Function generator output oscillations in an RLC circuit

[Braeden Hamson]

Hey guys,

(This post is a bit rambly, read and answer at your leisure)

Right now I'm covering RLC circuits. I had a series RLC circuit connected to an oscilloscope and function generator. I used a BNC T splitter on the output of the FG and a BNC to BNC to connect the FG to the scope. I then had a BNC to alligator which connected to the circuit. I then connected my oscilloscope probe to the circuit at the resistor. In the attached scope screen capture the input signal is in blue and the circuit measurement is in yellow. The RLC circuit was made to be underdamped and I saw the nice oscillations in the waveform. I was measuring exactly what I wanted to see out of the circuit which was nice.

However the oscillations in the input were a bit perplexing. It does make some sense to me, I wasn't expecting a nice square wave. I'm curious to know what the ramifications are, and if there is anyway to mitigate these effects on the output signal. Just waving my flashlight around in the dark here

Lastly, I'd like to run my assumptions by you guys. I have a few ideas of why this is happening. I think current is when there is an absence of electrons in one area and that pulls electrons off of the copper in the wire. Those atoms with missing electrons then pull electrons off the atoms near them and so one and so forth up the wire.) I imagine this process has a kind of inertia to it. That is, if electrons stop being pulled off at the... positive side? then the electrons at the "negative" end of the wire wont stop their shuffle at that same instant. I'll come back to this.

Firstly I'd like to look at the instant after the rising edge of the signal. It comes to a peak and then falls off very rapidly. >1µs. I think this is because when the FG switches to its 10V on state the BNC-BNC between the FG and scope is able to rise to 10V initially, but then the RLC circuit begins to draw current which causes the voltage output of the FG to fall because it can't output enough current to charge it and maintain 10V then when one of the components reaches full charge, I'm not sure which, the slope becomes positive. This is where my understanding becomes a little shakier.

After the slope becomes positive the component that built up its charge is now discharging its voltage into the circuit this voltage is then added to the voltage from the FG which causes the voltage to rise above 10V. If I'm not mistaken this is the basis of a boost converter? After the charge of that component is depleted the cycle starts again. This explains the yellow output waveform.

I think these voltage fluctuations are being applied to the FG's output because its not a perfect voltage source (not saying Tek makes bad equipment, just that is has to obey the laws of physics)

Am I anywhere near the target here?

Haha thanks for any light you can shine on this.
(And as you can see from the capture this Oregonian spent 4/20 in the lab. Nerd Life XD)

 

[kellys_eye]

However the oscillations in the input were a bit perplexing.

What 'input'. You have a LOAD connected across your FG output and the load has RLC elements and will invariably affect the FG output in the way shown.

It's not as if the rising edge of the waveform represents the 'input' and the trailing edge the 'output'...

 

[Braeden Hamson]

What 'input'.

It's not as if the rising edge of the waveform represents the 'input' and the trailing edge the 'output'...

Sorry, i haven't nailed down my terminology.. By output I mean the oscilloscope's measurement. And input is the FGs output.

 

[kellys_eye]

Yes, terminology - it's quite important to get this right as it often leads to misdirection and confusion. I thought you yourself were thinking that the signal had an 'input and output' when this is not the case.

The circuit might have an input and an output - depends on how it's configured* - but whatever you connect to any signal source it then becomes a 'load' on that signal and will change its output in some way. It affects the source signal and (again, depends on how the circuit is configured) can show a difference at the output. The output of your SG will ALWAYS be affected to some degree or other, that's the nature of things.

You have overshoot and ringing as a function of the L and C components - this occurs on both the rising and falling edges of the signal. Suggest you Google the words in bold.

I think these voltage fluctuations are being applied to the FG's output because its not a perfect voltage source

not 'applied' - as stated, the output of the FG (any FG, not just Tek!) is being loaded by the RLC circuit - but 'affected by'. As terminology goes it is important to get this right.

* - if your load is parallel to the source then the input/output will appear identical. If the load was in series (in any way) then you will have an 'input' and an 'output' - but the input will still affect the source signal in some way.

 

[hevans1944]

@Braeden Hamson: It would help us to help you if you provided a sketch (hand drawn and photographed with a cell-phone camera is okay) showing (1) the RLC circuit with labeled parts values (ohms, farads, henrys), (2) how it is connected to the oscilloscope, and (3) how the function generator is connected to the RLC circuit. All three items are necessary. Complete understanding requires knowledge of the make and model of the function generator and the oscilloscope (and its probe) to determine their output and input characteristic impedance respectively.

 

[Braeden Hamson]

@Braeden Hamson: It would help us to help you if you provided a sketch (hand drawn and photographed with a cell-phone camera is okay) showing (1) the RLC circuit with labeled parts values (ohms, farads, henrys), (2) how it is connected to the oscilloscope, and (3) how the function generator is connected to the RLC circuit. All three items are necessary. Complete understanding requires knowledge of the make and model of the function generator and the oscilloscope (and its probe) to determine their output and input characteristic impedance respectively.

Certainly, I detailed the set up of the SG and scope in my original post, and attached my LT SPICE simulation which has the circuit. I'm not sure what kind of probe I have, it's just a cheapo one I got at the school store. But it was set to have 1X attenuation I think its called? However, I'm confident that the scope was performing correctly. I did a comp test, not sure if thats what its called haha but I connected the probe to the scope's 1K square wave output and it had no ringing. I actually think everything in this system was performing correctly.

 

[Braeden Hamson]

Yes, terminology - it's quite important to get this right as it often leads to misdirection and confusion. I thought you yourself were thinking that the signal had an 'input and output' when this is not the case.

The circuit might have an input and an output - depends on how it's configured* - but whatever you connect to any signal source it then becomes a 'load' on that signal and will change its output in some way. It affects the source signal and (again, depends on how the circuit is configured) can show a difference at the output. The output of your SG will ALWAYS be affected to some degree or other, that's the nature of things.

You have overshoot and ringing as a function of the L and C components - this occurs on both the rising and falling edges of the signal. Suggest you Google the words in bold.


not 'applied' - as stated, the output of the FG (any FG, not just Tek!) is being loaded by the RLC circuit - but 'affected by'. As terminology goes it is important to get this right.

* - if your load is parallel to the source then the input/output will appear identical. If the load was in series (in any way) then you will have an 'input' and an 'output' - but the input will still affect the source signal in some way.

Thank you for helping me with my terminology, there are so many things in electronics that all have different names so I gotta get them straight so I don't confuse the poor people trying to help me XD

Yeah, overshoot is what's going on here. From what I've read just now it has to do with the electron's momentum.

Yes, I knew that the FG is being affected by the load and you've confirmed my suspicion that this is common and seems to be unavoidable. And when I mentioned that the FG wasn't a perfect voltage source I meant to say that it wasn't an ideal voltage source like the ones that are in LT SPICE.

 

[timff]

The ringing on your waveform is a result of the inductive and capacitive components attempting to release the energy they stored when the source (FG) polarity reverses.

 

[hevans1944]

I detailed the set up of the SG and scope in my original post, and attached my LT SPICE simulation which has the circuit.

Sorry, I somehow missed that LT SPICE simulation and schematic in your original post. And may I presume that SG in the quote above means Signal Generator, and this is the same as FG or Function Generator?

So, is the FG represented by the voltage source, V1, in the schematic image "Underdamped,JPG" that you just now posted in post #6, and also by V(n001) in the simulated oscilloscope trace located above the schematic in the same image? And, is what you are calling the output represented by V(n003) in the simulated oscilloscope trace located above the schematic in the same image? Let's hope so, but there is no indication in the uploaded image of where V(n003) is measured... perhaps between the connection of R1 to L1 and circuit common or "ground"?

In your first post you mentioned

I then connected my oscilloscope probe to the circuit at the resistor.

Umm. The resistor has two ends. May we assume you connected the probe to the junction of the resistor and the inductor in the schematic, that you just now posted? That would make sense because you already have an oscilloscope input connected to the FG output (the other end of the resistor) using a coaxial cable with a BNC on each end, connected to a BNC "tee" which is connected to both the FG as well as to a coaxial cable with Alligator leads on one end which presumably connects to the other end of the resistor. However, since it requires two points between which to measure a voltage, may we also assume the second point is the oscilloscope common or "ground" input and that this is connected to the "ground" shown on the negative terminal of V1 in your LT SPICE schematic?

You also mentioned in your first post

I used a BNC T splitter on the output of the FG and a BNC to BNC to connect the FG to the scope. I then had a BNC to alligator which connected to the circuit.

This is a pretty standard way to "breadboard" circuits, so are we to assume all those BNCs and their coaxial cable shields took care of connecting all the circuit commons together? Well, that's what anyone well-versed in the art would think, but it pays to be sure. Sometimes a BNC-to-Alligator lead set is used to isolate a BNC ground, although that does not appear to be the case here.

I hope you are beginning to understand some of the limitations of a circuit simulation program when compared against the reality of physical measurements using real components, real signal sources, and real instrumentation. It helps to know what to expect from a circuit, and simulation can provide this. But you also need to understand why the actual measurements depart from the simulation, and whether or not that is important.

Yeah, overshoot is what's going on here. From what I've read just now it has to do with the electron's momentum.

What are you reading? Overshoot has NOTHING to do with the electron's momentum.

Overshoot, or ringing, is all about the exchange of electrical energy, stored in electro-magnetic fields, transfering back and forth between an inductor and a capacitor, mediated by the flow of current between these two components. The electrons themselves hardly move at all in the connecting wires, but their electric fields do, moving nearly at the speed of light, slightly slowed by their environment.

You can probably run faster on an open football field than the individual electrons move in a wire conductor, which for them is akin to you trying to run through a tunnel crowded elbow-to-elbow with people. Things are packed really close together at the atomic and molecular scale of electron motion, so even so-called "free" electrons have a tough time physically moving through a wire without bumping into things.

The series resistor opposes the flow of current, which by definition is charges (electrons) in motion, causing the electrical energy acquired by the electrons in the circuit to be dissipated as heat and the oscillations to be damped (decreasing in amplitude). If everything was built from superconductors, which have zero resistance, the only way the oscillations would be damped would be by electro-magnetic radiation, which is caused by the acceleration of electrons in a conductor. EM radiation is hardly a subject for beginners not well-versed in field theory with a good understanding of Maxwell's Equations.

Near the beginning of my career, I worked as a technician with another technician who was also pursuing an electrical engineering degree on a part-time basis while working full-time. I was just beginning my formal college education and he was several years ahead of me. One day he struck up a conversation about Ohm's Law, with which I was moderately familiar (or so I thought). His completely unexpected comment was "Ohm's Law is unnecessary. Everything in a circuit can be explained by Maxwell's Equations."

Clearly, this was news to me since I had not progressed in my studies far enough to even know what Maxwell's Equations were, much less how to apply them to circuit analysis. I learned much later that just because some things are possible doesn't mean you should embrace them and discard everything else you have learned. It turned out he was right: you can use Maxwell's Equations to describe what happens in any circuit. That doesn't mean you have to, or even that you should want to. If the only tool you own is a hammer, then pretty soon everything begins to look like a nail. We do have a LOT of tools in electrical engineering, and with study and practice you will eventually learn how to use every one of them, as well as know which ones are appropriate to the task at hand.

BTW, an oscilloscope probe that attenuates the input signal by a factor greater than 1:1 needs to be "compensated" with a variable trim capacitor. A pretty good explanation of how this works can be found here in this short application note.