ACK! Do not confuse the volume of a cylinder with the amount of capacitance. It just muddies the waters. In this conversation, it is sufficient for capacitors of equal size have equal capacitance.

You could type for a week explaining all the different dielectrics, the advances in thin film technology in the last thirty years, and the different properties of each mechanical configuration, but it will not help Mr. Martinez with this question.

I'm not confusing the volume of a cylinder with the amount of capacitance -- I'm pointing that very thing out by showing an example in which the two do not relate, namely two different capacitors that have the same volume but different capacitance.

The notion that volume equates to capacitance has numerous problems. First of all, it creates the impression that the charge is stored in the volume between the plates instead of on the plates. Second, for a given configuration, the capacitance goes up as the plates are brought together, hence by decreasing the volume you are able to store more charge for the same voltage.

 

The notion that volume equates to capacitance has numerous problems.

You're right. Maybe I should've used the word area instead? And as I said earlier. I have no problem accepting different capacitive properties for different materials and configurations.

In the meantime, I'm carefully reading #12's suggested article. Thank you all for your help.

 

Without rereading a buttoad of datasheets, my general impression is that power MOSFET datasheets are trending toward gate capacitance, or at least adding gate capacitance to a parameter table that already includes gate charge. Makes it easier for gate driver circuit designers to figure out the timing.

ak

 

Without rereading a buttoad of datasheets, my general impression is that power MOSFET datasheets are trending toward gate capacitance, or at least adding gate capacitance to a parameter table that already includes gate charge. Makes it easier for gate driver circuit designers to figure out the timing.

ak

Believe it or not, all of this confusion of mine began when I noticed that some datasheets specified capacitance, and others gate charge. That's when it downed on me that I didn't have a perfectly clear and transparent understanding of the difference.
Thanks for the tip!

 

One thing that might help is to consider a similar situation with batteries. They are rated in ampere-hours and this is often used as a measure of the energy stored in the battery. But ampere-hours is a measure of charge, not energy. The linkage is that the delivery of that charge occurs at, roughly, a fixed voltage and so the ampere-hour rating is a usable proxy for energy, particularly for batteries of the same, or nearly the same, voltage.

If I have a gate capacitance then it is a fairly simple matter to determine how strong my driver has to be in order to switch the gate from "off" to "on" in a certain amount of time (and answer comparable questions for more analog type behavior). But two things complicate this picture -- the capacitance of a MOSFET gate is highly nonlinear, so is the given capacitance the total capacitance at a particular voltage, or just the dynamic capacitance? It can make a significant difference. Second, the computations to determine the drive strength are slightly more involved (just slightly). But if I have the total charge that I have to move onto or off of the gate in order to switch it, then I can just total that up for all the FETs and divide that by the desired switching time and I have a good first cut at my needed drive strength.

 

One thing that might help is to consider a similar situation with batteries. They are rated in ampere-hours and this is often used as a measure of the energy stored in the battery. But ampere-hours is a measure of charge, not energy. The linkage is that the delivery of that charge occurs at, roughly, a fixed voltage and so the ampere-hour rating is a usable proxy for energy, particularly for batteries of the same, or nearly the same, voltage.

If I have a gate capacitance then it is a fairly simple matter to determine how strong my driver has to be in order to switch the gate from "off" to "on" in a certain amount of time (and answer comparable questions for more analog type behavior). But two things complicate this picture -- the capacitance of a MOSFET gate is highly nonlinear, so is the given capacitance the total capacitance at a particular voltage, or just the dynamic capacitance? It can make a significant difference. Second, the computations to determine the drive strength are slightly more involved (just slightly). But if I have the total charge that I have to move onto or off of the gate in order to switch it, then I can just total that up for all the FETs and divide that by the desired switching time and I have a good first cut at my needed drive strength.

That was a pretty good explanation, thank you. I didn't know about a fet's gate capacitance non-linearity. That explains why charge is a better parameter to consider instead.

It just occurred to me that a capacitor can be compared to a mechanical flywheel, in which the moment of inertia of the mass would represent the charge, and its rotational speed the voltage. A flywheel's rotational energy equation is:

​

And a capacitor's stored energy is:

​

The equations are quite similar, and it makes it easier for me to visualize things that way.

 

A closer mechanical analogy for a capacitor is a spring under tension.
Mechanical inertia is the analogy for the energy stored in an inductor, 1/2 LI^2.

 

A closer mechanical analogy for a capacitor is a spring under tension.
Mechanical inertia is the analogy for the energy stored in an inductor, 1/2 LI^2.

You're absolutely right. I've always related inductance with inertia, too.
Thanks.

 

You're absolutely right. I've always related inductance with inertia, too.
Thanks.

The reason why Mosfet gates are measured in coulombs instead of farads is because the voltage change over time is not linear.

BUT, the voltage of a capacitor over time is linear so it is easy to predict what the voltage will be given x amount of coulombs.

So datasheets for Mosfets simply specify "if you give x amount of coulombs, the voltage will be x"

 

The reason why Mosfet gates are measured in coulombs instead of farads is because the voltage change over time is not linear.

BUT, the voltage of a capacitor over time is linear so it is easy to predict what the voltage will be given x amount of coulombs.

So datasheets for Mosfets simply specify "if you give x amount of coulombs, the voltage will be x"

I don't follow what you mean when you say that the voltage of a capacitor over time is linear, especially after having just stated that the voltage change over time is not linear.

The capacitance of a MOSFET gate is highly non-linear. How the voltage changes with time has no bearing on that. Since the capacitance is non-linear, the best way to describe the capacitive characteristic of the gate (there are other ways) is to provide data points of the stored charged versus the Vgs. Ideally this would be done with a plot, but providing important data points from it is often sufficient.

 

The size of a bucket is the capacity in Farads. The amount it is filled is the charge in Coulombs.

So, which should happen first?
Should the overvolted capacitor blow up or the bucket overflow?

 

So, which should happen first?
Should the overvolted capacitor blow up or the bucket overflow?

What does it mean for the bucket to overflow (in the case of a capacitor)?

As you charge the capacitor to higher and higher voltages, more and more charge will be stored on it. Eventually, something will give. What, exactly, that is will depend on the capacitor and the specific circumstances.

 

What does it mean for the bucket to overflow (in the case of a capacitor)?

As you charge the capacitor to higher and higher voltages, more and more charge will be stored on it. Eventually, something will give. What, exactly, that is will depend on the capacitor and the specific circumstances.

The charging and discharging of a capacitor are asymptotic processes. The voltage across a capacitor will approach, but never actually reach the limiting value. As more charge is added the voltage gets closer and closer to the limit. The same happens with the discharge it will get close to but never actually reach zero. The bucket will never overflow or underflow. It will get close enough to either limit to defy your ability to measure the difference. Think zeptovolts or yoctovolts.

 

So, which should happen first?
Should the overvolted capacitor blow up or the bucket overflow?

Define overflow.
If you change a capacitor rated at 6.3v much beyond, it will start leaking and may pop as the dielectric can no longer keep the charges separated.
If you have a different definition of overflow, let me know.

 

The charging and discharging of a capacitor are asymptotic processes. The voltage across a capacitor will approach, but never actually reach the limiting value. As more charge is added the voltage gets closer and closer to the limit. The same happens with the discharge it will get close to but never actually reach zero. The bucket will never overflow or underflow. It will get close enough to either limit to defy your ability to measure the difference. Think zeptovolts or yoctovolts.

The post I was responding to specifically asked about an "overvoltaged capacitor", which I took to mean a capacitor that was driven to a voltage above its rated specifications.

 

 

I liked the footage of the balloon the best.

I couldn't see their setup very well, but it looks like they were using a variac as their power source. I wonder if they were rectifying it or applying AC to those polarized caps.

They should also have done some dipped tantalum caps -- those explode very nicely and occasionally give some beautiful and rather large flame jets.

 

I liked the footage of the balloon the best.

I couldn't see their setup very well, but it looks like they were using a variac as their power source. I wonder if they were rectifying it or applying AC to those polarized caps.

They should also have done some dipped tantalum caps -- those explode very nicely and occasionally give some beautiful and rather large flame jets.

I want to get a slow motion camera and start filming stuff. It is absolutely incredible how much happens in an 'instant'. With the ease of YouTube channels and the widespread audience, random videos would almost be a guaranteed money maker.

You make a good point about if it was AC or DC. I'm pretty sure it's a Variac as I've seen other videos. Nonetheless someone could make two experimental videos and I'm sure they would each get a lot of views. How many to pay for the camera and break even? Haha

Anyone here run a YouTube channel?

 

I want to get a slow motion camera and start filming stuff. It is absolutely incredible how much happens in an 'instant'. With the ease of YouTube channels and the widespread audience, random videos would almost be a guaranteed money maker.

You make a good point about if it was AC or DC. I'm pretty sure it's a Variac as I've seen other videos. Nonetheless someone could make two experimental videos and I'm sure they would each get a lot of views. How many to pay for the camera and break even? Haha

Anyone here run a YouTube channel?

The cameras are so expensive that they don't even publish the prices, but most sites seem to converge on a starting price of $150,000. Most filmmakers rent them at the tune of about $2,500/day.