exit, this very same nature can be traced back to LTspice.
Isrc.qsch (1.0 KB)
Thank you @KSKelvin for explanation. I thought that when using Rpar parameter Spice will measure ideal source current with connected parallel resistance so the result would be the same as with connected external resistor. But now I know more 
About the small current I measured … I only had another B-source for cos signal, but these were not connected and B2 did not have any Rpar value assigned, so it was ( 
.
However:
A) if I remove B2 circuit, then I also measure 0A of current.
B) if both B1 and B2 have set Rpar=1, then I also measure 0A.
OK, it makes sense now.
You have another current source B2, which doesn’t have Rpar and nothing connected outside. This is a current source but without a complete current loop. Therefore, it cannot solve this circuit and V(BV2) is a floating node. Now, SPICE have to solve its DC solution with Gmin stepping method. From output window, a gshunt is added. gshunt is conductance added from every node to ground. It added not only at V(BV2), but also V(BV), therefore, an external path is formed and that why you see a very small current when B2 is included. Without B2, your B1 with Rpar in it, which give a valid current loop and therefore, it doesn’t requires to add gshunt.
That’s interesting. Thank you very much @KSKelvin for clear explanation.
