If I want to charge a capacitor off a bridge rectifier, should I bother with a ripple-reducing capacitor? It seems there would be advantages in terms of average DC voltage over time being higher, but the ripple-reducing cap would also therefore be in parallel with the charging cap while having a different capacitance/voltage rating to it.
TIA!

 

I did not know that ripple reducing cap and charging cap is two different things ?
I always choose the filter cap to suite the max current. Typically 1500uf/amp. Plus a 100nf bypass cap if needed.

 

I don't see a difference between the two unless you have some circuitry between the ripple-reducing capacitor and the charging capacitor.

 

I did not know that ripple reducing cap and charging cap is two different things ?
I always choose the filter cap to suite the max current. Typically 1500uf/amp. Plus a 100nf bypass cap if needed.

Well, I am thinking of having a large capacitor bank that will store a significant amount of energy off the DC feed from the bridge rectifier. These will obviously have different capacitive characteristics than your typical ripple cap, as per your example in your response.
It seems to me having a minimum ripple will (obviously) result in faster charging than two unmodified half waves coming off a bridge rectifier. If this is true, then a ripple cap with a very short charging time,once charged, will allow a basically flat and constant DC voltage to charge the caps with a much longer RC period to charge more efficiently over time?

 

I don't see a difference between the two unless you have some circuitry between the ripple-reducing capacitor and the charging capacitor.

Well, the ripple-reducing cap would have different capacitive values than a large , energy-storing cap bank placed after it.
Assuming it was a well-selected ripple cap that produced essentially a near-perfect DC output, would the subsequent energy-storing cap bank view this as an uncomplicated DC voltage, or would you have to incorporate the total capacitance equations for caps in series or parallel to include the ripple cap? OR, as I suspect would the rectifier "see" the entire capacitive load downline and the ripple effect would inadvertently be altered as a result?

 

Well, I am thinking of having a large capacitor bank that will store a significant amount of energy off the DC feed from the bridge rectifier. These will obviously have different capacitive characteristics than your typical ripple cap, as per your example in your response.
It seems to me having a minimum ripple will (obviously) result in faster charging than two unmodified half waves coming off a bridge rectifier. If this is true, then a ripple cap with a very short charging time,once charged, will allow a basically flat and constant DC voltage to charge the caps with a much longer RC period to charge more efficiently over time?

How much is "a significant amount of energy"?

Unless your large capacitor has some sort of unrealistically huge ESR value or the input AC frequency is very high the smaller capacitor in front of it will serve absolutely no purpose.

 

If I want to charge a capacitor off a bridge rectifier, should I bother with a ripple-reducing capacitor? It seems there would be advantages in terms of average DC voltage over time being higher

You can't get more than the peak AC voltage. If the capacitor is the only load, another cap in parallel would serve no purpose. If you leave the big cap connected long enough, it will charge to the peak voltage.

 

Sometimes we do put different value capacitors in parallel to filter a broader range of frequencies.
At AC mains frequency, the time constant is so long that a reservoir capacitor and a ripple capacitor are one and the same.

 

How much is "a significant amount of energy"?

Unless your large capacitor has some sort of unrealistically huge ESR value or the input AC frequency is very high the smaller capacitor in front of it will serve absolutely no purpose.

Aha, suspected so. So, do I take an efficiency hit in terms of capacitive storage over time because I don't have the on-average higher DC level charging the caps from reducing the ripple?
Eventually, once I stop taking metrics on this, I will switch to battery storage anyway.

 

You can't get more than the peak AC voltage. If the capacitor is the only load, another cap in parallel would serve no purpose. If you leave the big cap connected long enough, it will charge to the peak voltage.

Thanks. Looks like I take a hit in terms of how quickly it charges if I don't flatten the ripple, I guess.

 

Sometimes we do put different value capacitors in parallel to filter a broader range of frequencies.
At AC mains frequency, the time constant is so long that a reservoir capacitor and a ripple capacitor are one and the same.

Great answer, precisely what I was wondering. In this case, the input to the rectifier will consist of intermittent bursts of single AC pulses between 0.1-20 Ms wavelength separated by slightly longer intervals of zero input.

 

..........the input to the rectifier will consist of intermittent bursts of single AC pulses between 0.1-20 Ms wavelength separated by slightly longer intervals of zero input.

I assume you mean ms (milliseconds).
Time represents a period, not wavelength (given in distance).
What is the frequency of the AC?

 

Great answer, precisely what I was wondering. In this case, the input to the rectifier will consist of intermittent bursts of single AC pulses between 0.1-20 Ms wavelength separated by slightly longer intervals of zero input.

We use different capacitors in parallel because they have different ESR and inductances and hence have different frequency responses. This is applicable at frequencies above 1MHz.

If your frequencies are below 1MHz you will not benefit from having capacitances in parallel.

 

I assume you mean ms (milliseconds).
Time represents a period, not wavelength (given in distance).
What is the frequency of the AC?

I do mean ms, but my autocorrect thinks I am using the modernistic form of female salutation, sigh. And,yes, incorrect re wavelength, I meant pulse duration. Sorry re both, def imprecise on my part.
These are variable, intermittent single AC waveforms interspersed unevenly over time. I have a source that generates these in great abundance at no cost to me (not to be coy). Once rectified, they provide a substantial amount of DC power, which I want to store. In the prototype, I use caps so I can get an approximate feel for how much energy storage I can attain. Ultimately, storage would be by battery. My readings indicate that with modern-day battery storage (large cells) , no real harm is incurred by charging with with an uneven and intermittent pulse (which surprised me) as long as overheating isn't an issue. So, the caps serve as an intermittent test device, but the ripple question - which has been answered to my satisfaction in this thread - did arise.

 

Batteries are quite tolerant of ripple.
For example most vehicle battery chargers apply the voltage directly from the AC rectifiers to the battery with no capacitor smoothing at all.
The battery does the smoothing.
So you may not need much capacitance unless you need it to store energy between AC bursts.

 

We use different capacitors in parallel because they have different ESR and inductances and hence have different frequency responses. This is applicable at frequencies above 1MHz.

If your frequencies are below 1MHz you will not benefit from having capacitances in parallel.

These single AC pulses are far below 1 MHz. They vary in amplitude up to 100Vpp and are very slightly asymmetrical in the time domain (interesting, but not a consideration), and range between 0.1-20 ms in duration. They don't self-interfere, if you'll excuse the term.
Your comments regarding parallel caps are informative. My understanding re caps in series is that one benefits from an overall reduced RC constant while energy storage remains the same as with a parallel cap configuration comprising the same elements. However, series caps are vulnerable if one or more suffer performance degradation.

 

Batteries are quite tolerant of ripple.
For example most vehicle battery chargers apply the voltage directly from the AC rectifiers to the battery with no capacitor smoothing at all.
The battery does the smoothing.
So you may not need much capacitance unless you need it to store energy between AC bursts.

Thanks, great info!!

 

I have a source that generates these in great abundance at no cost to me (not to be coy). Once rectified, they provide a substantial amount of DC power, which I want to store.

Again without a definition of substantial no on here has any idea what size of capacitor bank Vs common storage battery equivalent you are working with.

Which. BTW in comparison any capacitor based storage system is dismally light on energy storage capacity for the size and financial costs involved compared to any basic battery. You would need a bank of ultracapacitors the size of a small vehicle worth $100,000+ to equal to same storage capacity of a common $100 12 volt deep cycle battery.

 

Again without a definition of substantial no on here has any idea what size of capacitor bank Vs common storage battery equivalent you are working with.

Which. BTW in comparison any capacitor based storage system is dismally light on energy storage capacity for the size and financial costs involved compared to any basic battery. You would need a bank of ultracapacitors the size of a small vehicle worth $100,000+ to equal to same storage capacity of a common $100 12 volt deep cycle battery.

All this I am quite aware of. Nevertheless, as an ad hoc indicator of RELATIVE energy storage capability in real time, measuring the DC voltage off a cap of known value is a neat, quick-and-dirty way of capturing info. Rest assured I also have oscopes , multimeters, etc.
I mentioned in other responses to the post that I am dealing with a large number of single pulse AC waveforms ranging up to 100Vpp and durations from approx 0.1-20 Ms. After rectification, I have successfully charged batteries using this source. It's just that I don't know how to calculate how much energy is stored in a rechargeable battery without draining it on a circuit of known resistance, impedance. That's time-consuming and inaccurate, since battery charges are non-linear over time.
Thus, the ripple question with regards to capacitor charging.
Everyone has been extremely helpful ;-)

 

All this I am quite aware of. Nevertheless, as an ad hoc indicator of RELATIVE energy storage capability in real time, measuring the DC voltage off a cap of known value is a neat, quick-and-dirty way of capturing info. Rest assured I also have oscopes , multimeters, etc.
I mentioned in other responses to the post that I am dealing with a large number of single pulse AC waveforms ranging up to 100Vpp and durations from approx 0.1-20 Ms. After rectification, I have successfully charged batteries using this source. It's just that I don't know how to calculate how much energy is stored in a rechargeable battery without draining it on a circuit of known resistance, impedance. That's time-consuming and inaccurate, since battery charges are non-linear over time.
Thus, the ripple question with regards to capacitor charging.
Everyone has been extremely helpful ;-)

Yes I follow, but if the average or peak currents of each pulse are known along with their voltage and duration it's not too difficult to get a rough estimation of the total available power that may be capturable.

Same with quantifying the relative battery size. Recharging a button cell Vs a sub C multi AH cell Vs a group 31 deep cycle LA battery are all pretty easy to define and be understood by most everyone here.