Exper. Confirm. of Energy Cons., not Charge Cons., in Two-Capacitor Problem

[Transpower]

Based on my paper, "Theory of the Capacitor--Third Revision", http://transpower.wordpress.com, I conducted an experiment to see whether energy conservation holds, rather than charge conservation, in the well-known two-capacitor problem. I used a Rigol 1102E oscilloscope, a Fluke 289 digital multimeter (the most accurate DMM available), and the Elenco SC-750R Electronic Snap Circuits. Capacitor one (electrolytic) has a nominal rating of 100 mu-F; the Fluke measured 102 mu-F. Capacitor two (also electrolytic) has a nominal rating of 470 mu-F; the Fluke measured 476 mu-F. The DC voltage source (lithium batteries) has a nominal rating of 4.5 V; the Fluke measured 5.44 V. The Reciprocal System predicted that with these values the voltage of the second capacitor (in parallel), after the switch is closed, would be 4.936 V; conventional theory says that it should be 4.48 V. The experimental result is 4.938 V; the difference being some residual voltage in the second capacitor. Before running the experiment, I shorted out both capacitors, before charging the first, opening the first set of switches, and then closing the second set of switches. I repeated this experiment several times with the same result. Note that it was difficult to get the voltages of the capacitors down to zero before running the experiment, but I managed to get them down to about 1 or 2 mV. Note also that because of capacitor leakage, it was important to set the Fluke to measure the maximum voltage in order to capture the value before leakage. I also measured the connecting resistance and the current to the second capacitor. The resistance was only .25 ohm, and the maximum current was only 3.864 mA. Joule heating loss (i^2 x R) is thus negligible.

Of course, my experiment will need to be replicated by other scientists. If it holds up it will mean that conventional electrical theory is dead wrong. The electrons in ordinary electrical circuits are actually uncharged, and there is therefore no conservation of charge. There has to be energy conservation! And that's what the Reciprocal System says, and what the experiment has shown.

[cjameshuff]

You appear to have gotten your capacitor values backwards. As you have it, it sounds like the 100 µF capacitor was the one that was charged and then discharged into the other. (This would in fact be the better test, resulting in a larger difference between the two predictions...0.96 V mainstream theory vs. 2.29 V RST. Did you try it this way? Or with two equal-sized capacitors?)

Your stated maximum current is not consistent with your voltage and resistance...given negligible inductance, attaching a capacitor charged to 5.44 V to a 0.25 ohm load would result in a current of over 20 amps, diminishing to near zero in less than a millisecond. An ordinary multimeter is not suitable for measuring such events, and will also likely have difficulty getting accurate measurements of a disconnected capacitor due to self discharge and the load of the meter itself.

Mainstream theory conserves both charge and energy. Your theory in fact does not conserve either. That "negligible" resistor in fact dissipates about 18% of the initial energy in your experiment, 82% if the capacitors are switched. Your theory predicts the capacitors have the same stored energy afterwards, despite this.

[Transpower]

I discharged both capacitors. Then I charged the 470 mu-F capacitor first, then disconnected the voltage source. I then connected the 100 mu-F capacitor, in parallel, and then closed the switches. The maximum voltage reached in the second capacitor verifies the Reciprocal System calculation and disconfirms conventional theory, which calculates a much lower voltage and violates energy conservation. Try the experiment yourself! The Fluke DMM measures 30 samples/second, which is sufficient for this experiment. It is important to measure the maximum voltage which occurs when the switches are closed linking the two capacitors, because leakage begins immediately thereafter into the connecting circuit from both capacitors.

You appear to have gotten your capacitor values backwards. As you have it, it sounds like the 100 µF capacitor was the one that was charged and then discharged into the other. (This would in fact be the better test, resulting in a larger difference between the two predictions...0.96 V mainstream theory vs. 2.29 V RST. Did you try it this way? Or with two equal-sized capacitors?)

Your stated maximum current is not consistent with your voltage and resistance...given negligible inductance, attaching a capacitor charged to 5.44 V to a 0.25 ohm load would result in a current of over 20 amps, diminishing to near zero in less than a millisecond. An ordinary multimeter is not suitable for measuring such events, and will also likely have difficulty getting accurate measurements of a disconnected capacitor due to self discharge and the load of the meter itself.

Mainstream theory conserves both charge and energy. Your theory in fact does not conserve either. That "negligible" resistor in fact dissipates about 18% of the initial energy in your experiment, 82% if the capacitors are switched. Your theory predicts the capacitors have the same stored energy afterwards, despite this.

[Strange]

a Fluke 289 digital multimeter (the most accurate DMM available)

Really? That claim sounds about as plausible as the rest of your theory.

The Fluke DMM measures 30 samples/second, which is sufficient for this experiment.

That sounds extremely unlikely. How long did it take for the voltage on the two capacitors to stabilize?

I noticed you carefully avoided answering any of cjameshuff's points that show your understanding of your own experiment is deeply flawed. Why is that?

There are so many potential sources of error, just based on your vague description of the setup and your apparent poor understanding of theory and how the instruments you use work, it is not too surprising that you got the wrong result.

[cjameshuff]

That sounds extremely unlikely. How long did it take for the voltage on the two capacitors to stabilize?

The meter samples considerably faster than 30 Hz...that is likely the display update rate. The manual states that it can capture peaks at least 250 µs long. That's not nearly long enough, however...it only takes about 20 µs for the current to drop to 36.8% of its peak value in this case.

edit: Hopefully obvious, but I meant to say not nearly fast enough.

There are so many potential sources of error, just based on your vague description of the setup and your apparent poor understanding of theory and how the instruments you use work, it is not too surprising that you got the wrong result.

Indeed. The experimental procedure is poorly described, the chosen setup minimizes the difference between the two predictions, what little is described of the measurement procedure sounds error prone, with the measurement instruments being clearly inadequate for some of the measurements...

[cjameshuff]

And here's my own replication of the experiment, a 470 µF electrolytic discharging to a 100 µF electrolytic through a 1 ohm resistor after initially being charged to 6.2 V from a set of alkaline D cells, measured with a Rigol DS1102E oscilloscope.

RST prediction of equilibrium voltage: 5.6 V
Mainstream prediction of equilibrium voltage: 5.1 V
Measured: 5.1 V.



And the reverse, the 100 µF discharging to the 470 µF:

RST prediction of equilibrium voltage: 2.6 V
Mainstream prediction of equilibrium voltage: 1.1 V
Measured: 1.1 V.

[slang]

And that reminds me why Measurement Technique (literal translation, don't know what it's called in English) was one of the toughest classes during my education, about 20 years ago..

[Transpower]

I've repeated the experiment numerous times today with the oscilloscope on the second capacitor, and in every case the voltage is higher than that predicted by conventional theory. There is a time lag between when I open the first set of switches and close the second set, and this explains why in most of the repetitions the voltage is lower than the RS predicts--but always higher than the conventional theory predicts. Of course, there is also a small voltage loss in the resistance. The faster that one closes the second set of switches, the closer one will get to the RS prediction. There is no "stabilization" here because leakage occurs right after the peak voltage is obtained. Physicists should never doubt conservation of energy. Charges are very easily created and destroyed, and so there is no conservation of them.

And here's my own replication of the experiment, a 470 µF electrolytic discharging to a 100 µF electrolytic through a 1 ohm resistor after initially being charged to 6.2 V from a set of alkaline D cells, measured with a Rigol DS1102E oscilloscope.

RST prediction of equilibrium voltage: 5.6 V
Mainstream prediction of equilibrium voltage: 5.1 V
Measured: 5.1 V.



And the reverse, the 100 µF discharging to the 470 µF:

RST prediction of equilibrium voltage: 2.6 V
Mainstream prediction of equilibrium voltage: 1.1 V
Measured: 1.1 V.

[cjameshuff]

I've repeated the experiment numerous times today with the oscilloscope on the second capacitor, and in every case the voltage is higher than that predicted by conventional theory. There is a time lag between when I open the first set of switches and close the second set, and this explains why in most of the repetitions the voltage is lower than the RS predicts--but always higher than the conventional theory predicts. Of course, there is also a small voltage loss in the resistance. The faster that one closes the second set of switches, the closer one will get to the RS prediction. There is no "stabilization" here because leakage occurs right after the peak voltage is obtained. Physicists should never doubt conservation of energy. Charges are very easily created and destroyed, and so there is no conservation of them.

These plots were not dependent on switch timing. The initial and final voltages are visible right there on the plots. If there was substantial leakage, the initial voltage would simply be lower, and both RST and mainstream theory predictions are based on the actual measured initial voltage.

A resistor carrying current heats up. It dissipates energy that was initially stored in one of the capacitors. If you do the calculations, the dissipated energy works out to half the charged capacitor's stored energy. You can not remove energy from the capacitors and end up with the same stored energy...this is however what RST predicts, and thus RST violates conservation of energy.

As for conservation of charge, if charges were so easily created or destroyed, there would be evidence of such creation or destruction. Such evidence has never been found, and it certainly isn't evident in the operation of capacitors.

[Strange]

I've repeated the experiment numerous times today with the oscilloscope on the second capacitor, and in every case the voltage is higher than that predicted by conventional theory.

You are not going to provide any evidence to support this claim? Or address the fact that someone else with an equivalent setup gets the expected results while you repeatedly get wrong ones? Or the flaws in your own analysis (e.g. "negligible" heating loss)?

As the computer (and scope, DMM, etc) you are using would not work if capacitors did not behave precisely as predicted, we can safely say that, once again, your theory fails the "reality test".

[Transpower]

James, when you run the experiment please display Vmax on the Rigol 1102E screen. I've done this and confirmed that the value is much higher than conventional theory predicts. The resistance and inductance of the connecting circuit elements quite naturally cause the measured value to be somewhat lower than the RST calculation. For example, for many of the runs the DC voltage source and capacitor one were at 5.3768 V, I opened the first set of switches, and then closed the second set of switches. During the time lag, there is leakage from the first capacitor, because it is no longer connected to the DC voltage source. Hence, the longer the delay, the lower the actual voltage to the second loop. The Vmax displayed in this case, for the second capacitor, was commonly 4.64 V. The conventional theory value is 4.42 V, the RST value is 4.88 V. I can explain a drop, but conventional theory cannot explain an increase. I'd very much like to obtain jumper wires with negligible resistance and negligbile inductance and rerun the experiment. I predict that then I will get much closer to the RST value, and further away from the conventional value. If anyone on the forum can suggest where I can purchase such jumper wires I'd be interested. The ones I have from Radio Shack measure .13 ohm, and the Snap-Circuit elements measure .25 ohm.

[Strange]

James, when you run the experiment please display Vmax on the Rigol 1102E screen.

I'm not sure what you are asking for; his plots show all voltages throughout the entire cycle.

[cjameshuff]

As Strange said, the plots show voltages through the point where the switch contact is closed. Leakage (and thus switch timing) is not a factor, the initial voltage is what it is when the switch closes and the oscilloscope triggers. It turns out that leakage in my capacitors is low enough that the voltage is essentially identical between experiments.

It is not in fact inexplicable that the oscilloscope can measure a higher instantaneous peak. Switches bounce, there are high currents when the discharge starts and non-zero inductance. The same basic principles are used in switching power supplies in all sorts of electronic products. In this case, it's irrelevant noise manifesting as glitches at the beginning, and the actual value can easily be read from the plot.

Lower resistance and inductance can be had with soldered connections and short, thick wires. However, lower resistance will make the inductance a bigger factor, due to the higher currents and higher discharge rate. The energy dissipated in the resistance is independent of the amount of resistance...a larger resistance means lower currents for a longer period of time, with the same total energy dissipated.

[Transpower]

Yesterday I didn't use jumpers to connect the Fluke and Rigol test leads to the capacitors, but it was tedious. So, today, when repeating the experiment numerous times, I used alligator clip jumper wires to connect the test leads--this is much easier to do, but of course less accurate than pressing the probes against the capacitor terminals, so naturally the results were not as close to the RST values as yesterday but still way above the conventional theoretical values. I've e-mailed Digi-Key to see if they have a resistance comparison of their various female BNC to alligator clip cables. In the meantime an EE friend of mine has said he will hard-wire the experiment using very low resistance wire.

[Transpower]

Today I bought some insulated alligator clips from Radio Shack (or "The Shack" as they now call themselves) and used one to clamp to the hot terminal of capacitor two and stuck the hot oscilloscope probe into the alligator jaws against the terminal; this setup has slightly lower resistance than using a jumper wire (with two alligator clips) between the probe and the terminal. I then ran the experiment again numerous times; the results were slightly closer to the RST values and slightly further away from the conventional theory values. In the coming weeks and months we can continue to refine the experiment, but I'm very pleased with the results so far!

[amazeofdeath]

How do you reconcile the fact that another independent experiment has already shown a discrepancy with your results in this thread? I'm not sure why you are trying to minimize the resistance in the first place, as the time constant for an RC circuit is RC (duh!), so using a variable load resistance would get you lots more data points.

[Strange]

I'm not sure why he is messing about with alligator clips (1). What is wrong with a switch. Obviously, the experiment is as flawed as the theory as they appear to be in agreement. Although, as he is reluctant to actually post any data I'm not really convinced. Anyone can say, "the results are what I predict".


(1) I've always called them crocodile clips; maybe that's a US/UK thing?

[cjameshuff]

As I've pointed out multiple times, the resistance only changes the time constant...the equilibrium voltage is the same no matter what the resistance is. Trying to minimize resistance only maximizes the inductive and radiative effects and makes the discharge itself more difficult to measure. The capacitors have internal resistance as well...the smaller the external resistance, the more uncertain the exact overall resistance is. And as long as it's high enough that inductive effects are negligible, it doesn't even change the losses in the resistance...halving the resistance halves the discharge time but doubles the current through the resistance, increasing I^2*R by 2 and leading to the same total heat dissipation.

Speaking of which, you still haven't accounted for this energy. It is a simple fact that energy is dissipated in the resistance. How can the capacitors end up with the same total stored energy afterward? Where does the extra energy come from, and how do you reconcile this with your claim that RST conserves energy?

[slang]

Transpower, can you show us some detailed images of how you set everything up?

[cjameshuff]

Transpower, can you show us some detailed images of how you set everything up?

And captures from the 'scope. I've just got stuff plugged into a breadboard, and am getting consistent results that agree completely with mainstream theory. The second plot in particular is quite unambiguous...the result is clearly 1.1V, not 2.6V.

Have you tried it with the 100 µF discharging to the 470 µF instead of the other way around? With the 470 µF capacitor being charged and then discharging into the 100 µF, the RST prediction is only 110% of the mainstream prediction. With the capacitors reversed, the RST prediction is 240% of the mainstream prediction, and the difference in the predictions is nearly 3 times greater in absolute terms.

[John Mendenhall]

Transpower, can you show us some detailed images of how you set everything up?

Yes, please. Trans, your descriptions are inadequate.

[Transpower]

I've attached a scan of the setup. You can see that it's a Snap Circuit, the easiest way to build electric circuits. There is a DC voltage source, two capacitors in parallel, and two sets of switches, plus Snap-2's, etc. I've run the experiment numerous times and proven to my own satisfaction, that there is energy conservation, not charge conservation. I've connected the Rigol 1102E to the second capacitor and the Fluke 289 to the first capacitor, and vice versa. I measure Vmax at both capacitors, after I open the first set of switches and close the second set of switches. I've found that I get results very close to the Reciprocal System values if I connect the Fluke to the second capacitor, rather than the first, and the Rigol to the first capacitor, rather than the second. But, in either case, I always get much higher voltages than conventional theory says I should. This is because conventional theory uses charge conservation, not energy conservation, unlike the Reciprocal System. Once I modified the Reciprocal System space-time dimensions of capacitance from to s to s^3 / t, because of my RC time constant experiments, everything fell into place.

Yes, please. Trans, your descriptions are inadequate.

[slang]

I've attached a scan of the setup.

If I may ask, why a scan? Is this your setup, or did you set it up just like this?

The reason why I asked for detailed images (plural) is that I'd like to see the actual circuit that you built, and the instrument settings. I'm not sure if it's applicable to these instruments or not, but with the various Fluke and other instruments I've used it was possible to get quite wrong results by not configuring them properly.

You can see that it's a Snap Circuit, the easiest way to build electric circuits.

Easier than a breadboard? A matter of taste, I suppose.

[cjameshuff]

I've attached a scan of the setup. You can see that it's a Snap Circuit, the easiest way to build electric circuits. There is a DC voltage source, two capacitors in parallel, and two sets of switches, plus Snap-2's, etc.

Why do you have so many switches? The largest number that would be useful is 3, one for charging the "source" capacitor, one for discharging the "destination" capacitor, and one for connecting the two. You have 4, and none is connected to be suitable for the second purpose, so you may as well have 2.

You also have no series resistance. As already explained, this will make the time constant extremely small, making measurements of the discharge more difficult, in particular making the Fluke useless for measuring things like peak current.

I've run the experiment numerous times and proven to my own satisfaction, that there is energy conservation, not charge conservation. I've connected the Rigol 1102E to the second capacitor and the Fluke 289 to the first capacitor, and vice versa.

I have that model of oscilloscope, and used it for the measurements I posted earlier. It has two inputs. The problems with using the Fluke have been described in detail, why are you not using the oscilloscope to measure both capacitors as I did?

I measure Vmax at both capacitors, after I open the first set of switches and close the second set of switches. I've found that I get results very close to the Reciprocal System values if I connect the Fluke to the second capacitor, rather than the first, and the Rigol to the first capacitor, rather than the second.

Don't you think that getting different results with two arrangements that should be equivalent might indicate a problem with your measurement procedure?

But, in either case, I always get much higher voltages than conventional theory says I should. This is because conventional theory uses charge conservation, not energy conservation, unlike the Reciprocal System. Once I modified the Reciprocal System space-time dimensions of capacitance from to s to s^3 / t, because of my RC time constant experiments, everything fell into place.

Once again, conventional theory conserves both charge and energy. Your theory has been demonstrated to break conservation of energy. Please explain this, and quit ignoring the issue and repeating claims that have already been shown to be incorrect.

[Strange]

You can see that it's a Snap Circuit

What a wacky thing - I've never come across that before. (I am mildly amused by the fact it is aimed at children. And presumably others with little knowledge of electronics ...)

I too am curious why four switches. I am also concerned about if/how you discharge the destination capacitor. The fact you get significantly different results by using the scope and DMM to measure the voltages is also worrying. And then the reluctance to perform the experiment with the capacitors reversed...

It is not too compelling so far.

[ShinAce]

What a wacky thing - I've never come across that before. (I am mildly amused by the fact it is aimed at children. And presumably others with little knowledge of electronics ...)

We use that garbage in our physics labs. Considering the setup is the simplest circuit in the world, there's no reason not to build it properly on a copperclad board.

I've already done this experiment in the past, and the, classical, predicted voltages were always right. Why you wouldn't use the 100uF(charged) to energize the 470uF is beyond me(as previously pointed out). And why you would use such leaky capacitors instead of film capacitors is again, beyond belief. It wouldn't cost much(maybe 5$ for some good caps) to finally prove that all electrical engineers don't know jack about electricity.

[Transpower]

I've attached two jpegs, one from the Rigol and one from the Fluke. 4.56 volts on capacitor one initially, then 4.22 volts after the first set of switches are opened and the second set of switches closed. This is just one result of many, and they fully confirm the Reciprocal System theory. In ordinary circuits, the electrons are uncharged and so the capacitor stores uncharged electrons. In an electrolytic capacitor, there are an equal number of positive and negative charges on the atoms of the electrolyte and so in order for the capacitor to be neutral overall the electrons stored cannot be charged. Also, conservation of energy cannot be violated, whereas conventional theory does do so for the sake of charge conservation. But there are no charges on electrons here! You'd have to go to a Van de Graaff generator for that.

As far as electrical engineering goes, if you use the Reciprocal System space-time dimenions for voltage (t/s^2), resistance (t^2 / s^3), current (s/t), capacitance (s^3 / t), and inductance (t^3 / s^3), you will find that Kirchoff's law, etc., all work fine.

[cjameshuff]

I've attached two jpegs,

These are garbage. Failure to show useful measurements supporting your claims noted...contrast with my own plots posted earlier, which clearly and unambiguously show the capacitors behaving just as mainstream theory predicts. You need to quit stalling and address the points made and questions asked about the problems with your measurement procedure, and show some measurements. This isn't complicated, you have all the equipment required.

Also, conservation of energy cannot be violated, whereas conventional theory does do so for the sake of charge conservation.

Once again, RST violates conservation of energy, and mainstream electrical theory does not. Claiming otherwise does not change this. Stop ignoring the issue and address the problems raised with your claim.

[Transpower]

Garbage? They are the actual results for one run. You don't like it, too bad; I've repeated the experiment now over a hundred times. Try using a Fluke 289 on capacitor two, and the Rigol on capacitor one (with trigger set to 50%), and measure the maximum voltage on each. If you use the two channels of the Rigol you may measure the voltage right after the second set of switches is closed, which will then tell you just the second voltage, not the first and the second. I have answered all the questions--it's you who refuse to measure the proper voltages. Again: it is the maximums which count, because the capacitors start leaking right after they're connected, and so you are not measuring the correct values. As a matter of fact, if conventional theory were right, you should actually be measuring less voltage than the conventional calculation because of losses in the non-ideal real circuit.

These are garbage. Failure to show useful measurements supporting your claims noted...contrast with my own plots posted earlier, which clearly and unambiguously show the capacitors behaving just as mainstream theory predicts. You need to quit stalling and address the points made and questions asked about the problems with your measurement procedure, and show some measurements. This isn't complicated, you have all the equipment required.




Once again, RST violates conservation of energy, and mainstream electrical theory does not. Claiming otherwise does not change this. Stop ignoring the issue and address the problems raised with your claim.

[cjameshuff]

Garbage? They are the actual results for one run. You don't like it, too bad; I've repeated the experiment now over a hundred times. Try using a Fluke 289 on capacitor two, and the Rigol on capacitor one (with trigger set to 50%), and measure the maximum voltage on each.

They're garbage. Your measurements are clearly inaccurate, as evidenced by your own failure to get consistent results when switching the instruments. Numerous flaws in your approach have been pointed out, including your use of the Fluke. Quit wasting everyone's time repeating flawed measurements and do it right.

If you use the two channels of the Rigol you may measure the voltage right after the second set of switches is closed, which will then tell you just the second voltage, not the first and the second. I have answered all the questions--it's you who refuse to measure the proper voltages.

That's nonsense. Have you looked at those plots? They clearly show the voltages immediately before the switch was closed, the discharge curves as both capacitors come to equilibrium, and the final equilibrium voltage. Your supposed problem with using two channels is pure fantasy...this is what multi-channel oscilloscopes are for.

So, I have measured the voltages. That's what those plots in post #6 are. Including the case with the 100 µF capacitor being charged and discharged into the 470 µF capacitor, a variation you apparently refuse to try despite it giving clearer results. (or is your reluctance because it gives clearer results?)

And no, you have not answered all questions...you have not even begun, and have instead ignored them. You are required to answer them. To repeat:

What does the discharge look like on the oscilloscope? Both capacitors, simultaneously. Forget about the Fluke, it's useless for this.

What are the results you get when charging the 100 µF capacitor and discharging into the 470 µF capacitor? Again, measured with the scope.

How do you account for the fact that energy is lost to resistance in the circuit? RST predicts that the amount of stored energy is the same after the discharge, despite some of it having been dissipated, a clear violation of conservation of energy. Do not simply repeat your claim that RST conserves energy.

Show that the mainstream accounting of energy loss does not conserve energy. Do not simply repeat your claim that this is so.

Again: it is the maximums which count, because the capacitors start leaking right after they're connected, and so you are not measuring the correct values. As a matter of fact, if conventional theory were right, you should actually be measuring less voltage than the conventional calculation because of losses in the non-ideal real circuit.

The plots I posted take place over less than a millisecond. Loss due to leakage is minuscule over such a short timeframe. Other non-ideal factors do not affect the final stored energy, only how the remainder of the energy is dissipated.