Different result in LTspice and Qspice when simulating the same circuit.
Untitled0.qsch (657 Bytes)
Missing LTspice file and two libraries for correct checking
.lib C:\Users\bordodynov\AppData\Local\LTspice\lib\cmp\standard.mos
.lib C:\Users\bordodynov\Documents\LTspice\user.mos
Rename attached .txt to .cir to run in both Qspice and LTspice. This netlist is basically from Untitled0.qsch but delete library link. No different if running the same netlist.
Untitled0.txt (342 Bytes)
1 Like
I made a netlist file from the LTspice file, but with this file there was no oscillation in Qspice. I figured out what was going on (bypassing the inductor with a resistor) and added the Rpar=0 parameter. I didn’t think that the default inductor bypass in LTspice could affect the simulation result so much. You surprised me. I didn’t know that Rpar=0 works the same in LTspice and Qspice. Although I knew about the default inductor bypass in LTspice, I didn’t pay attention to it. I guess this will be news for other LTspice users too.
LTspice has options for Rpar and Rser. If I disable “Supply a min. inductor damping if no Rpar is given” and remove Rpar=0, it also does the trick. I have noticed that your circuit is usually sensitive to Rpar, which is why this is the first area I typically check when I see your post.
- LTspice in default : Rser = 1m, Rpar (I don’t know the formula, it enable depends on option)
- Qspice in default : Rser = 0, Rpar = Inductance / 15.91 / Gmin
Currently, LTspice and Qspice have quite different default parasitic settings for inductor. One of the checking items, try declaring these two values for inductors if the results do not match between them.
1 Like
Hi，Kelvin
There’s difference between LTspice and QSpice when using B voltage source, please help check. Thanks a lot!
Netlist:
V1 in 0 SINE(0 1 10k)
B1 out 0 V=idt(v(in)-delay(v(in),Ts))/Ts
.param Ts=1/10Meg
.tran 1m
.end
Hi, Bordodynov
That works if define maxstep, thanks.