Capacitor discharge

[prasannan82]

Hello ,
I want to discharge a capacitor to ablolutely zero value.As per the capacitor theory it is not possible to either charge a capacitor to the input voltage value or to discharge it fully from the charged value.
Ex if i charge a cap using a current source to 10v and discharge ,some small voltage will be prresent say 0.1v.If i charge the cap again it charges to 10.1v and after discharge it will be 0.2v.If i continue this cycle the capacitor voltage will increase and bursts.
SO how will i discharge my cap to zero value.
Kindly sugegst.

 

[Nigel Goodwin]

Solder a piece of wire across it and leave it for a few years!.

The capacitor won't charge to any higher than the supply voltage, so your concerns are imaginary!.

 

[ericgibbs]

If it really mattered, you could apply a voltage of equal potential but of opposite polarity to the capacitor.

 

[prasannan82]

Capacitor charging

But I am charging a capacitor using a current source for a specified amount of time and discharging it for some time.This cycle repeats.
As it is the current source the same current flows in each cycle and same voltage is added to the capacitor.

 

[ericgibbs]

Hi,
Can you explain the purpose of the charge/discharge cycle, why are you doing it?

Im just curious.

 

[prasannan82]

Capacitor charging

In one of my application i am charging a capacitor using a current source for some time x sec and discharging the capacitor for y sec.But some voltage is present after the discharge also.
When this cycle repeats the charge accumulates in the cap and voltage increases in each cycle and the cap burns.
I want to discharge the cap fully to ablolute zero value.

 

[Nigel Goodwin]

In one of my application i am charging a capacitor using a current source for some time x sec and discharging the capacitor for y sec.But some voltage is present after the discharge also.
When this cycle repeats the charge accumulates in the cap and voltage increases in each cycle and the cap burns.
I want to discharge the cap fully to ablolute zero value.

Didn't you read my previous post?, the capacitor can't charge higher than the source voltage - your application sounds like a simple oscillator?, which works fine.

 

[prasannan82]

But the source is not the voltage.Its the current.I am giving same current in each on period .As the charge accumulates the voltage will build up.Right?

 

[Nigel Goodwin]

But the source is not the voltage.Its the current.I am giving same current in each on period .As the charge accumulates the voltage will build up.Right?

NO!!!! - the voltage can't rise higher than the supply voltage - if it was a true constant current it could, but there's no such thing, and the current is only constant within it's supply voltage.

 

[G8RPI]

Capacitor voltage

Nigel, a capacitor will charge to more than the supply voltage if the is series inductance! Many pulse discharge systems (flashlamps etc) use this technique. A series inductor limits the current drawn from the supply during charging and causes the capacitor to charge to almost twice the supply. As the supply is typically rectified mains, the rectifiers stop the voltage being clamped by the source.
I guess Prasannan82 has a capacitor rated at less than the supply used for his constant current source. His problem with residual charge is caused by dielectric absorption. The best capacitors for this application were polycarbonate, but they are no longer made (I think the plastic film was single source and the supplier stopped making it). The next best options are Teflon (PTFE very rare for capacitors), polystyrene and polypropylene.
Any type of electrolytic will be very poor, even if you do leave it shorted for a week!

Robert.

 

[Nigel Goodwin]

Nigel, a capacitor will charge to more than the supply voltage if the is series inductance! Many pulse discharge systems (flashlamps etc) use this technique. A series inductor limits the current drawn from the supply during charging and causes the capacitor to charge to almost twice the supply.

There's never been any suggestion that it's any kind of switch-mode converter, so it's silly to assume so - particularly when what he has told us is that it charges and discharges over a period of time, fed from a 'constant' current source.

As the supply is typically rectified mains, the rectifiers stop the voltage being clamped by the source.
I guess Prasannan82 has a capacitor rated at less than the supply used for his constant current source. His problem with residual charge is caused by dielectric absorption. The best capacitors for this application were polycarbonate, but they are no longer made (I think the plastic film was single source and the supplier stopped making it). The next best options are Teflon (PTFE very rare for capacitors), polystyrene and polypropylene.
Any type of electrolytic will be very poor, even if you do leave it shorted for a week!

You appear to have imagined a completely different scenario from what the meagre information supplied suggests?. As with almost all posts on here, he would get far more helpful answers if he actually gave some information!.

 

[Sceadwian]

Nigel, you gave him assurances his concerns were imaginary without any knowledge of the circuit he was working on, apparently G8RPI isn't the only one reading too much into the scenario. A simple "Your question can not be answered without a detailed schematic of the circuit you're working on" reply was probably the best to the original post.

Perhaps whatever circuit he's working on is setup in such a way that it would cause an uncontrolled boost conversion and blow the cap like he feared?

Probably not, but no one is right or wrong or reading too much into anything unless a specific answer is given without enough information. G8 got to it before me because I knew it was going to lead to an argument but a switched capacitor array or a relatively large inductance/capacitance together could easily create voltages WELL over supply. Either that or the CCFL inverter and flasher circuit I have is powered by miniature gerbils => And it's a VERY simple circuit for someone to be asking questions on. I think the last thing to add is..
PRASANNAN82 YOU NEED TO POST A COMPLETE SCHEMATIC FOR THE CIRCUIT YOU'RE ASKING A QUESTION ABOUT OR YOUR QUESTION CAN NOT BE ANSWERED

 

[Hero999]

the capacitor theory it is not possible to either charge a capacitor to the input voltage value or to discharge it fully from the charged value.

That's nonsense, no where in capacitor theory does it say that you can't fully charge or discharge a capacitor. It just says it is impossible with a series resistor.

A capacitor will charge linearly if given a constant current source and discharge linearly if given a constant current load.

A 1F capacitor charged to 1V will discharge through a 1A constant current load in 1s.

The formula is:
dv/dt = CI

 

[Sceadwian]

Constant current loads only exist in theory Hero, at least ones of sufficiantly low series resistance to truly completly discharge a capacitor. Then again you're talking about nano volts in a practical circuit if even that. Thermal noise is probably higher than the voltage left over in a real world capacitor when it's discharged through a low ohm load.

 

[Nigel Goodwin]

Nigel, you gave him assurances his concerns were imaginary without any knowledge of the circuit he was working on, apparently G8RPI isn't the only one reading too much into the scenario. A simple "Your question can not be answered without a detailed schematic of the circuit you're working on" reply was probably the best to the original post.

Well I do tend to get a little tired of typing that, as almost every poster doesn't give enough information

Perhaps whatever circuit he's working on is setup in such a way that it would cause an uncontrolled boost conversion and blow the cap like he feared?

Perhaps?, but the little information he DID give made no such suggestion, and didn't give any reason at all to suspect any such circuit - in fact what it did tell you pretty well eliminated any such type of circuit.

I am charging a capacitor using a current source for a specified amount of time and discharging it for some time.

This can hardly apply to a switch mode or flyback circuit?.

 

[Ambient]

It sounds like your current source can provide more voltage than the cap is capable of taking in. In a constant current source, won't the voltage increase automatically to try and maintain the current? If thats the case you need a higher voltage rated capacitor or a current source that is capable of lower voltages.

Am I close or way off?

At any rate, try decreasing the charging duty cycle and leave the frequency the same. This way the cap has more time to discharge. Or just increase the time of the discharge and leave the charge time the same. Or increase the discharge current/ or decrease load resistance.

 

[Sceadwian]

Switch mode supply topology is diverse. A very low voltage moderate current boost converter could bootstrap itself to a higher voltage for operation. Any configuration which allows switched capacitor or inductive elements has the ability to create higher than VCC voltages. Charge 5 capacitors in parallel and discharge them in series.. Inductors are capable of even higher boost voltages based on inductance, as the CCFL and flash inverters I mentioned are capable of easily.

But we are digressing SEVERLY off the original topic which we must likley wait some time to receive an answer to if we ever do. Anything up until that point is speculation.

 

[Nigel Goodwin]

Switch mode supply topology is diverse. A very low voltage moderate current boost converter could bootstrap itself to a higher voltage for operation. Any configuration which allows switched capacitor or inductive elements has the ability to create higher than VCC voltages. Charge 5 capacitors in parallel and discharge them in series.. Inductors are capable of even higher boost voltages based on inductance, as the CCFL and flash inverters I mentioned are capable of easily.

But none of those bear any resemblance at all to the description we did get!.

 

[Sceadwian]

That's because we got NO description of a circuit really. Only a vague indication of charging and discharging a capacitor through a current source. I'm not sure if the OP really gets that the voltage imballance in a modern circuit with a capacitor connected to it tends to be bellow the noise floor, useable voltage is almost alway used.

But you still can't say that you can never charge a cap past VCC and just rest on your laurels with no specific circuit as an example =) Dangerous!

 

[Nigel Goodwin]

But you still can't say that you can never charge a cap past VCC and just rest on your laurels with no specific circuit as an example =) Dangerous!

In the existing context you can, using a constant current source to charge a capacitor you can't charge it higher than the voltage source the constant current source is fed from.