Switcher Cad III

[Russlk]

I am questioning the accuracy of Switcher Cad. I tried simulating the LTC1872B but the waveforms were sine waves, not the switching waveform of the real device. The DC output of a standard circuit seems reasonable, but if I try to use it in a non-standard circuit, the results are not what I expect.

I tried simulating a different circuit using transistors. I had a current sense resistor which should turn on an NPN at 0.6 volts. I reduced the resistor value to get more current but the current stays the same, the transistor turns on at lower voltage. When it got to the point the transistor was turning on at 50 millivolts, I quit and tried something else.

I simulated the same circuit in ISPICE (an old DOS program) with good results, the transistor always turns on at 0.6 volts.

I think Switcher Cad is faking it in order to avoid "time step too small in transient analysis", an error message that I get a lot using Berkley Spice.

 

[Optikon]

I am questioning the accuracy of Switcher Cad. I tried simulating the LTC1872B but the waveforms were sine waves, not the switching waveform of the real device. The DC output of a standard circuit seems reasonable, but if I try to use it in a non-standard circuit, the results are not what I expect.

I tried simulating a different circuit using transistors. I had a current sense resistor which should turn on an NPN at 0.6 volts. I reduced the resistor value to get more current but the current stays the same, the transistor turns on at lower voltage. When it got to the point the transistor was turning on at 50 millivolts, I quit and tried something else.

I simulated the same circuit in ISPICE (an old DOS program) with good results, the transistor always turns on at 0.6 volts.

I think Switcher Cad is faking it in order to avoid "time step too small in transient analysis", an error message that I get a lot using Berkley Spice.

Can you attach the .cir file? for us to try?

Also, try using a real NPN part number with a real spice model behind it.. i.e. 2n2222 etc... Maybe just something wrong with the model. I seriously doubt there is though..

I'm confused about your last statement.. what does time step too small error have anything to do with turn on thresholds..?? I don't think they're related or SWCAD is faking something to avoid a legitimate error..

 

[plot]

Welcome to the world of simulation... it's not always accurate... at all... :wink:

although, i do love switchercad III... post your circuit and special parts used in it? maybe something wasn't done right.

You could also try starting a new circuit in switchercad and building it from scratch... then resimulate it... see if that works.