Please Give Insights and Reasons Into as Why the Voltage Controlled Resistor is "Timestep Too Small"-Error-Prone

[freakcrazy]

Hi, I found that a voltage controlled resistor is very "timestep too small"-error-prone. If I simply connect a voltage controlled resistor with a AC signal source, without anything connecting to the controlling inputs, it will already generate a "timestep too small" error. I picked this scenario out of a whole bunch (say a add a V.C.R. to a parallell RLC circuit) because I found the voltage controlled resistor to have unpredictable, at least to me, behaviors; sometimes gives an error, sometimes not (such in the parallel RLC circuit example, with the controlling inputs connecting to the parallel nodes).

I found this error propensity of the V.C.R when I connected it in parallel with a voltage controlled capacitor (which is controlled by another circuit) connected to an AC signal source. I found that there is no problem whatsoever with the V.C.C; what causes the V.C.R to generate "timestep too small" errors?

Two more small questions: 1) I am trying to plot a frequency Vs current plot (to show the resonance) of a parallel RLC circuit. How do I do that in Multisim? 2) If I try to use the nonlinear dependent source to implement a current, do I connect the outputs + and - so that the N.D.S acts as a branch between two nodes (as oppose to connect the + and - to the two ends of an existing wire/in the same way as implementing a voltage)?

[yyao]

Hi there,

Can you post some circuits that demonstrate these problems? I quickly wired up some VCR circuits and they seem to work fine.

Remember, you must connect the control terminals to a purely positive input. Any 0 or negative input will lead to 0 resistance and convergence errors.

Also, it helps to ground both the control and resistor side of your VCR.

Please attach a schematic if possible. Thanks.

[freakcrazy]

Dear YAO! Thank you for answering my question so quick again! "Remember, you must connect the control terminals to a purely positive input. Any 0 or negative input will lead to 0 resistance and convergence errors. Also, it helps to ground both the control and resistor side of your VCR" I didn't know that!!! I now do what you said and worked perfectly fine!!! So that's the rules for the VCR...

Yao, may you also please answer my two short questions below?

1) I am trying to plot a frequency Vs current plot (to show the resonance) of a parallel RLC circuit. How do I do that in Multisim? 2) If I try to use the nonlinear dependent source to implement a current, do I connect the outputs + and - so that the N.D.S acts as a branch between two nodes (as oppose to connect the + and - to the two ends of an existing wire/in the same way as implementing a voltage)?

Kudos! Thanks!

[yyao]

To get a frequency response plot of any simulation variable, you would want to do an AC Analysis. Click Simulate>>Analyses>>AC Analysis. You can specify the starting and ending frequencies of the sweep as well as the variables for analysis such as voltage, current, power, etc.

Once you see the plot, you can modify the axis to your liking. This is how you generate Bode plots in Multisim

As for current and voltage sources, think about how you would connect a multimeter to a circuit to measure these values. You connect a multimeter:

• in parallel with circuit components to measure voltage drops
• in series to measure branch currents.

So, no, you can't simply wire it up like a voltage source, it wouldn't make sense. Be careful of the polarity. In current source mode, current enters the + node and exits the - node.

Hope that helps.

Message Edited by yyao on 04-15-2010 11:00 AM

[freakcrazy]

YAO! Thanks ALOT man! Connect like a multimeter, ha, much better way of describing than my "connect two nodes... as a branch..."  And thanks for being so detailed ("Be careful of the polarity. In current source mode, current enters the + node and exits the - node."). Got it :manwink:.

I have one more problem; sorry I really need a person to walk me through these, it's my first time using Multisim and I have no idea how to fix issues. The problem right now is that I did what you told me of AC analysis, but immediately it gives me matrix singular error. I attached my circuit here (it's the same one as in the school project thread I posted, except that I got rid of the VCR and replaced with a NDS to fix the timestep too small error). Everything is fine, it shows the correct resonance at 60 Hz (the L and C values are calculated for a resonance at 60 Hz) at the multimeter. Yet, AC analysis don't work. May you please share some pointers with me of what causes the AC analysis to fail?

Message Edited by freakcrazy on 04-15-2010 01:43 PM