Troubleshooting Multisim Circuit Simulation with Astable Multivibrator

[Crazymechanic]

Hello so I'm struggling to understand why is multisim telling me different outcomes in the exact same circuit with same values just in two different simulations.
I have attached the photos for you to see the circuit of astable multivibrator.
In one of the multisim simulations the values work perfectly I tried a new one with the same circuit and it's not working properly or not at all not even with different values selected in the margins of what is possible for such a circuit.

By the way I forgot to tell that the transistors in both the simulation and my drawing are the "mpsa44".

Thanks.

 

[the_emi_guy]

It can be a challenge to get Spice to do what you want in circuits which have feedback. Regenerative feedback is often quite difficult for Spice. You realize that there are more than one valid result of this simulation depending on initial conditions. There is a metastable state where the circuit is in equilibrium with no oscillation. How are you telling the simulation that you do not want that solution?

One way of coaxing Spice to behave in these types of circuits is to make good use if the nodeset (.NODESET) and set initial condition (.IC) cards.

By the way, a better description of "not working" would help us troubleshoot with you.

 

[Crazymechanic]

Well yes I understand what you mean but basically spice just like any computer program works like the binary code and if i put in some certain value then some certain outcome must be there and changing these numbers also the outcomes should change.
Well basically I need a square wave from the multivibrator to switch some mosfets , ofcourse there will be other circuitry but this is the main one.
The thing is that I made the generator and it worked all ok.Then I did some analysis and all was ok , then I thought hey I should try it on a copy file just to see if all is ok and then well you see it nothing is ok.:D
Also I changed the base resistor values from the npn totally open to totally closed that would be some 100k in range or more and it just gave me a weird triangle like distortion with a varying frequency or something like that or it just gave me a straight line dc.
But before in that same range I adjusted my needed resistance with potentiometers and set in a resistor after that and all worked.
So basically which scenario should I believe now ?
The one that worked or the one that failed at many different input settings?

 

[the_emi_guy]

Is this multivibrator a circuit that you need to simulate because it will be built, or are you just using it to create a squarewave so that downstream circuitry can be simulated?

 

[Crazymechanic]

Well the downstream circuits are pretty clear to me and the way they work is not needed to be simulated basically the fundamental thing is always the most important one.
yes the mltivibrator will be a real life circuit and that's why I'm so depressed upon it as I need it to work as nicely as possible as that will affect all further efficiency and performance of the switching circuit.

I guess I'll just have to stick with basic good old math and ohm's law and transistor datasheets.
But anyhelp would be appreciated as I believe multisim being a quite nice simulation software.

 

[AlephZero]

You could try making your circuit not perfectly "symmetrical". For example make the two collector or base resistors or the two capacitors slightly different values (say 1% or even 0.1% different).

The reason why this type of circuit starts to oscillate in real life is because the pairs of component values are not exactly the same. If the simulation is too "perfect", it has probably got into a stable state with both transistors on, or both off.

 

[Crazymechanic]

Thanks Aleph for your suggestion although I already did that multiple times usually with the base resistors I made them even 1k different and still it did nothing to help the simulation.
In real life I wouldn't even have to make that difference as the consumer parts and resistors have a pretty high tolerance and hence the circuit should start to oscillate by any means.

Is there some more ideas why is this happening?
I wonder why because the circuit is so simple and basic I mean what can go wrong with the simulation anyway?


In the beginning I thought that I have chosen transistors that don't have the right characteristics so I made my choice to the mpsa44 ones and I think they are pretty ok for the job to be done.
Ok I'll be happy for any insight.
Thanks

 

[the_emi_guy]

I'm assuming you have DC analysis turned off?

 

[Crazymechanic]

Well I guess yes they are switched off but I don't know correctly as multisim is kinda complicated haven't found out all the features yet.I'm doing more in my head and on the paper multisim is just for easing the process.

Well I uploaded the file with my multisim simulation on a server maybe some of you if yopu have some time can take a look at it ?

the link to the file :
http://failiem.lv/u/lrxfbbd

 

[Crazymechanic]

maybe somebody can look at this multisim simulation and try to make the same from scratch and I believe his will not work and I don't know why.
I'm sorry to bump a thread I just forgot to say this in last post yesterday before sleep. :)

 

[Crazymechanic]

Hey and here is the simulation in which the circuit doesn't work as planned although the tolerances and values are the same.
Can anybody please take a moment to check it out?

http://failiem.lv/u/qdmbvcb

Thanks.

 

[Crazymechanic]

Anybody?

 

[vk6kro]

I don't have Multisim, so I tried this with LTSpice.

It oscillates if I change the power supply to 2 volts and the capacitors to 50 nF.

There is also a delay before it gives any output, so you could miss this if you put it on a sweep of a few microseconds.

Start off with a horizontal scale of about 1 second so you can see if it oscillates at all.

To get higher voltage operation, you can put resistors from base to ground for each transistor. I found 560 ohms worked for a 300 volt supply, but this will vary with the exact transistor you are using.

The output really slows down for oscillating circuits like this, so you may have to wait a while to get a display. If it goes part way across the screen and then stops, you should wait to see what it is going to do.

If it doesn't do anything, measure the voltage on one of the bases and adjust the base resistor or the supply voltage so that the base voltage is about 0.7 volts.

The emitter resistors didn't seem to do anything so you could omit them.

 

[Crazymechanic]

So I have my test circuit up and ready and you can see my circuit in the attached image.
The problem is I have gone so far as to make it oscillate and it does so very good, but after a while it locks up and the oscillation stops and the transistors stay opened or botch closed.

Well I tested the circuit in multisim as I said earlier but the values in life needed to be a bit different.
The resistors on collectors are 3.3k 5 watts each because they have like 100 mA going through.
Maybe this happens because the resistors heat up unequally and their resistance changes and so it disrupts the oscillation? Because in the first some 15 seconds everything runs smoothly and after that the frequency suddenly increases the amplitude of the frequency shrinks and then it hits zero and the oscillation stops.
Maybe someone has some insight on as to why it happens like this.

 

[vk6kro]

When one of the transistors is turned on, the 3.3 K resistor connected to its collector is dissipating about 32 watts. If it is turned on for half the time, this is still 16 watts.

So, that is a significant overload for the 5 watt resistors. They must be getting very hot.

I would also try much larger capacitors than the 50 pF ones you have now. Make sure the capacitors can cope with the voltage.

There is an effect called Miller effect which makes the input of each transistor appear to have a much larger shunt capacitance than you might expect. If this is enough to make the 50 pF capacitors unable to drive the transistors, then the oscillation will stop.

In a simulation, I couldn't get any oscillation with 50 pF capacitors. 330 pF capacitors gave oscillation at a frequency of about 13 KHz with your resistors, but this was with 2N3055 transistors.
So, yours might be only just managing to oscillate and then failing when they get hot.

 

[Crazymechanic]

Hi thanks for the answer , Ia m using the Soviet made 5watt resistors , they are not of a oxide layer but of a wire wound type so the heat is no problem, even checking them with hand seemed quite ok.
So basically your saying that the capacitors are the ones that fail?
What would be the chances for me to correct something yes ofcourse I could put a higher capacitance ones in like the 330pF which you mentioned but that would lower the frequency significantly and that is not desired.
Also lowering the collector resistor value and then putting higher capacitance caps in would help in saving the frequency but the current through tha CE of transistor would increase and I can have only maximum 300mA on these mpsa44 ones.

 

[Crazymechanic]

Almoust forgot to mention I have 500k potentiometers to each base of the transistors so that I can adjust how much they open , Maybe changing decreasing the base voltage would help or should I change the capacitors?

 

[vk6kro]

What frequency do you need?

Do you have access to a 5 volt power supply?

 

[Crazymechanic]

Well I need about 100khz, I just changed the capacitors to different 100pF ones.Now the oscillation doesn't die off , well at least not in the time when I did the measurements.
The only problem with the 100pF ones is that I have heavy harmonics now but I guess i'll adjust the base voltage because the capacitors are different.

Yes I have some spare computer psu's and a battery so I can have 5v.Why are you asking?
Ofcourse it would have been smarter to make a multivibrator operating from 12 or 24 or whatever little volatge and then add the amplifying stages to get the oscillation up to 330volts but i guess now I will do fine with this setup I only need to correct some errors and adjustments.

 

[Crazymechanic]

Well I went through different types of ceramic capacitors in the pF range and found ones that work pretty well they are 33pF.
Now can someone tell me I am getting a sine wave out of my vibrator but it should have been a square wave is it right or there is something wrong with my oscilloscope?