Can any one solve the Multisim DigitalCounter sample is now broken puzzle?

[Euler's_Identity]

I just tested one of the samples attached to Multisim titled "DigitalCounter" and it worked, but then I

added my oscillator circuit in the upper right hand corner, and

the Multisim DigitalCounter sample stopped working, while

my oscillator pumped out a very wrong somewhere around 4+MHz frequency.

I then deleted my oscillator, restoring Multisim's circuit.

The Multisim sample circuit worked again.

Of course, I then did an undo, and my oscillator circuit was restored along side Multisim's circuit.

The Multisim sample circuit was broken again.

What is it about my oscillator that can kill working Multisim circuits, including, now, Multisim's own circuits, which have to be without errors, right?

 

[Euler's_Identity]

I tried one more thing, and it's still broken.

[lacy]

It didin't break it. With you Oscillaotr in circuit it is dragging the simulation speed way down and causing it to appear like it isn't doing anything. Wait until the counter/timer in the lower right hand corner reaches 1ms and you will see it change.

I see the oscillator is putting out a frequency now, but there is still a problem and I have run up against this also. The circuit is not oscillating at a frequency determined by the R/C network. It is free running at the rise and fall time of the gates. Your feedback loop from U1B to U1A is setting this up. Basically it is ignoring the R and C in the circuit. You could remove these components and get the same results. Also, you have the Digtial settings on IDEAL instead of REAL and this is why you are probably seeing this oscillation now even though it isn't correct.

Like I said CMOS Oscillators do not work no matter how hard you may want them to.

[Euler's_Identity]

That's one of the things I noted: the time.  For it was possible that adding the oscillator to the circuit could have slowed down the simulation and that I just wasn't allowing it enough time.

Run the sample circuit, and within 22ms the counter has started.  I consider this a checkpoint, a constant of the circuit itself, to determine if the circuit is working properly.  Add the oscillator circuit, and the counter never starts.   Now, thinking about the circuits themselves, no matter how long it takes for the simulation to play out, nothing should have changed with respect to that 22ms checkpoint, that constant.  This is why I concluded that the original circuit stops working when the oscillator circuit is added to the schematic.



 

[Euler's_Identity]

Scratch that last.  I may have missed a decimal.  I'm running it again now, and I'm thinking I mistakenly saw 22ms for 22us.  waiting

[Euler's_Identity]

Ok, yes, I did make a mistake.  The circuit still works, even with the oscillator circuit in the schematic.   Right now the circuit is running, and, on my end, I'm seeing the counter start at very close, if not dead on, 1ms.

Understand that I need some sort of a constant to solve this problem.  I'm thinking that this two-fold constant is the best way.  First, we start with a known to be correct circuit, the Multisim sample circuit, and we also determine a drop dead time.   Perhaps I'll make that the 1ms counter start time as opposed to the 22ms counter stop time.  (I was in error there also.  The counter STOPS at 00, if you don't touch anything, at 22ms, not starts.  It was late last night...)

waiting

[Euler's_Identity]

I'm letting the whole simulation play out (get to max count), so this is going to take a while.  Adding that oscillator dragged the simulation down to a crawl, but I've an idea in mind that may help that when I finally get there.  One piece at a time...

waiting

[Euler's_Identity]

Incidentally, to see where I'm coming from, I think that this is great that this circuit is still "running" (we aren't at the stop yet, but I'm not expecting a Multisim fail at this point) with the same characteristics, even though the oscillator is in the circuit.  This is light at the end of the tunnel.   This is showing that the oscillator is creating too many points that must be solved; the oscillator really is the cause, not Multisim's response to CMOS.   (Lacy, yeah I know you're laughing at me over all this, but I need to prove this to myself, prove that my CMOS designs do work even when I drop in this oscillator hurdle.)

still waiting, and just crossed 10ms...
 

[Euler's_Identity]

The circuit ran until completion, where I stopped it at 22ms.

Now, I'll switch to "real," use the 1ms counter start time, and see if I can sneak up on the oscillator.

[Euler's_Identity]

Oops!  Forgot another variable.  Uncharged capacitors start as shorts, so I'm also switching to "set to zero."