Are Shunt Capacitors in the Signal Path?

[Edgar]

4 minutes ago, mboxler said:

The current flow will decrease until the voltage across the capacitor is 12 volts, then will stop.

I guess I wouldn't call this blocking DC.

 

 

DC isn't DC if it changes. Going from 0 Volts to 12 Volts is a change.

[captainbeefheart]

7 minutes ago, mboxler said:

 

Agree, but as I understand it no current can pass through a capacitor.

 

 

As I stated above, no current can pass through a capacitor.

 

 

Since a capacitor is an open circuit, current can only flow from one plate to the other via the circuit itself.

 

Current only flows when there is a change in voltage, and the quicker the voltage change, the greater the current flow.  This current flow is instantaneous, and is 90° ahead of the voltage change.  

 

An example that I would be afraid to try would be a capacitor is series with a driver.  Hook this circuit to a 12 volt battery.  The instant the battery is connected, current will flow through the circuit, including the driver.  The current flow will decrease until the voltage across the capacitor is 12 volts, then will stop.

I guess I wouldn't call this blocking DC.

 

Again, this is how I understand it.

 

Mike

 

 

 

 

You cannot think of it in the same terms as a the way a resistor passes current which is different and is in phase. Picture the electrons flowing through a conductor or resistor and back to the source. This exact mechanism doesn't happen in a capacitor, you have this part understood.

 

If you take one plate of the capacitor and electrons flow into it, the only reason they can is because the other plate's electrons are being repelled and pushed out, this allows the electrons to stay on the first plate. You cannot think of it in terms of water, and the water cannot pass through, which for the analogy is correct but we don't care whether the same electrons make it though or not, there is still an electron flow of new electrons from the second plate being repelled by the force. Electrons still move and current still flows and the electrons are from the same source.

 

This is why we can model the capacitor as an impedance the same way we use a resistor. We still view a current passing through the capacitor but not taken so literal as with a conductor or resistor where the same electrons move the complete 'distance' if you will, but that's not important, what's important is there is still a flow of current where the amplifier is providing the work.

[captainbeefheart]

Here watch this video, there is no math but it shows the AC current loop very well and explains why the impedance of a capacitor changes with frequency. Clearly you still have to view AC signal current loop through capacitors but as I said don't take it so literally like with conductors or resistors.

 

 

 

[mboxler]

54 minutes ago, Edgar said:

 

DC isn't DC if it changes. Going from 0 Volts to 12 Volts is a change.

 

I agree,  but it's not AC either.   My point is that the phrase "blocks DC and let's AC pass"  may be misleading to others interested in how capacitors work.

 

 

[mboxler]

23 minutes ago, captainbeefheart said:

Here watch this video, there is no math but it shows the AC current loop very well and explains why the impedance of a capacitor changes with frequency. Clearly you still have to view AC signal current loop through capacitors but as I said don't take it so literally like with conductors or resistors.

 

 

 

 

Actually that's the video that I came across a couple of years ago.  Really helped me understand reactive components.

[Edgar]

4 minutes ago, mboxler said:

 

I agree,  but it's not AC either. 

 

 

Joseph Fourier would disagree. A step function has infinite frequency content.

[captainbeefheart]

Just now, Edgar said:

 

Joseph Fourier would disagree. A step function has infinite frequency content.

 

Yes one of the wonderful mysteries of the universe.

 

Any waveform is nothing more than the sum of multiple sine waves.

[captainbeefheart]

17 minutes ago, mboxler said:

My point is that the phrase "blocks DC and let's AC pass"  may be misleading to others interested in how capacitors work.

 

Why is that misleading? That's exactly how they work. How else do you explain why we can model them as an impedance?

[captainbeefheart]

This is actually a great discussion for circuit theory and capacitors. I'll try and summarize in my own words.

 

When you have alternating current the electrons do not flow through from plate to plate inside the capacitor, this is what is confusing. But one needs to view the capacitor as a whole mechanism and not just view the movement of electrons from one plate to the other as if they were touching. As electrons flow in one direction into one plate they repel electrons from the other plate forcing them to move in the same direction and out of the second plate. As the polarity alternates the flow happens in reverse. Although the electrons do not pass from plate to plate there is still a reaction of same charge types repelling and creating force. For all intensive purposes the capacitor allows AC current to flow within the circuit, there is a current loop through the capacitor, force of electrons still move through the storage of charge and repulsion between plates. As I mentioned we don't really care if it's a continuous water flow of the same water in the loop making the complete trip from source through the loop and back to source. The capacitor still closes an alternating current loop when in circuit to flow in the same manner. So going back to the cap in the woofer network you still view the capacitor as an impedance which mathematically works out to the same amount of current flow within the current loop. The current loop is still source (amplifier) through inductor and since the capacitor is in circuit creates a current flow through it back to source. It doesn't matter if it's the same electrons making the complete journey or not, the current flow is still the same.

[mboxler]

51 minutes ago, captainbeefheart said:

This is actually a great discussion for circuit theory and capacitors. I'll try and summarize in my own words.

 

Thanks!  Your words are much clearer than mine.  I like your use of the phrase "alternating polarities".  I guess in my simple mind the capacitor "creates" AC in the circuit as opposed to "passes" AC.  Perhaps that thought process only works for me. 

 

 

[captainbeefheart]

1 hour ago, mboxler said:

 

Thanks!  Your words are much clearer than mine.  I like your use of the phrase "alternating polarities".  I guess in my simple mind the capacitor "creates" AC in the circuit as opposed to "passes" AC.  Perhaps that thought process only works for me. 

 

 

 

Well, I suppose we use the term "passes" as opposed to "creates" because the source itself is what is creating the current, when you close a loop with a capacitor it "passes" the current from the source through the circuit back to the source. When using a water analogy it's correct to say the water isn't flowing through the capacitor, but electrons are not water. The fact the electrons on one plate are pushing the electrons on the other plate in the same direction there is a flow of energy through the repelling nature of like charges. So in a literal sense no the capacitor doesn't pass current in the same way a conductor does, but it still passes the current flow in both directions at the same frequency as the voltage but out of phase.

 

An open circuit is just that, open and zero current would flow. So if a capacitor behaved truly like an open circuit it wouldn't pass any current, we know that not to be true, when we close a circuit with an AC source we know the circuit is closed and current flows. Of course we know that the plates don't touch and pass the current in the same manner as a conductor but still it behaves like an impedance in circuit and allows current to flow.

 

Ya completely confusing I know. Transformers do not touch primary to secondary and pass current like a typical conductor yet it still pass AC current due to electromotive force. Think of capacitors as a similar phenomena, although the plates don't touch and pass current like a conductor, the like charged particles (electrons) repel each other with force so when the electrons flow into one plate they are only allowed to do so because there is another plate in close proximity and the electrons repel the ones on the other side pushing them in the same direction. So you can say there is current flow "through" the capacitor even though part of it cannot be seen, it is the repelling force of the two negative electrons between the plates pushing current on the other side in the same direction.

 

Hope that helps.

 

 

[captainbeefheart]

Let's look at it as a model.

 

The impedance of the capacitor at 1kHz is 1.59 ohms.

 

From the image I posted where the green line is voltage across the capacitor and the blue line is the current through the capacitor, we can say:

 

V= 1.1v

R= 1.59 ohms

 

Using ohms law

 

1.1 / 1.59 = .705

 

You can see in the simulation the blue current waveform is 700mA so the math works out, the capacitor is passing 700mA of current completing the circuit. It's just another load in parallel with the woofer. The amp has to produce current for both paths, total current from amp is the sum of the two.

[mboxler]

Thanks to all, and yes, It's great that Ohm's Law applies to reactive impedance as well as resistance.

 

Using the components from my earlier post, I plotted the following...

 

Top graph shows current.  Red is resistor, blue is capacitor, and pink is the sum of these two components.  Green is the current through the inductor.  Since the inductor current is equal to the sum of the parallel components, the current through the parallel components is hidden (equal).

 

Bottom graph shows reactive impedance.  Green is the impedance of the entire circuit, and red is the impedance of the parallel resistor/capacitor.

 

I realize it's hard to see, but they closely match the figures in my earlier post.

 

Getting back to the original question, perhaps it's up to the speaker builder to determine the quality of the shunt capacitor.  Or the shunt inductor for that matter.

 

Mike

 

[Deang]

17 hours ago, mboxler said:

…alternating polarities

That’s how I always looked at it since it’s about all my brain can handle. 
 

Good stuff in this thread. 
 

The following is from my notes folder. 
 

https://electronics.stackexchange.com/questions/123906/ground-vs-earth-vs-common-vs-negative-terminal

[captainbeefheart]

I'd say we have proven the video wrong, the shunt capacitor current is not passing through the woofer so it can't have an impact on sound. That doesn't mean the addition of the capacitor doesn't do anything which some comments I read in youtube people confuse as the same thing by making comments like "if the capacitor has no effect of the sound then remove it and nothing should change". That's a completely ridiculous statement, of course the capacitor is doing something for the circuit, it gives the filter a steeper slope, what it also does is give the amplifier another load, a reactive one to need to drive. Can that extra load on the amp make a difference? Possibly, but a film cap is still a reactive load and the amp will see the same thing.

 

Now as far as quality of capacitor, we can see there is some decent current passing through that capacitor, we know electrolytic capacitors have a higher loss angle, this means the more current the capacitor passes the more internal heat will build within the capacitor. The increased heat and high current will cause the life span of a electrolytic capacitor to be shorter vs a lighter duty application. It's cost effective to use an aluminum electrolytic, and it has a smaller footprint. Choose a high quality one rated for high current and low ESR (low loss angle) and it should last quite a while but still it eventually will need to be replaced. The higher the volume you listen to the more current it needs to pass. Remember I was using a 10v input. So yes a film cap would be ideal for longevity without needing to replace components but seriously replacing a $1 capacitor every 30 years is no big deal in my book and worth the smaller footprint and cheaper price. I can solder so that's easy for me to say. 

[DrWho]

On 3/5/2022 at 2:53 PM, mboxler said:

I agree,  but it's not AC either.   My point is that the phrase "blocks DC and let's AC pass"  may be misleading to others interested in how capacitors work.

But it is AC... And the capacitor is letting only the AC component through.

 

This is a very important fundamental principal, and I hope you take the time to grapple with it. I ask a very similar question when interviewing EE candidates. They have a degree and years of experience in electrical engineering, but most of them still get it wrong...and that's unfortunate because i want to hire people with strong fundamentals. My point is that although it's fundamental, it's still not simple. Or at least not simple to me. It took me a few years to really grapple with it.

 

The fundamental equations are always true:

I = C x dV/dt

V = L x dI/dt

 

There is no current if there's no change in voltage. That instantaneous change from 0V to 12V is a change happening in zero time (dt =0)... Which means you're dividing by zero, and that means the dV/dt term approximates infinity.... Which means infinite current if it weren't for the other impedances in the circuit.

 

I think in your mind, you're thinking you have a "DC source" and you're hoping a "DC blocking cap" will "protect" your tweeter from that "DC source". I understand why these abstractions are attractive, but your analogy doesn't include a (mathematical) abstraction for the "switch" that instantaneously connects that DC source to the tweeter. That switch needs a mathematical model too. And as such, the cap blocks DC, but it doesn't "protect" against the AC event that happens when you flip the switch.

 

So if it makes you feel better, think of the switch as a DC to AC transducer. And once you're comfortable with that, then you're well on your way to understanding Class D amplifiers.

 

And then maybe some time after that, you can start exploring "parasitics" where we move from a simple capacitor to adding circuit elements that describe the true physical structure of the cap.... Which has its own ESR, inductance, magnetic fields, shunt capacitance, hysteresis, piezoelectrics, leakage, temperature variance, thermal losses, thermal expansion, etc.... The things that make the different capacitor structures important when choosing a part. You can put most of this in LTspice if you want to take the time to explore it. I certainly do when I'm designing filters.

[mboxler]

21 hours ago, DrWho said:

But it is AC... And the capacitor is letting only the AC component through.

 

I guess in my simple mind it can also convert varying voltage DC into alternating current.  Simpler yet, the capacitor removes the DC offset, if any.

 

Perhaps this graph will explain my fuzzy thinking... 

 

The voltage source is a 100hz sine wave, 20 volts peak-to-peak, with a 10 volt DC offset (green plot).

A series capacitor of 13uf.

An 8 ohm resistor (red plot).

 

As I see it, if the capacitor was not in the circuit, current would only flow in one direction, and that current would vary from 0 to 2.5 amps.  The current through the resistor is not alternating(?).

 

With the capacitor, the now current flows in both directions through the resistor...AC.  This happens because the capacitor's plates are continuously charging and discharging through the resistor and voltage source.

 

As everyone can tell, I'm no EE, so thanks for your patience.

 

[Edgar]

Perhaps a thought experiment will help.

 

Suppose you have a signal that switches from -1V to +1V, then a half second later switches from +1V to -1V, then a half second after that it switches again from -1V to +1V, and so on forever. Is that DC or AC? Easy -- it's a 1 Hz square wave, so it's AC.

 

Now change the frequency so that instead of switching once every half second, it switches once every 50 years. Is that DC or AC? Also easy -- it's a square wave with a period of 100 years.

 

Now change the frequency so that instead of switching once every 50 years, it switches once every lifetime of the universe. You may witness one of the switches, but you'll never witness the following switch back. But it's still AC!

 

The point is, once the level switches at least once, it's no longer DC. It doesn't matter how long it takes to switch back (if ever), it's still AC.

 

[mboxler]

32 minutes ago, Edgar said:

Suppose you have a signal that switches from -1V to +1V, then a half second later switches from +1V to -1V, then a half second after that it switches again from -1V to +1V, and so on forever. Is that DC or AC? Easy -- it's a 1 Hz square wave, so it's AC.

 I agree.  However, if you change the signal to switch from +0V to +2V, is it still AC?  I think, with a series capacitor, current will flow the exact same way in both circuits???

[Edgar]

9 minutes ago, mboxler said:

 I agree.  However, if you change the signal to switch from +0V to +2V, is it still AC?

 

Yes, it's AC with a DC offset.