SIDE 1299, W11. INTERNAL 4060 OSCILLATOR DRIVERS

I try to use the internal drivers of the 4060 to obtain an oscillator as the 4060 manual says, but I always obtain a constant frequency even if the capacitor value changes, for example.

In the physical circuit it works fine. Tell me please.

Same. Works only with an external generator.

This post is not addessing to you, please refrain from answering. I only care about the response of the programmers.

I will not interfere. It is clear that the question to the developer. But I just wanted to show that the pulse generators on the three inverters also do not work. Such schemes work in version 0.4.15----

In Simulide using discrete inverters and the same structure as 4060, they oscillate, but always at 50Mhz...

I have used the 4060 phisically many times and it oscillates perfectly over a wide range of frequencies according to its datasheet.

I remember, W11 and r1299.

Arcachofo, tell me, please.

This issue has no simple solution, specially for the general case.

I think it can be "forced" to work tweaking propagation delays and input voltages.
By now I could not find a combination that can work in a Logic Subcircuit, but it should be possible.

Finally I implemented a solution for the general case at Rev 1301:
There is a new property to set the initial state of Gates.

This can solve unwanted oscillations in some circuits like this, some other ring oscillators, Flipflops implemented with Gates, etc.

74HC4060 required some extra adjustments of input voltages to get RC oscillator working.

Test 74HC4060

In addition I'm adding an "Initial Voltage" property to Capacitors.
This will help with Transistor based oscillators, but can also be used with these Logic Gates oscillators.

Probably will add "Initial Current" to inductors as well.

@Alex68:

4060 seems to work in 0.4.15, but indeed is not working correctly.

And in any case 0.4.15 is a completely different simulator, so 1.0.x requires different solutions.

- 0.4.15 does not simulate Gate propagation delays.

- 0.4.15 has an algorithm to suppress oscillations, but it has its own problems:
In the case of 4060 it will oscillate a few times and after that it will start to work properly, but those oscillations will increase the flipflops, which is not correct.
In addition it can suppress oscillations that are correct.

For example none of this is correct:




With the last changes everything is working correctly in 1.0.1.
For example the same circuits with 10 ns propagation delay:

Good job, Arcachofo, thanks, but... is it worth spending time on such an obsolete version (0415)?

I put a test on how this counter works in hardware. Now in the latest version the inverters are running at their delay frequency. 50 MHz.

Yes, it makes sense, thanks. But it would be much better if you attach the Simulide schemes, all in english (The only universal language) to facilitate the work of the programmers without them having to draw it and translated. This is much appreciate. Please keep it in mind in all your posts (Otherwise they are only useful for Russia and its countries of influence).

Defran wrote:Good job, Arcachofo, thanks, but... is it worth spending time on such an obsolete version (0415)?

These fixes has nothing to do with 0.4.15.
What Alex means is that 4060 seems to work in version 0.4.15.
He probably thinks that something got broken and it is only a matter of returning things to 0.4.15.
I just tried to explain that 1.0.x works in a different way and that the solution used in 0.4.15 was a bad one.

Alex68 wrote:I put a test on how this counter works in hardware. Now in the latest version the inverters are running at their delay frequency. 50 MHz.

arcachofo wrote:With the last changes everything is working correctly in 1.0.1.

You don't have the last changes yet.

To clarify the solution to this problem:

Version 0.4.15 used an "automatic" solution: it tried to suppress "suspicious" oscillations.
But this is a bad solution:
- It didn't work in every case.
- It worked when it shouldn't.
- It didn't suppress the first few oscillations.


The solution I just implemented for 1.0.1 is not "automatic":
The user must detect and solve the problem.

To solve the problem there are 2 options (for this case):
1- Set the initial state of some Gate to High.
2- Set the initial voltage of the Capacitor to some value enough to stop the unwanted oscillation.

In addition some other other changes could be needed, for example set some hysteresis in the inputs of some Gate.

In this case even more things could be done to get a better result, for example adding clamping diodes to the input used in the first Gate of the RC circuit.
With no Diodes the voltage will swing far above and below the Max. and Min. voltages, increasing the oscillation period.

In any case using the RC oscillator in the simulation will substantially decrease the simulation speed and might differ from the real values obtained in a real chip ( every datasheet shows different values ).
So in my opinion the best approach is using a "Clock" component and set the frequency to the exact value you want.


I forgot to mention:

Defran wrote:is it worth spending time on such an obsolete version (0415)?

Version 0.4.15 is the last "stable" version and it is still being maintained (only bug fixes).
Indeed it was updated just 2 days ago: https://www.simulide.com/index.html

Also with R1336... Tell me please.

Your 4060_TEST.sim1 is working for me.

4060 works OK IN R1336, thanks...

But this circuit (same structure as in 4060) In Windows 11 fails ...

The new solutions implemented were discussed in some previous post:

arcachofo wrote:To solve the problem there are 2 options (for this case):
1- Set the initial state of some Gate to High.
2- Set the initial voltage of the Capacitor to some value enough to stop the unwanted oscillation.

Also about the capacitors, note that they are showing a warning:
https://simulide.forumotion.com/t938-simulation-speed-reactive-components

Thank you. Let me study this, if I find something strange i'll let you know.