Does a capacitor charges smooth, or in stairs?

[StillTrying]

Would you post a schematic of the setup you used?

There's not a lot to it! The 5V square is from the same thing I was using to apply 1us pulses to 5mm LEDS, the 390R was the first one I picked up around 500R, and the coax just happened to be ~2.5m long.
There's no GND wire or clip on the X10 probe (SP) at the coax, just the GND ring resting on the 25mm of shield and the probe tip resting on the 390R and core junction.
Captured at 5 Giggle Sa/s I don't think there's any scope ringing in the yellow waveform, I think any ringing is real coming from the large amount of real 68MHz ringing on the 5V square.

[Labrat101]

Would you post a schematic of the setup you used?

There's not a lot to it! The 5V square is from the same thing I was using to apply 1us pulses to 5mm LEDS, the 390R was the first one I picked up around 500R, and the coax just happened to be ~2.5m long.
There's no GND wire or clip on the X10 probe (SP) at the coax, just the GND ring resting on the 25mm of shield and the probe tip resting on the 390R and core junction.
Captured at 5 Giggle Sa/s I don't think there's any scope ringing in the yellow waveform, I think any ringing is real coming from the large amount of real 68MHz ringing on the 5V square.

I did not use the inverter I just connected in series but got the same results as you.
but if you turn on the FFT you can see there are aliases . If you use a 50 cable it must have the same impedance as the scope or you will see all sorts of weird stuff .
My cable was 56ohm when matched it lost all the steps you are not seeing what you think .. Its late we are 2 hrs different  tomorrow . also Dave did some videos on this sort of thing its distortion call what you like but its not
due to the charging effect.

[StillTrying]

I am not knocking your work or the scope picture it was good . 

Knock all you like, it was just a 5 min. experiment.

I can't see any images in your post, or even find links to any in the html.

Yes it's just a transmission line rather than a capacitor.
Vid from previous page https://youtu.be/gdiPwbZm0aM?t=2256

[T3sl4co1l]

Now consider what happens when:
1. Transmission line impedance is very low, and length is very short.  This is roughly descriptive of typical film and ceramic capacitors, where layers are stacked or rolled up, and the ends are shorted across; the electrical length is approximately the component length, no more.
2. Connecting leads or traces have high impedances, and comparable lengths.
3. When these TL segments (such as they can be considered to be) have high HF losses, so that if any oscillations do appear, they dissipate quickly.  What does it mean for a "wavefront" event to dissipate?!
4. When the signal source and/or scope have much less bandwidth than the electrical lengths in #1 and #2.

The TL model doesn't go away, it just becomes more complicated; and potentially less useful -- instead of tracing one seemingly discrete event back and forth (like for the one-end-driven TL stub), there are myriad reflections to keep track of, and the slow-moving signal can no longer be considered as a discrete wavefront, but must be handled as a continuous, tsunami-like flow of charge.

In general, we can split a connection into two ports of some impedance, and consider the incident and reflected waves.  Which are no longer discrete wavefronts, but themselves continuous functions of time -- waveforms, just like any other.  This leads to peculiar cases, like a loose battery forever transmitting continuous power, only for that power to immediately and constantly reflect back, in phase, to the battery, hence deliver no charge.

This is a perfectly equivalent description of current flow, and an interesting one to wrap your head around!  Thought-experimenting with ordinary cases like DC and mains power supplies, is useful for practice itself, if not very useful for practical calculation.

Ed: typos

Tim

[rhb]

Would you post a schematic of the setup you used?

There's not a lot to it! The 5V square is from the same thing I was using to apply 1us pulses to 5mm LEDS, the 390R was the first one I picked up around 500R, and the coax just happened to be ~2.5m long.
There's no GND wire or clip on the X10 probe (SP) at the coax, just the GND ring resting on the 25mm of shield and the probe tip resting on the 390R and core junction.
Captured at 5 Giggle Sa/s I don't think there's any scope ringing in the yellow waveform, I think any ringing is real coming from the large amount of real 68MHz ringing on the 5V square.

Where was the source, where was the resistor and where was the DSO on the transmission line?  I've got  sources which don't ring so badly.  The 11801 calibrator at <20 ps rise time, the <40 ps Bodnar  BNC pulser and a Tek 7104 to check them against :-)

Reg

[jmelson]

It had a bunch of He-3 detectors...

Did you say Helium 3?!   

Yes, indeed.  And, right after the Pons/Fleishman disclosure, it was VERY hard to get He 3 detectors, the only company in the US that made them had a huge backlog.  My boss had to pull some strings to get a few.

Jon

[jmelson]

I've also seen "capacitor quakes".

What I'd like to know, is how long did it take to discover the culprit!

Well, embarrassingly long, I'm afraid.  We were in a hurry to get some measurements, and just hooked it up and started counting.  But, we also hooked the signals up to a scope.  We expected to see signals all of about the same energy, but they were all over the place.  That should have been the clue.
Anyway, eventually, I had to go back and be methodical.  So, we hooked up the bias supply, WITHOUT the detectors.  All should be quiet, right?
But, no, we got the same pattern of pulses.  Then, we saw that they decreased in frequency, as time since bias was turned on increased.  Well, all we had in the circuit was a known clean bias supply, series resistors and coupling capacitors.  A quick look in the literature showed that pulses from high voltage ceramic capacitors was a fairly well-known phenomenon.  But, where are you going to get a .01 uF 3000 V cap EXCEPT ceramic?

Jon

[EEEnthusiast]

Yes indeed it follows steps, when you measure it using any digital equipment. The voltage and currents are quantized by these measuring instruments.
When you use math,  an analog scope or analog meter, they follow an exponential decay.

[T3sl4co1l]

Hah... I'd have gladly given you this one, if I had salvaged it about three decades earlier?...



Tim

[Labrat101]

I am not knocking your work or the scope picture it was good . 

Knock all you like, it was just a 5 min. experiment.

First you miss read   I said I was "NOT KNOCKING"

I spent 5 hours late at night running Tests to simulate theory .
I got a lot of interesting results . But when I have finished I will show the screen shots .
But at the moment my results are not showing what Prof said . I have got dam close.
 But I am not going to do the tests @ 200v so it has to be scaled down to give the same results.
our Prof in the video also mentioned the reflections .
  ('house insurance does not cover Against... 'Mad Scientists' )   
      I have a good stock of Caps & Cables etc

Its been a many years  since I worked on transmission lines . so I want to see if this experiment
can be setup  as per his drawings . With different caps .

   
If one of you guys are Radio hams . You will be better equipped with transmission 

[StillTrying]

I said I was "NOT KNOCKING"

I know you weren't, I was saying you could if you want, I wouldn't mind, it might lead to me making an easy minor improvement that I've missed.

Assuming there's was a long thin plate capacitor, I don't see the problem in it being a just a capacitor at very low speeds, and just a transmission line at very high speeds, the problem would be in the middle were it's it a bit of both.

[Nominal Animal]

Something in all this reminded me of a funny thing when dealing with heat flow (as opposed to equilibrium).  You get similar impedance mismatches that affect how and where the heat is transmitted; it too behaves in many ways as a wave.  Feels very unintuitive that you can change how heat flows from hot to cold by just slipping in a very thin sheet of material with a different heat transfer characteristic, but that's what happens in real life... impedance mismatch, really.
It's too easy to forget that a lot of thermodynamics applies to equilibrium states, and the dynamics are much more complicated and interesting.

Same here.  I can imagine a circuit board having a semicircular disc shaped trace when dealing with a high frequency signal where the staircase is undesirable and a smoother behaviour wanted; the trace then acts sorta like a spectrum spreader – not a filter because it does not dampen any frequencies, it just retards some of the wave so that the arrival of the wavefront is spread out in time.

I'm beginning to think that even if this does not matter for many/most practical circuits, as an educational approach to ensure the students grasp how different models are needed to describe the important features depending on the situation/context, this is superior.

[GlennSprigg]

Wow...  I chose to not read beyond the 1st few replies... 

Let's talk the most simple basics...
Capacitors charge/discharge continuously/smoothly on an 'ascentotic curve'...
which never actually peaks!!  I don't care how long it has been charging, or what
'curve' you draw, it is ALWAYS still charging, but by lesser & lesser & lesser amounts.
There are no 'STEPS'. Just a continuous analog reductive progression. But there's a problem.

Consider a ball thrown perfectly vertical, which eventually comes down again. OK...
Did it ever STOP at the top??  NO!!  It passed ZERO 'speed' for ZERO time!!So it never stopped.
I know where I'm going here mathematically... but do you ??

[Labrat101]

Ok I got something.

I change some or the resistor to try and get a better results.
The Cap is 10nf @ 1.6kv  the resistor is 1k ohm . 
I tried 100ohm it worked but as using lower voltage was hard to get a clear picture.
the pulse Gen is set @ 2Mhz .. 9volt .. 100ns
The resistor is in parallel with the cap.
I did not show the pulse as a second channel as the probe was loading the circuit .
The Sharp peek is the pulse .
I used the analogue as it has a faster response .
Is this the staircase or is it something else     or dodgy cap  ...

[DannyTheGhost]

This reminded me about one thing I was studying in earlier university courses.
In mid 80s scientists started developing SETs - single-electron transistors.
They were creating thin films of Au, typically, above some dielectric thin film on metallic substrate, but they stop the process before Au creates uniformal coating, so there is islands of Au clusters, typically ~10nm. That gives it around tens of aF (attoFarads) of capacitance, which in turn, started creating Coulomb Blockage effect. Then they were using scanning tunnel microscope (STM) and its tip acting like electrode. Coulomb blockade occurs by 'charging' cluster with STM tip and tunneling effect, and because of really low volume of Au cluster when it charges by a 'single' electron tunneling effect stops and there is no current. And, of course, it creates steps in voltage needed to charge or discharge this cluster.
In this extreme example 'capacitor' indeed charges in steps   

[RoGeorge]

Capacitors charge/discharge continuously/smoothly on an 'ascentotic curve'...
which never actually peaks!!  I don't care how long it has been charging, or what
'curve' you draw, it is ALWAYS still charging, but by lesser & lesser & lesser amounts.
There are no 'STEPS'. Just a continuous analog reductive progression. But there's a problem.

Consider a ball thrown perfectly vertical, which eventually comes down again. OK...
Did it ever STOP at the top??  NO!!  It passed ZERO 'speed' for ZERO time!!So it never stopped.
I know where I'm going here mathematically... but do you ??

That's only on paper.  In the real world, at some moment the last electron will take its seat on the capacitor plates (or will leave the capacitor if we are discharging) and that's it.  The process stops, because we can not have a charge smaller than an electron.

As a thumb rule for the real world:
- infinite and zero does not exist,
- forever and nothingness does not exist,
- everything comes in "lumps",
- at some point those "lumps" can not be divided into even smaller lumps,
- there are small and apparently random fluctuations in everything, a background noise that is all over the place, and it seems to be coming out of nowhere, and it can not be dialed down to zero, to me that's the most intriguing feature of this Universe:  the noise and it's randomness.  (not talking here about the microwave background radiation, but about any kind of random fluctuations, either at a quantum or at a macroscopic level)

With enough randomness, I think a whole Universe can be built:  the equivalence between mass and energy is already known, there is also an equivalence between energy and information, therefore it should be possible to generate matter out of information.  We might still need two essential ingredients, space and time.  If you ask me, matter emanates space (similar with a light source that emanates light), and that will also fix the need for black matter and black energy problem, but that's way out of mainstream physics.  The only ingredient left to get is time, and if we buy the idea of spacetime as a continuous, than we can get the time for free, time will be generated by mass, too, together with space.

In conclusion, from randomness alone, we can get some unintended structure here and there, or else said information, and from information we can get energy, then matter, then space and time, and if we wait long enough, we get dinosaurs, and monkeys, and us.   

Now, where from is all this randomness coming, I don't know... 

[Labrat101]

As a thumb rule for the real world:
- infinite and zero does not exist,
- forever and nothingness does not exist,
- everything comes in "lumps", 
- at some point those "lumps" can not be divided into even smaller lumps,
- there are small and apparently random fluctuations in everything, a background noise that is all over the place, and it seems to be coming out of nowhere, and it can not be dialed down to zero, to me that's the most intriguing feature of this Universe:  the noise and it's randomness.  (not talking here about the microwave background radiation, but about any kind of random fluctuations, either at a quantum or at a macroscopic level)

With enough randomness, I think a whole Universe can be built:  the equivalence between mass and energy is already known, there is also an equivalence between energy and information, therefore it should be possible to generate matter out of information.  We might still need two essential ingredients, space and time.  If you ask me, matter emanates space (similar with a light source that emanates light), and that will also fix the need for black matter and black energy problem, but that's way out of mainstream physics.  The only ingredient left to get is time, and if we buy the idea of spacetime as a continuous, than we can get the time for free, time will be generated by mass, too, together with space.

In conclusion, from randomness alone, we can get some unintended structure here and there, or else said information, and from information we can get energy, then matter, then space and time, and if we wait long enough, we get dinosaurs, and monkeys, and us.   
Now, where from is all this randomness coming, I don't know... 

 <sup>Now You are saying when my Capacitor starts to bulge & smoke . This is How the dinosaurs were created ..
   The BIG bang Theory   

      Some How the subject has turned into something else.
        And infinite zeroes is possible but not on your Pay Check 
            means I think you Have been Drinking the IPA again   and strayed from the original subject at Hand ..
           - everything comes in "lumps",     Yes you got that right 1 coffee 2 lumps of Sugar.
 I would Like to know were I can BUY some time from ..</sup>   

[T3sl4co1l]

Ok I got something.

I change some or the resistor to try and get a better results.
The Cap is 10nf @ 1.6kv  the resistor is 1k ohm . 
I tried 100ohm it worked but as using lower voltage was hard to get a clear picture.

A core on the ground lead(s)?  What?

At least draw the lumped equivalent circuit first!

Tim

[Labrat101]

This is special for our Friend   T3sl4co1l

T3sl4co1l
   A core on the ground lead(s)?  What?

If you need to be shown how to wire something up .. don't laugh at the core.

  There was a very nice schematic further up .
      And also the Professor spent an hour drawing it on his blackboard .
          O you slept though it all . Ok  ''it was a bit boring''     

   Put a resistor in parallel with a capacitor  you need a diagram For this.!!    . 
 
   

  At least draw the lumped equivalent circuit first!

     
       I took a photo especial for you . Be careful its really complicated   

[T3sl4co1l]

A mechanical schematic, showing the physical components and connections yes, but that's not what I asked!

Tim

[Labrat101]

A mechanical schematic, showing the physical components and connections yes, but that's not what I asked!

Tim

I suggest you take this question to RoGeorge
Who's idea it was. We have just just tried to figure out if his hair brain theory is correct.
There are No 2 components that will be
Anatomically the same. . (Close enough.)
You would have to use the exact same components for all experiments.
Watch his videos.  And you can decide
How he makes a square peg fit in a round hole .
My experiments proved nothing other than I can pulse a capacitor to failure point.
RNS.
I know what you want .. but these things were not give in any of his documentation.
So that's why his theory is not possible to solve . .
Even two lengths of coax from 2 different rolls Will not be exactly the same.
..
We would all be the same ..
Enjoy.

[StillTrying]

Is this the staircase or is it something else

I think it's probably something else, maybe oscillations (which happen all the time) superimposed on the the discharge curve.

The middle of the video is definitely describing transmission line effects, along the plates of the capacitor, so I chose a capacitor that was 2.5m long.

To get a good view of effects only a few ns wide you have to get rid of all those long wires. To get the GND and resistor as close as possible to the GND and gates of the 74AC14 I soldered them to the back of the board, and arranged the end of the cable so I didn't have to use any GND clip on the scope probe.

[rhb]

Thanks for the photo.  I shall repeat the experiment with either a 1 GHz 7104 and 36 ps step or an 11801 with a <20 ps step and 20 GHz sampling head and varying lengths of coax.  Likely the 11801 as I have some other experiments I need to set it up for.

Both of those should give nice neat staircases as they have very little overshoot.

Have Fun!
Reg

[rhb]

OK  These are with a Tek 11801 & 20 GHz SD-24 sampling head.  I fed the coax stub through a 150 ohm resistor.

10-12 ft coax



50 cm



16 cm



(sorry about the burred image.  I didn't notice it until I posted it)



The 250 mV step has a <20 ps rise time.  The only thing changing is the travel time and the capacitance.  As the coax has approximately the same capacitance per foot, the number of steps stays constant and the only thing that changes is the time delay for each step.  If I expand the vertical scale a lot I can count about 13 steps before the steps fall below the noise and quantization.

Q.E.D.
Quod erat demonstrandum. 
It has been shown.

Have Fun!
Reg

[rhb]

Just for fun, here's a close up of the trailing part of the stairstep for a 30 cm piece of high quality RG402 with >1024 sweeps averaged and at maximum gain.  The little bump at the top of the step is connector saver and wear on the 3.5 mm connector of the SD-24.



Reg