Why resonant frecuency signal disappears

Honestly, I'm not sure. But here's my best guess. I think EC is wrong, and LTSpice is right this time. Because there are no resistive losses in the circuit, there should be no damping of the natural response (resonate frequency). This question on the electrical engineering stackexchange, "Transient Current in an LC Circuit with a DC Supply," may be helpful. I hope others weigh in on this, too.

Might be because there is no resistance built in the power supply. I really don't know if that is it

In EasyEda and LTspice, with resistance or without, the resonant frecuency is present in steady state. In wikipedia "LC circuit" entrance, they show differential ecuations with the frecuency permanent response.I think that maybe there is a bug in Everycircuit...

From what I understand, the resonant component fades because the frequency source doesn’t actually match the resonance frequency of the values your using here. Because of that, you see a bit of almost noise at start up (2 mixed conflicting Frequencies), but it dissipates quickly, and begins to only pass what it’s been fed on the input. What you have right now seems to be acting more like a Band-Pass filter that’s operating close to the edge of the actual pass frequency, resulting in just a dampened version of the input signal. The actual resonant frequency I calculated from the values you currently have is: 225.0791 Hz (for 500mH & 1uF). Change the source to that frequency and watch the results differ drastically

Hmm, these are some interesting points you brought up, Issac, but don't think that is what's going on. Here's why. If 2 frequencies are combined, one will not act to dampen the other. They will merely interfere. I think you're correct that this is a filter (run a frequency sweep and see the resonate gain). But filters do not dampen/dissipate frequencies but rather attenuate/reduce them.

Still, I may be off here, but I think what is confusing about this circuit is that the driving/forcing function is a sine signal. Check out this circuit I put together, which is the same sort of thing, but uses a step function instead (DC). https://everycircuit.com/circuit/6429033277161472 Thoughts?

Regarding the Bode curve (pause->f), your sine source is not set on the resonant frequency (225 Hz).

My bad on some of the terminology there. What I was trying to describe is the starting conditions… The frequency source is the dominant frequency simply because it’s the source with infinitely low impedance; and because it doesn’t match the resonant frequency of the LC filter here it is not going to resonate… BUT, you can see at start up where the source excites the LC circuit and the product becomes a mix of the two frequencies. Early on you can actually zoom in enough to see the higher (resonance frequency) superimposed onto the sine wave source. But because they don’t match, it can’t sustain the other component, (the one that’s not a source; merely a byproduct) so it dampens, and instead you soon get just the source frequency.

Ok, I get what you mean. I agree that the source is the dominant/driving function that excites the natural LC response. The frequencies must match to sustain oscillations, IF the natural resonate frequency decays due to resistive losses. But there is NO resistor in this circuit, so shouldn't the resonate component last forever after being excited once? Did you look at the circuit I linked to above?

Ah I see what your saying now. I think this is simulating properly, because there’s no isolation between the positive and negative transitions of the source frequency, so because the source and resonance frequencies are different, there’s gonna be a lot of out-of-phase feedback that kills the oscillations regardless of any resistance (present or not). You would still need an amplifier and a mixer to sustain and superimpose them… or a simple but clever little way introducing feedback from the source into the LC only when the phases line up so that it the resonant oscillations sustain. Oh and I checked out your link now, I didn’t realize you included one earlier, looks like a good visual comparison of the effects with various series resistors.

I didn't consider negative feedback or lack of isolation. Thanks for checking out my circuit link. The point I was trying to make with it was that the LC circuit with no resistance should oscillate forever, but it doesn't. I thought the same 'bug' was happening here. But I guess I need to think about it some more...

Or you can just modify it as I’ve done in this example to avoid needing an amp and a mixer, so you can sustain the oscillations using only the same frequency source: http://everycircuit.com/circuit/5071126350528512

The only slight difference worth noting to the above approach is that the extra components (particularly the additional capacitor) will have some affect on the resonance frequency and amplitude, how adverse depends on the value(s) chosen

Ok, I tried to do some research. Your circuit was helpful. I found out that to start oscillations, the LC circuit must be "kicked" at the resonate frequency. The oscillations will fade if the frequencies don't match. Your circuit demonstrates that well. BUT I still think that fading would not necessarily be caused by negative feedback but rather by resistance.

I'm looking at it like this. The input sine function and resonate LC frequency coexist at the beginning. The LC "pushes back" just as hard as the source because, at resonance, the impedance is 0. Thus, they should continue together. Back to the original post, LTSpice shows it continuing, and EC doesn't. I still think LTspice is correct.

Hmm interesting, I’m not too sure about why it’s simulating so differently in LTSpice to be honest… But regardless, the series LC also has to have a return path that must lead back to the inductor node in order to oscillate the energy that was built up in the capacitor/inductor back and forth, which can only happen alternately, between the source transferring current to and from the resonant LC. Because the current is in series with everything here, the only return path is through the source. You cannot just examine the two sides as each a source that push back or adds onto each other, but rather have to examine it as net positive and negative magnitudes of current. Maybe try the LTSpice simulation again with varying sim speeds and see what happens over a longer stretch of time with and without some series resistance..?

Good ideas 👍. As soon as I find time, I'll try out LTSpice with your suggestions. I totally get that the current must flow back and forth through the source, which is where the non-isolation feedback comes in. The reason I figured that I could examine an LC circuit as 2 parts that superimpose is because "natural response + forced response = net response" (I think that was in my Circuit Theory class). However, I will look into it further...

I tried simulating everything in LTSpice, and I will say, it does seem to act more like you’ve been trying to describe, even with resistance no matter how much. and I honestly don’t understand why cause I’m pretty sure this wouldn’t be the case in real life unless you were close enough to the resonant frequency. I’m gonna admit, I’m feeling pretty clueless now, and I’m kinda out of explanations.

Do you have a function generator at home? Cause if so, I’d like to settle this once and for all in real life rather than comparing simulations. I could do it with what I have, but I’d have to build my own Sine Wave generator and buffer it with a push-pull (and probably hand wind a 500mH inductor; or just try it using different frequencies/values) to really be able to test this.

What?! Frankly, I'm a little confused myself. No matter how much resistance it still oscillates? I can say one thing for sure. In real life, the oscillations would not continue by themselves because that would be a perpetual motion machine! Unfortunately, I have no signal generator, so I can't run any tests.

OK, I experimented with LTSpice and Proto (which I highly recommend), and I got results that seemed to confirm my theory (so I think). With no series resistor, the resonate frequency stays present, and with R = 10ohms, the resonate frequency fades out. Please let me know if you simulate again and/ or if you build it. My curiosity is aroused 😄.

Haha yeah will do. I think I’ll build it if anything, I’ll design something within a day or two and post it here so you can test it as well if you’d like!

Alright I finally got around to testing it, I’ll say i didn’t have that large of an inductor for this but I got as close to it as I could by using part of a 60Hz transformer, results weren’t all that exciting I’ll admit, but nevertheless, the 50Hz signal generator was a fun little design challenge. First, it might be important to note that the driving capability of the circuit I was using for this is nothing great, only about 100 Ohms, but I think that’s totally fine for this. As for the results… not too much to say; it was simply an attenuated version of the 50Hz driving signal, no mixed resonant frequency was present from what I could see (however I didn’t try doing a single shot trigger, which may have possibly revealed different initial results)…. As for the testing circuit, I’ll link 2 schematics, the 1st will be basically a replica of the circuit I used (with the exception of one additional amp stage that wasn’t actually necessary to function), and the 2nd will be just a simplified version, to make it easier to see how it generates the signal in case anyone is curious or would like to build it… Actual: http://everycircuit.com/circuit/5284666655113216 Simplified: http://everycircuit.com/circuit/6714681921896448

Yep, those are the results I expected, too. Guess some things are only possible in simulation 😁.

Yep ;) guess so

Ok. It's not really a matter of specific values. If you change the inductor to 100mH, the source to 60Hz and the capacitor to 10uF the effect is still there. The tempos are different, but it is qualitatively the same phenomenon. I'm very interested in knowing how is possible that LTSpice, a very highly regarded software, is making mistakes in the simulation of suck a basic circuit. Which is right, LtSpice or EveryCircuit?

Did you read the thread? I don't know if Isaac agrees with me or not, but I'm convinced that LTSpice is correct. The signals would both continue forever in the absence of resistance (something which never happens practically).

I know this is a little late, but I came across two circuits recently that really helped make sense of this (at least for me...) https://everycircuit.com/circuit/6127621154799616 This one shows how the resonate frequency fades with resistance (like in real life) but remains without resistance (like in simulation).

https://everycircuit.com/circuit/5428841145171968 This one shows the effect of making the driving function close to the resonate frequency (both signals build off each other). But making the frequency differ will result in interference. But both frequencies remain present.

Now, I think that the reason LTSpice and EC differ is because EC is undersampling the resonate frequency. If you slow the simulation WAY down, EC gives the same result as LTSpice. Ok, that's my best guess! Good luck.