capacitors do not conduct in DC circuits

[rigidigital]

After being told that capacitors do not conduct in DC circuits I am even more confused. The simple LED flashing circuit i built seems to use the capacitors to 'time' the flashing of the lights. I guess they receive current until their limit is reached and then let it go in a bang ?

On the other hand I built a circuit that used a capacitor to connect the negative and positve early on in the circuit for a microcontroller. I thought this would have made a short circuit but of course it worked, the controller ran as was meant to !

Ive also read it can be a good idea to connect a capacitor to both points on a small electric motor to smooth out the current ? Huh

 

[Michael 2727]

Think of a Capacitor as a bucket with a baloon
stretched across the opening.
Once full it that's it no more will go in but it can
come back out. This is where they are used to store
energy as a filter/bypass Cap and it is released when
the supply drops, acting like a battery to keep the
supply level smooth-er.

Only a fixed amount can go in and back out.
(unless you exceed the cap V rating and blow
the ass, sorry bottom out of the bucket )

In the case of a Motor Back EMF they reduce the
high voltage spikes by dumping them into the bucket.
When charging any capacitor the very first few % of
the input charge sees a very high resistance which
drops very quickly as the capacitor charges upward.

Google - Capacitors, how they work, tutorials etc.
There are 123456 Trillion pages of stuff and much
of it is very good.

 

[hippy]

A capacitor is really like a bucket. It is empty to start with and can fill up with charge until it is fulll. The rate at which it fills up depends upon the size of the hose filling it ( a current limiting resistor ) and the bucket also has a small hole in it through which it slowly leaks ( self discharge ).

For the Led flashing circuit there is probably something in there which fills the bucket with charge and then kicks it over when full to control the flashing of the Leds, the process then repeats, so your basic analysis would be correct.

When used as a 'reservoir capacitor', when put across the supply rails, it is effectively a dead-short, but only for a very small fraction of time. As soon as power is supplied, current rushes in to the capacitor and it ceases being a short.

With a bucket full of water, stabbing your finger in it usually only causes small ripples, but overall the water level doesn't change. This is similar to the smoothing function of a capacitor. It's a bit more complicated for capacitors across motors but they take out any nasty spikes the turning motor may cause so they do not travel back to the controlling hardware.

The real electronics experts will have much more detailed and more accurate explantions of what is actually happening and how capacitors work.

 

[BeanieBots]

Electronic components also have mechanical equivalents as far as any analysis is concerned. If you can get your head around mechanical things that you hold in your hand, then it is also possible to do the same with electronic components.

A resistor is good old fashioned friction.
An inductor is equivalent to mass.
The less comonly know equivalent is that of the capacitor. It is actually equivalent to a spring! (stiffness inverse of capacitance)

Knowing that, it should be quite easy to see how something can be made to "vibrate" (turn on and off) with a capacitor and an inductor. The rules are the same. Big mass and it will vibrate slowly. Big inductor and it will oscillate slowly. Stiff spring will make it vibrate fast. Small capacitor will make it oscillate fast. Stick in some friction, the vibrations die down quickly. Stick in some resistance, the oscillations die down quickly.

Also, just like mechanical devices, electronic components have limits.
Pull a spring too far. It breaks. Put too much voltage on a capacitor, it breaks. Vibrate a spring too much, it explodes. Put too much current through a capacitor, it explodes.

There is nothing mechanical that cannot be simulated in electronics and (the one that many do not appreciate) there is nothing in electronics that cannot be replicated mechanically.

 

[Ralpht]

"There is nothing mechanical that cannot be simulated in electronics and (the one that many do not appreciate) there is nothing in electronics that cannot be replicated mechanically."

BB - I agree with your comparison between Mechanics and Electronica but -I'd agree with the first part of the above quote but the second half - I'd love to see someone make a plasma VDU; for example, out of mechanical component only !!

 

[BeanieBots]

A display device is probably the easiest. Pen & paper
I'm talking about behaviour. Not sensors and displays which are physical reactive devices.
Have you seen the wooden mirror? Quite an impressive mechanical display!
If not, google should be able to find you a video of it.

Actually, thinking about it, a plasma display IS mechanical.

 

[Rickharris]

Perhaps you are being a little too specific in your demands! A VDU is quite possible to replicate mechanically - indeed Logi baird did just that with his telivisor.

Here http://www.tep.org.uk/Frames/_f_trweb.html if th link works is a modern "toy" version using an LED as the illuminator and electronics to sync the picture - J-L's version was electro mechanical.

 

[BeanieBots]

Good point Rick. And the electro bit (spinning disk) could be water wheel driven if required.

 

[hippy]

I hate frames which give all pages the same url. For the televisor ...

http://www.mutr.co.uk/catalog/product_info.php?products_id=1376

 

[Ralpht]

"there is nothing in electronics that cannot be replicated mechanically ".

Even that televisor on your link Rick, had some electronics attached. Look at the PC boarfd behind the televisor in the picture.

I did say 'mechanical only'. Are you telling me that that original mechanical televisor can display, as an example the 1024 X 768 resolution of a very ordinary display in the literal unlimited colours etc that are available these days?

I conceede the point that a display was made mechanically back in the old days. But the effort was not viable so in reality a mechanical device could not do the job of an electronic device. There fore could not replicate an electronic device.

How do you define replicate ? Simpliy put - "to make a copy or duplicate of something."

OK then, replicate a core 2 duo processor mechanically please. Mine runs at 3Ghz.

Recon a mechanical device can be made to do the same function (replicated) at those speeds. A crystal can mechanically vibrate at 3Ghz but I'm talking about the whole function of a bog common electronics device.

Sorry guys, I tend to take things literally when a statement is made that:
"there is nothing in electronics that cannot be replicated mechanically ".

If the word 'simulate' was used instead, then I would agree.

 

[Ralpht]

After reading my post above, I may be a reasonable engineer but a p@!s poor typist.

 

[Michael 2727]

Years ago they had Petrol Pump digital displays that
were a bank of digit segments operated by relays which
flipped the correct segment to show a complete digit.

The same thing can be seen on older airport arrival
and departure information screens, I suppose they
have all gone to LCD or Plasma these days.

 

[andrew_qld]

I used to teach amateur radio classes years ago and I like the bucket analogy. I have seen people teach electronics using just plumbing- resistors thin pipes, current water speed, volts "header tank" pressure, diodes one way valves etc.

I still like drawing a mental picture of lots of electrons in a race to get onto to a capacitor plate before it gets all charged up. I have even been known to draw stick figure electrons with little smiley faces.

 

[jodicalhon]

This may help, rigidigital:
http://www.ibiblio.org/obp/electricCircuits/
Lots of good information here.

 

[BeanieBots]

"there is nothing in electronics that cannot be replicated mechanically ".[/I]

If the word 'simulate' was used instead, then I would agree.

My fault. I should have used EMULATE function.
Also, the devices you are using in your example are those which by their very nature cross the boundry between the two worlds and consequently require BOTH mechanics AND electronics.

To put it more simplistically. Anything mechanical can be represented by MLT (mass length time) and anything electronic can be represented by CLR (capacitance inductance resistance). MLT and CLR can be mathematically linked for equivalence. Hence, one can FULLY EMULATE the other.

Let's consider your other example, a 3GHz processor. Granted, it might not appear that it is possible to 'replicate' that mechanically. However, first off, it CAN (in theory) be done even if we don't yet have the technology. Secondly, don't get hung up on component for component equivalence. What is the processor actually doing? Running software? Maybe a complex calculus algoryhtm, which in turn is a simulation of some mechanical system. And round we go again.

Bottom line, IO devices don't count becaue they are translating the 'equivalance' values between the two worlds. eg the vibrations in a car travelling down a road exist whether you measure them or not. For us to "see" them, you need to 'translate' them into 'our' world using perhaps a G-meter. In the electronic version of a car travelling down an electronic road, the voltage (or current) fluctuations that represent G still exist, but again, to "see" them, we need to use a DVM (or similar) to translate them into 'our' world.

 

[Rickharris]

"there is nothing in electronics that cannot be replicated mechanically ".

Even that televisor on your link Rick, had some electronics attached. Look at the PC boarfd behind the televisor in the picture.

As I pointed out the toy model uses modern electronics for synchronisations but Logi baird didn't and electro mechanical means can be just as good.

I did say 'mechanical only'. Are you telling me that that original mechanical televisor can display, as an example the 1024 X 768 resolution of a very ordinary display in the literal unlimited colours etc that are available these days?

I guess the answer is yes - The resolution is set by the televisor disc - we could easily today micro engineer a fine enough grid to give the resolution - Indeed if not for the fact the VDU was developed at about the same time TV could well have been mechanical today. Colour can also be done mechanically.

I conceede the point that a display was made mechanically back in the old days. But the effort was not viable so in reality a mechanical device could not do the job of an electronic device. There fore could not replicate an electronic device.

How do you define replicate ? Simpliy put - "to make a copy or duplicate of something."

OK then, replicate a core 2 duo processor mechanically please. Mine runs at 3Ghz.

Recon a mechanical device can be made to do the same function (replicated) at those speeds. A crystal can mechanically vibrate at 3Ghz but I'm talking about the whole function of a bog common electronics device.

Sorry guys, I tend to take things literally when a statement is made that:
"there is nothing in electronics that cannot be replicated mechanically ".

If the word 'simulate' was used instead, then I would agree.

I think we are saying replicate the functionality - I originally worked on tracking radars that were using mechanical computing systems to solve complex movement equations totally mechanically in real time.

Babbage's differencing machine resolved complex mathematics totally reliably in real time - can't get a lot faster than "real time operation".

It.s fairly easy to make a mechanical resonator to operate at high frequency - a dog whistle for example - trouble we don't need to or nano technology would be a lot more advanced!

 

[moxhamj]

A capacitor is a bit like a tiny battery. It does not conduct as such. What it does is charges up (so is drawing current) and once it is charged up it stops drawing current because the volts are now equal.

The simplest capacitor is two metal plates seperated by air or something else (a dielectric). Or you could get a glass jar, put some alfoil on the inside and some on the outside. The glass is the dielectric. Connect one plate to +V and one to -V and for a brief moment current will flow as the capacitor charges up. Disconnect the battery and measure the volts and the volts will stay there. If the dielectric of air or glass or plastic is replaced with other materials (like an electrolyte in electroytic capacitors) the capacitor can be made much smaller.

The water bucket analogy isn't quite right as there isn't really an electrical equivalent of an overflowing bucket. A bucket might have a static pressure of 1/2 a psi at the bottom and we are filling it from a 100psi hose. Maybe that is like charging a 5V capacitor from the mains and maybe overflowing is the same as the capacitor exploding. Bottom line is don't exceed the voltage rating on a capacitor. But if you are under the voltage rating essentially a capacitor charges up to however many volts you put on it and then current stops flowing and it keeps that charge if things are disconnected. Sometimes the charge will stay for hours or even days.

Perhaps the water analogy for a capacitor is a long pipe going down a hill with the bottom blocked off. Run water into the pipe from a tank and it will fill up (volts increase) until it comes to the same level as the tank. Current is flow. Volts are water pressure. Resistance is the narrowness of the pipe. Inductance is the momentum of the water flowing (water hammer when a tap is suddenly turned off). A transistor doesn't really have an equivalent - maybe a sprinkler solenoid but not the electrical solenoid - rather the way a little hole is opened which then uses its own water pressure to open a much bigger hole.

Then again, capacitors can get complicated. I'm trying to get my head around reactance where the capacitor turns into a resistor at a certain frequency. That sort of makes sense as it resonates, but why in a tank circuit does the ratio of a C and a L determine the Q? Capacitors are more complex than they first appear.

Just follow the recipe and put them in where the circuit says. You don't need to understand egg protein binding to bake a cake!

 

[Rickharris]

My instructor in the RAF once said "Electronics is easy - there are only 4 things to worry about 5 if you include the wires joining the bits together: Resistance - Capacitance - inductance and the active device (valve, transistor, IC or whatever)."

How right, BUT the interesting bit is what happens when you put the 4 bits together in different ways.

Much of the practicalities of electronics can be understood if we remember two facts:

1. Electrons are negatively charged. (lets assume an electron is a particle)
2. Like charges repel and unlike charges attract (just like magnets)

From this we can see that electrons are repelled from the negative pole of the PSU and attracted to the positive. You need a complete circuit or they will stop flowing when the conductor gets full.

Conductors have a lot of free electrons able to move about - Insulator do not have many free electrons so little movement takes place. IF you push the electrons hard enough they will go through even an insulator - about 10,000 volts will make electrons jump an inch or so - if the insulator is air we call this jumping a spark.




OK the basics of capacitors has been well described and for many functions the charge and discharge function are the ones most used e.g. in timing circuits. the rational here is very simple - The capacitor takes time to charge or discharge, if you restrict the flow of electrons with a resistor then the charge/discharge takes longer.

The relationship of the capacitor and resistor is expressed as C x R= charge time in seconds - often called CR time - Cr time is in seconds when capacitance is in farads and resistance in ohms.

This is slightly confused as the capacitor charging rate slows down as it gets full, there is less room for the electrons to go into (perhaps) the charge curve is exponential so we,by convention, say a capacitor is fully charged when the voltage reaches 60% of the supply.

As for the PSU functions generally the capacitor acts as a reserve tank to fill in any slight drops in the supply - if the supply is not perfect DC this tends to smooth out the ripples in the almost DC current.

Capacitors can also protect against fast changes in current flow (spikes such as from a set of motor brushes) as far as I know there are 2 ways to look at this either the capacitor can charge very rapidly for a very short time thus absorbing the spike.

Or, alternatively the capacitor isn't able to charge instantly and so a sharp change in voltage on one side initially results in a correspondingly sharp change at the other i.e. for an instant the capacitor looks like a short circuit to this rapid change thus the spike at the motor brushes is shorted out and will not get to the microprocessor.

For this reason putting these capacitors as close to the source of the interference is critical or alternatively putting them as close as possible to the microprocessor input prevents interference picked up by the wireing affecting the processor.

 

[BeanieBots]

If you just want an anlogy to "get your head around" for capacitors, then think about a compressed air line.
Consider a narrow air hose.
When no air flows, the pressure will be the same at both ends.
If you allow air out, the pressure at the open end will almost instantly drop.
If you now fit a resevoir tank to the open end and close the open end, the pressure in the tank will increase until it is the same as the pressure of the far end of the pipe. Just like a capacitor charging.
If you now allow air to escape close to the reservoir tank, the pressure will only gradually drop as the tank empties. However, the volume of air comming out will be greater than the amount the line could supply.
Close the whole and the pressure will start to build again.
That is the same principal as a decoupling capacitor.

 

[D n T]

What is a capacitor, Hippy, that answer is the best on I have heard

Hippy, that parrallel is the best on for a capacitor I have heard or read, i think might have to use it in my classes. Very simple. Very on the money. Thanks

 

[premelec]

What distinguishes a capacitor from a battery [electrochemical] is the voltage characteristic on charge and discharge... batteries have a small voltage drop while discharging most of their capacity but capacitors have linear voltage increase or decrease with a steady current draw or charge...

A very useful relationship for capacitors is Q [charge] = CV = IT . C=farads, V=volts, I=amperes, T=time[seconds].

Real world capacitors are often not constant with regard to voltage and temperature etc... often times this can be ignored with good result but it's good to know these effects exist when troubleshooting timing circuits incorporating capacitors - for instance... some capacitors have a 'bounce back' or soakage effect where voltage reappears after they have been shorted out and discharged - slow electrons I guess!

 

[Rickharris]

when I was a young sprog learning the basics I recall being told a farad was a enormous value and a capacitor with a 1 farad capacity would be as big as a house.

As I sit here I have a 10 farad capacitor slowly discharging into an LED which it will keep alight for perhaps the next 4 or 5 hours. So much for training!

 

[evanh]

One thing I like to say to those that ask is, Capacitors resist change in voltage while Inductors resist change in current.

When one applies the above statement to Ohms law - If the conditions change around a capacitor then it will react with a change in current to maintain it's voltage. And if the conditions change around an inductor then it will react with a change in voltage to maintain it's current.

Compare that to a Resistor where both current and voltage evenly change with the changing condition.


Evan

PS: That's the DC point of view. The AC p.o.v is out of my depth and deals with frequencies, phasing and attenuation.

 

[BeanieBots]

Good analogy evanh.
Clearly explains why when an inductor (such as a relay coil) is turned off, the voltage shoots up in an attempt to keep the current flowing. Even a small inductor can produce many thousands of volts trying to keep the current going which is why catch diodes are so important to protect fragile transistors from massive voltages.

The AC theory of Ls and Cs is still only ohms law. V=IR
It's just that V, I and R need to expressed fully to take into account the time domain and the sums become rather horrible.
Think of two cars, one going at 10mph the other at 5mph.
It's very easy to calculate how long it will be before they collide if they are driving towards each other and you know the distance. It gets a little more complex if they are at an angle to each other but all you need to do is take into account their velocity rather than the simple speed.