simulation of simple NOT gates oscillator

[Ridis]

I'm trying to simulate this simple oscillator made with two NOT gates.

The simulation don't start.

The real circuit is very common and works without any problem. I realized it and it oscillate immediately at right frequency.

[lacy]

What I have found is that the problem is centered around U1C and U1D. The way they are connected in parallel is causing the simulator to fault this as a short circuit. Try placing a very small resistor between the output connections of U1C and U1D This will at least get the simulation to run without the error. But there is another issue.

This other issue has to do with CMOS Oscillators not simulating properly in MS. Most of the time when I tried to design one of these things it would not oscillate at the frequency as determined by the resistor and capacitor. Instead it would free-run at an extremely high frequency reguardless of the component values. Once I got yours to running it was running at 12.3 MHz. I am sure that is not the frequency you want. I come to refer to this as "oscillating at the rise and fall times of the gates". In this case the rise and fall times of the Inverter Gates.

I have not had much success in getting CMOS Oscillators to work in MS. I think it has to to with the way the models are contructed for the digital components, but that is just a guess. 

This is really a pain as a majority of oscillators now are constructed using digital gates. I really wish NI would look at this and  come up with something so that we can use digital gates to make oscillators. As it stands right now, you can't and that shouldn't be that way.

I hope some of this helps. If you would happen to somehow get this to oscillate at the correct frequency, then please post what you did to accomplish this as I would be most interested to know.

[Ridis]

Thanks Lacy.

After a lot of try I found the same solution. But this means that you can't use the same schematic for simulation and for pcb ! You loose one of the most interesting features of CDS, the first reason for which I bought it.

I also found how to put the oscillator working in the right way.

As you can see from the schematic included it seems that with any IC > |5 V| the oscillator works well.

About high frequency of “free-running” you said exactly what I think. You can have a confirmation if you look into the model where you'll see

+ ( rise_delay = 50n fall_delay = 30n)

and the sum of two times is exactly the period of “free-running” oscillation.

This means that's almost sure that problem come from model. But in the data sheet you can see the schematic of one gate I include and it's unbelievable NI can't make a good model for this!

By the way a good model could works with any power supply voltage and could allow to use always the same component instead to choose among the two now available (4049BD_5V and 4049BD_10V).

It could be very interesting to know the advice of NI EWB Group.