simulation and complete lack of real time

[pintail120]

Hi,

I have installed multisim and put down a a very basic 555 timer where  I noticed everything is in slow motion.

So I hooked up a signal gen to a scope at 50 hz, the scope then draws the signal wave at less than 1hz?

My processor is almost on idle so I cannot forsee any lack of processing power.

What completly blows me away is that NO where in the manual is this made obvious, how can simulation of electronics be done at 1/10th of the real speed?

What even confuses me even more is that if I up the freq of the signal gen, I get  proper high speed results plotted on the scope, so this means it has nothing to do with CPU usuage.

When I plot 50hz I get this rediculous slow motion less than 1hz on the output of the scope.

please help explain this?

cheers

Gavin

[lacy]

I usually have the opposite situation. My low frequency signals are usually fairly fast and my high frequency signals simulate extremely slow.

Multism using the transient simulation (pressing the run button)  does not run in "real time". The time indicator at the very bottom of the screen indicates what I would refer to as circuit time. It is the exact time in the real world that it would take for a circuit to accomplish what it is doing even though it may take 1 minute of our clock time to simulate a 1 second circut event. The reason behind this is that it take time for all the spice commands and circuit mathmatics to be executed, interpreted, and then a solution arrived at and then all the outputs plotted on the instruments. It is this number crunching "on the fly" that makes the simulator run out of sync with real time.

One thing that can be used to speed things up is to look under SIMULATE>INTERACTIVE SIMULATOR SETTINGS then go down to where it says Set Maimum Timesteps (TMAX). Click the circle next to Maimum Time Step (TMAX) and in the the little box to the right change it from 1e-005 to 1e003.

With the obve setting you can approach real time with simple circuits, but with more complex circuits or high frequency circuits this could cause problems as the simulator doesn't have time to converge to a solution and you will get "time step too small" errors or the output will not look as you expected it to look.. This setting is basically the resolution setting of the simulator. The lower this setting is the less accuate the results but the simulation speed is faster. The higher this setting is the more accuare the results but the slower the simulation speed.

I hope I have helped somewhat and that this helps explain what you are seeing. 

Message Edited by lacy on 10-28-2007 08:30 AM

[pintail120]

thanks for that, I knew there must have been a simple explanation like this,

Cannot understand why they dont have the default setting set to this and some where deeper in the documentation give the more accurate simulation settings.

many thanks again for your most helpful advice

cheers

Gavin

[pintail120]

Hi

I get these errors, are they serious when I do the rules check?

Can someone explain them

Multisim  -  2007-10-28 20:50:23

ERC [monostable Gavin]  <2007-10-28 21:09:11>
   Error: Connecting 'Bidirectional to Power';   [V1 pin 1, monostable Gavin]  to  [U1 pin 8, monostable Gavin]
   Error: Connecting 'Power to Bidirectional';   [U1 pin 8, monostable Gavin]  to  [D1 pin K, monostable Gavin]
ERC [monostable Gavin] completed;  2 error(s), 0 warning(s);  Time: 0:00.03

[lacy]

I will have to read up on the Electical Rules Check function in Multisim. I rarely use it cause I have found that it reports errors that are not really errors. For example your circut. The only reason it is reporting an error is that the pins on the power source and diode are labeled a Bi-directional and you are hooking it up to another pin that is labeled Pwr (on the 555 Timer) which would be considered non-bidrectional. Why it is like this on a DC power source and diode is something I don't know. The pins should be labeled differently and this could be a bug/issue that you may have come across. Any bi-directional pin  connected to a non bi-directional pin will cause this. Sometimes it is an error and sometimes it isn't 

Nestor needs to read this about the these type components retruning ERC Errors and maybe address this in case it is a known issue or he may have additional information that I don't kniow about.. The one thing I will say that if you simulate your circuit and don't receive any additional error messages then I would be comfortable in saying that everything would be o.k and you shouldn't worry about these ERC errors.

I hope this helps some and that I was able to explain this in a way that is understandable.

Message Edited by lacy on 10-28-2007 05:08 PM

[lacy]

There is a way around this ERC Error issue. I read up on it and you can change it by going to TOOLS>ELECTRICAL RULES CHECK>ERC RULES. From there go down to the second row that says "Pwr". Then go to the right under the column labeled "Bi" and click on the square until the color changes from red to green. This should eliminate this from returning an ERC Error because in my opinion this shouldn't be considered a violation of any design rules based upon my observation of the way the pins are labeled on certain components.

I hope this helps as it worked on mine.

[nestor-sg]

Hi Lacy,

ERC violations will NOT cause your circuit not to simulate or not to transfer to PCB Layout... ERC is for your own guidance when designing a circuit.

The most important thing to know is that ERC is set by the user using the ERC Rules Matrix that you just mention on your message. No matter what we set as default in that matrix, users would eventually define their own rules. As a designer after you start placing components you would run ERC from time to time to help you identify possible bad connections according to your rules.

I hope this information helps...

[lacy]

Thanks for the explaination Nestor. You stated it better than I did and that is really appreciated.