Simulation working with 8 NonLinear Dependant Source but not with 9, Timestep too small

[PeterFreedom]

Hi! I'm Trying to simulate the current consumption and voltage drop of amplifiers on a Leaky Feeder communication system in our Mine with Multisim 9. Each Amplifier, normaly feeded by 24 VDC draw 100 mA. If the voltage drops, the current gets higher. At 12 VDC, the current will be 200 mA. The only way I have been able to simulate these amplifiers is to use Nonlinear Dependant sources. it acts like a constant power source. If any of you knows an easier way to do it please tell it to me.

Everything is working fine and I was able to build a big part of the mine communication system. When I try to place one more Nonlinear Dependant Source, Multisim 9 is unable to do the simulation. I have this message:
Output from instrument analysis  
   Unknown: starting Gmin stepping  
   Gmin step failed
   Gmin stepping failed
   Unknown: starting source stepping
   source stepping failed 
   DC operating point failed. Resimulating with UIC.
   TRAN:  Timestep too small; initial timepoint: trouble with node $8 
   doAnalyses: timestep too small 

What can I do to resolve this? I don't want to study the Transient state for now but only the establish state or what I think is called the DC operating point. Thanks all.

[kkwan]

Hi Peter,

I am encountering the same problem here, did you get any feedback from NI engineering support or other experts?

Thanks,

K. Kwan 

[PeterFreedom]

No. They said they would come back to me with an answer but they just forgot me. They said that I could upgrade from Multisim 9 to 10 for 2300$. How can I be sure that it will solve the problem? I finally did it in an Excel Spreadsheet. Yes! I made it in MS Excel using the Iteration or Circular reference. It is not easy to make changes and reconfigure the circuit but at least it is stable. Tell me if you want to have the file.

[OStep]

PeterFree,

This appears to be a classic case of non-convergence. I tried the netlist in other simulators and experienced identical behaviour. The best solution would be to revise the constant power source's model.

[PeterFreedom]

How can I do that? It is a very simple model. Could it be simplier? Thanks.

[kkwan]

Thanks for the response, Peterfree, Max!  I have been using MultiSim 9 to simulate a lot of circuits but this is the first time not going through...I may need to alter the circuit design a little and let you know if there is anything new I can find in this regard.

Thanks,