LC circuits in series with Diodes

[Samson4]

I need help understanding what will happen when the switch in closed in this circuit.

What I want to happen is for Cap B to charge first and then discharge into Cap C.
When the charged capacitor begins to discharge, will it charge Caps B and C at the same time? It will have to overcome the impedance of both Inductors to do so.

<< Mentor Note -- Schematic has been updated, please see the discussion below >>

 

[berkeman]

I need help understanding what will happen when the switch in closed in this circuit.
https://www.physicsforums.com/attachments/221453
What I want to happen is for Cap B to charge first and then discharge into Cap C.
When the charged capacitor begins to discharge, will it charge Caps B and C at the same time? It will have to overcome the impedance of both Inductors to do so.

You show only the left cap as charged, so when you close the switch, not much happens (other than the reverse I_s currents flowing through the diodes)...

 

[berkeman]

Unless the left cap has a negative charge value?

 

[Samson4]

You show only the left cap as charged, so when you close the switch, not much happens (other than the reverse I_s currents flowing through the diodes)...

You mean the charged capacitor won't charge either capacitor? Why is that? Assume the first capacitor is charged correctly for the current to flow through the diode while discharging. The other capacitors are not charged before the switch is closed.

 

[berkeman]

Assume the first capacitor is charged correctly for the current to flow through the diode while discharging.

So you posted the schematic upside-down? The convention for most schematics is for the more positive voltages to be near the top of the schematic...

Can you show us the differential equations for the KCL or KVL analysis? That's the right way to approach this after the circuit is inverted...

 

[Samson4]

So you posted the schematic upside-down? The convention for most schematics is for the more positive voltages to be near the top of the schematic...

Can you show us the differential equations for the KCL or KVL analysis? That's the right way to approach this after the circuit is inverted...

I apologize, I never learned these methods. I have searched them and after a little reading I'll report back.

However; I have corrected the diodes in the schematic.

 

[berkeman]

Have you learned Calculus yet? If not, we can try to offer a simplified explanation (with less accurate results).

Do you have an application in mind, BTW?

 

[Samson4]

Have you learned Calculus yet? If not, we can try to offer a simplified explanation (with less accurate results).

Do you have an application in mind, BTW?

I have yet to learn calculus. As for the accuracy; I just need to know if the capacitors will charge simultaneously or one after the other.

As for the application, I am going to make this very circuit but with connections and diodes that return from the last capacitor to the first. I want to do some experiments with this setup. If the capacitors won't charge 1 by 1 then it won't be useful.

 

[berkeman]

If the capacitors won't charge 1 by 1 then it won't be useful.

No, it will be a continuous re-distribution of the charge. As soon as each diode starts to be forward-biased, it will conduct current in the forward direction.

Instead, look at "switched capacitor" circuits, which use transistor switches to transfer charge between discrete capacitors (used in voltage boosting, or in signal processing, or in image processing, etc.).

 

[Samson4]

Is this what you mean by continuous re-distribution of the charge?
I found this video this morning. Particularly at 3 minutes in, it describes a signal reflecting because of differences in inductance. I'm wondering if using diodes would mimic this reflection by having a higher inductance in one direction.


Im still re-watching it to try and figure out how diodes would effect this. Also wondering how a sparkgap would effect the circuit if it replaces the switch and adding high frequency ac at the resonant frequency of each section of the schematic.

 

[berkeman]

Oh, you are looking into transmission lines? That's different.

Diodes in a transmission line will make it lossy, which is usually a bad thing. Can you just please say what you are wanting to do? That will make this go a lot faster... Thanks.

 

[Samson4]

Sort of like a transmission line I guess. I want the capacitor of an lc circuit to discharge into another lc circuit with identical capacitance and inductance. When the second capacitor charges I want it to discharge into another lc circuit and that one to discharge into the original lc circuit.

The schematic will be standalone. It won't be used to power anything.

If I power it with an ac signal at the resonant frequency of each lc segment, will this satisfy my needs?

 

[anorlunda]

If I power it with an ac signal at the resonant frequency of each lc segment, will this satisfy my needs?

Not without some switching as @berkeman said.

But I wonder why you care if they charge sequentially or not. Why? What are you trying to accomplish?

Are you thinking of something like this? https://en.wikipedia.org/wiki/Voltage_doubler

 

[CWatters]

Sort of like a transmission line I guess. I want the capacitor of an lc circuit to discharge into another lc circuit with identical capacitance and inductance. When the second capacitor charges I want it to discharge into another lc circuit and that one to discharge into the original lc circuit.

If this is a circuit made of discrete components it will quickly settle to an equilibrium state. I think you need switching and amplifying elements between the stages.

 

[Samson4]

I like the idea that the energy is moving in one direction at all times. This is only possible if they charge sequentially. I could do this with electromagnetic waves but the wave lengths will be huge or the radiation is problematic.

That's 3 times I've been told I'll need switching. Back to the drawing board.

If this is a circuit made of discrete components it will quickly settle to an equilibrium state. I think you need switching and amplifying elements between the stages.

Wouldn't equilibrium mean that all capacitors are discharged?

 

[anorlunda]

I like the idea that the energy is moving in one direction at all times. This is only possible if they charge sequentially.

Energy can split and flow in two branches in parallel. It's like the water in a river going around both sides of an island. I think you need a course on basic circuits because you are confusing yourself with some incorrect concepts.

 

[CWatters]

I suggest you try and simulate it.

If the first capacitor was replaced with a DC voltage source then the following caps would end up at a slightly lower voltage due to the diode drops and other losses. However it's not a voltage source, the source voltage will be falling as the current through the first inductor increases.

Exactly what happens will depend on the various time constants. It's likely the caps will end up with some DC voltage but that may leak away rapidly through the (non-ideal) capacitors.

It sounds like you are trying to build something like a ring oscillator? Bucket brigade device? Delay line?

 

[Samson4]

Energy can split and flow in two branches in parallel. It's like the water in a river going around both sides of an island. I think you need a course on basic circuits because you are confusing yourself with some incorrect concepts.

What do you mean? If I'm standing on that island all the energy is moving in the same direction.

I suggest you try and simulate it.

If the first capacitor was replaced with a DC voltage source then the following caps would end up at a slightly lower voltage due to the diode drops and other losses. However it's not a voltage source, the source voltage will be falling as the current through the first inductor increases.

Exactly what happens will depend on the various time constants. It's likely the caps will end up with some DC voltage but that may leak away rapidly through the (non-ideal) capacitors.

It sounds like you are trying to build something like a ring oscillator? Bucket brigade device? Delay line?

Thank you for those terms. I'll have to search them these evening when I get off.

 

[Merlin3189]

Have you thought of trying neon bulbs instead of the diodes?
If I'm showing my age too much, perhaps diacs would behave similarly. (Though I haven't seen one of those for decades either! Perhaps there is a modern equivalent.)

 

[CWatters]

https://en.m.wikipedia.org/wiki/Bucket-brigade_device

 

[CWatters]

As for the application, I am going to make this very circuit but with connections and diodes that return from the last capacitor to the first.

Not really sure what that will achieve. At every stage there are losses in the diode and resistance of the inductor so any energy that starts in the first capacitor will rapidly be lost as heat. The energy won't keep going round and round if that's what you were hoping for.

 

[Samson4]

Have you thought of trying neon bulbs instead of the diodes?
If I'm showing my age too much, perhaps diacs would behave similarly. (Though I haven't seen one of those for decades either! Perhaps there is a modern equivalent.)

I understand using the diac for the switch; but, why use the neon bulbs instead of diodes? I was contemplating a sparkgap for the switch so the diac is actually more modern than what I had planned.

 

[Samson4]

Not really sure what that will achieve. At every stage there are losses in the diode and resistance of the inductor so any energy that starts in the first capacitor will rapidly be lost as heat. The energy won't keep going round and round if that's what you were hoping for.

I'm not expecting perpetual motion but I am expecting to measure the resistance around the circuit and use ohms law to estimate how many times around the circuit the current will flow.

 

[Merlin3189]

Diac or neon bulb is like the spark gap which will work at lower voltage. A neon bulb breaks down at about 100 V, a diac lower at about 30 V.

If I understand your intention, the first capacitor would be charged well above breakdown voltage. When the switch is closed, the first device would break down and conduct, charging up the second capacitor. When this capacitor reached the breakdown voltage, the next device would start to conduct and charge up the third capacitor.
In principle I suppose this could be extended to further stages, though with the neons there would be a large voltage drop at each stage due to the sustain voltage. Probably the diacs would be better in that respect. You might need to add resistances (or maybe inductors) to limit the peak currents.

 

[Baluncore]

Here is an LTspice simulation of the response of a lumped LC transmission line with series diodes. Initial C1 = 12 volts, all other capacitors zero volts. Notice how the energy is transferred from C1 at the start to C7 at the end of the line. Diodes prevent reverse flow but causes losses.

LTspice circuit.asc and plot.plt files attached, remove the .txt extension to run.

 

[Baluncore]

Since that simulation worked with a switch, here is one that uses a trigger pulse and recirculates the wave.
Here is the lumped transmission line that includes diodes and a feedback link. Less stages can be used.
The short rectangular pulse is used to insert the initial wave energy, that pulse is then turned off and the sine wave allowed to recirculate.
The wave voltage has been increased to dominate the diode voltage and so reduce attenuation with time.
Notice that the diodes force the wave to circulate in the forward direction only.

LTspice files attached, remove the .txt extensions to run.

 

[Samson4]

Since that simulation worked with a switch, here is one that uses a trigger pulse and recirculates the wave.
Here is the lumped transmission line that includes diodes and a feedback link. Less stages can be used.
The short rectangular pulse is used to insert the initial wave energy, that pulse is then turned off and the sine wave allowed to recirculate.
The wave voltage has been increased to dominate the diode voltage and so reduce attenuation with time.
Notice that the diodes force the wave to circulate in the forward direction only.

I found this on wiki about Grain-Oriented electrical steel:
" The magnetic flux density is increased by 30% in the coil rolling direction, although its magnetic saturation is decreased by 5%. It is used for the cores of power and distribution transformers, cold-rolled grain-oriented steel is often abbreviated to CRGO."

My thinking is that a transformer core made with the laminations oriented in the same direction would have an asymmetrical inductance. In your last circuit, replacing the diodes with such inductors would restrict the current in 1 direction linearly in respect to the increased inductance in that direction. CRGO cores are obviously not good enough for this use because the asymmetry is not pronounced enough. However, my point is that it is possible. I have no idea if such components exist already. I have had no luck in finding them if they do. I did sketch one up though. Maybe something like this would work:

The high inductance coil behaves like a toroid. It's ends are closer together than the low inductance coil. Shouldn't this mean more magnetic energy can be stored in it's field? If it was designed to have 10 times the inductance in one direction than the other; it should be capable of replacing the diodes in the circuit.
Does this sound correct or like nonsense?

 

[Baluncore]

@ Samson4.
What are you really trying to do here? It seems like you need complexity beyond understanding.

I do not think you will be able to replace the diodes by using saturable core inductors with a fixed magnetic bias. Saturable cores and magnetic amplifiers were in vogue in the 1930s and 1940s. They will add to the complexity rather than reduce it.

 

[Samson4]

@ Samson4.
What are you really trying to do here? It seems like you need complexity beyond understanding.

I do not think you will be able to replace the diodes by using saturable core inductors with a fixed magnetic bias. Saturable cores and magnetic amplifiers were in vogue in the 1930s and 1940s. They will add to the complexity rather than reduce it.

Those were not saturable cores. They are permanent magnets that allow the magnetic field of the coil to terminate on it's poles; or circle around the exterior of the coil.

 

[berkeman]

What are you really trying to do here?

@Samson4 -- Please let us know what you are trying to accomplish.

 

[Samson4]

I want an apparatus that I can experiment with. I am most interested in; the wave nature the capacitors will charge with, the momentum of the moving charges, thermal coefficient of resistance. Basically, an analog of em waves that have acceptable wavelength without producing harmful radiation. It is ideal if I can keep the time between input pulses long.