Multisim has trouble simulating simple triac lamp dimmer

[funberry]

triac lamp dimmer

I've just installed Elecronics Workbench Multisim 9, and was tinkering with a classic triac/diac lamp dimmer circuit.

Waveforms are extremely spikey, inconsistent, and at times just plain wrong. Some signal excursions are even higher than the AC input voltage.
If I breadboard the circuit, and put a scope to it, I get completely different (and correct) waveforms. Nice and clean half-lemon slices.

For some component values, Multisim 9 even stalls and cannot simulate (component values which work well on the breadboard).
If simulation is supposed to replace breadboarding, how can someone trust simulation data under these conditions?

Is it just me, or does the softare have known limitations with switching circuits?

Thanks
Adrian

 

[ali007]

dimmer multisim

yes, i have noticed that problem as well. I think it might have to do with the spice model of their components. It's been known that Multisim supports xspice models rather than Pspice models. There could potentially be a parsing problem with their simulation engine.

 

[EDA_fan]

multisim dimmer circuit

Hello there.

I don't have much experience simulating the triac/diacs in Multisim. Did you know since National Instruments bought Eelectronics Workbench you can post your questions to NI's online forums? NI has dedicated engineers that will periodically troll these forums and answer questions posted there.

h t t p ://forums.ni.com/ni/board?board.id=370

The only caveat is that you have to complete a [free] registration. They are pretty good about not emailing you to hassle you though.

Happy simulating!

-EDA_fan

I've just installed Elecronics Workbench Multisim 9, and was tinkering with a classic triac/diac lamp dimmer circuit.

Waveforms are extremely spikey, inconsistent, and at times just plain wrong. Some signal excursions are even higher than the AC input voltage.
If I breadboard the circuit, and put a scope to it, I get completely different (and correct) waveforms. Nice and clean half-lemon slices.

For some component values, Multisim 9 even stalls and cannot simulate (component values which work well on the breadboard).
If simulation is supposed to replace breadboarding, how can someone trust simulation data under these conditions?

Is it just me, or does the softare have known limitations with switching circuits?

Thanks
Adrian