Help me pls - Oscillator Question - Why change the timeline the frequency will be also changed

121

02:09:06

Is it an App bug? When i changed the timeline, the oscillator can’t work or the frequency been changed. UPDATE: I have fixed the bug by add a voltage source with connect null, and it work good now. Thanks @592azy2circuitdude

published 2 years ago

Strav

2 years ago

Takes less time to fill 730pF cap with 9V than with 5V. Less time=higher frequency you are seeing. Notice amplitude of wave also increases with switch to 9V.

wyoelk

2 years ago

Not a bug. Turn off the bottom line that is connected to the base of the transistor from the oscilloscope. Then wait for the display to show frequency. As you slowly increase the time you will see the frequency get closer together. (((NOT changing the frequency just the display.))) As it flatlines, you can increase the cap values, keeping them identical in value, and the frequency will begin to display again.

wyoelk

2 years ago

http://everycircuit.com/circuit/4835982209449984 and basic circuits, http://everycircuit.com/circuit/5224974367064064

592azy2circuitdude

2 years ago

Hi. I'd rather not call it a bug, but it is a limitation of a circuit engine causing wrong results. The problem you described of the frequency changing or not working as the time scale is increased is due to undersampling. As the time scale increases, the sampling step size also increases. If the frequency of the circuit is higher than the sample rate, you get this problem (which is called "aliasing" or a fake signal). In real life, this can happen with ADCs. Is this helpful?

victorjee

2 years ago

Thank you guys. Let me describe the issue in detail: when the timeline is 200ns, the frequency is 12mHz. Then change the timeline to 500ns, STOP and RUN it again, it’s 13mHz. Or save it Close and Open it again, it also is 13mHz. My app is on iPad OS

592azy2circuitdude

2 years ago

Good details. Yes, I see the same thing on my Android device. I'm convinced this is aliasing due to undersampling (i.e. simulation step size or speed is too large). I guess it is kinda an issue, but I think you just are just wanting a more powerful simulator.

Strav

2 years ago

Simuation speed set at 53.6 ns/s (200ns timeline) Frequency: 12MHz Simulation speed set at 90 ns/s (500ns timeline) Frequency: 12.5/6MHz Simulation speed set at 102 ns/s (also 500ns timeline) Frequency: 13.1MHz Notice the 200ns/500ns timeline bar gets longer or shorter. The space between the bars measures the frequency for that sim speed For example, at sim speed 53.6ns/s there are 2 visible waves and 0.5 faded wave as measured by the 200ns bar At 102 ns/s there are 5 visible waves and 1 faded wave measured by the 500ns bar

Strav

2 years ago

It's completely possible that someone else is correct and I'm mistaken. Just looked at one of my own circuits and didn't notice this freq. change behavior when different sim times were used. *shrug*. Sort of confusing.

592azy2circuitdude

2 years ago

I suppose it can be confusing. But I do think it is explained by errors caused by not sampling the frequency fast enough. Here's a cool way to fix it. Add a 12MHz voltage source to the circuit. It doesn't even need to connect to anything. @jason9 uses this all the time to help stability in oscillator circuits (example https://everycircuit.com/circuit/4926685110337536 ).

victorjee

2 years ago

Thank you @592azy2circuitdude ,you’re right! I add a 30mHz voltage source and it work good now. It’s real a bug!

592azy2circuitdude

2 years ago

For the record, I didn't come up with this idea myself. I learned it from @jason9. Glad I could help.

add comment in editor