AC dilemn

hi guys,
this might be stupid but i was wandering why cant we store ac as dc in batteries..it should not necessarily be batteries but in like a capacitor bank..even if it cant be stored for a long time ,isnt there a way to store it for say a minute or two...
thks

It's kind of because if you input AC into a battery or capacitor or something, the second half the cycle empties out the charge that the first half of the cycle put into the storage device since it they are of opposite polarity.

It's not so much storing AC or DC as storing energy. Using something like a flywheel would probably work to store AC since the spinning can be hooked up to something else to make the sinusoid waveform...or you can just store DC and use an inverter to turn it back into AC again (that's what motor drivers do and power converters).

WHat are you trying to do exactly? Power an AC device from stored energy?

You can store AC power, but not for very long, if you connect a parallel resonant capacitor and inductor an AC power supply then disconnect it, the AC voltage will be stored for a short time before dieing away.

If you think about DC like water flowing in a hose, and AC like a vibration, you should begin to see the analogy - you can pump water into a water tower to store it, and later you can get a lot of the energy back by releasing it. How would you store a vibration? Something along the lines of a tuning fork would be the idea - this is the analog of a resonant inductor/capacitor circuit (AKA "tank circuit") - but this isn't really feasible to implement for effective energy storage. In the tuning fork, friction from the air and the material it's made from prevent it from storing vibrational energy for very long - it's even worse in the electrical case.

Hero999 said:

You can store AC power, but not for very long, if you connect a parallel resonant capacitor and inductor an AC power supply then disconnect it, the AC voltage will be stored for a short time before dieing away.

Click to expand...

To add to what Hero said, the reason that it will only be stored for a short time is that the inductor and capacitor have resistance, because they're not ideal, which causes energy loss. Generally it's far too significant to possibly store energy for even a second, let alone a few minutes.

I believe I once read an article about using superconductors to make a resonant tank circuit which could store energy much more efficiently - though you had to keep it immersed in liquid nitrogen or something to make it superconducting...

It's easier to convert the AC to DC, store that and then use an inverter to create AC. Direct storage of AC is pretty useless for anything more than a few cycles.

WHat are you trying to do exactly? Power an AC device from stored energy?[/QUOTE]
Not exactly but like the UPS converts ac into dc then store it into batteries ,i was thinking of eliminating the bulky batteries and find some way to store ac for some time(the least required to safely turn off the pc).i was thinking of converting the ac supply frequency into a very low frequency say 1 Hz then store it into resonant circuit and then reconvert it into 50Hz ac when needed.

I thought PCs use supercaps on the mother board after the AC-DC converter to store DC in order to allow the processor and HD a few seconds to finish writing all data to the harddrive. Maybe this was only some computers.

If you want to risk it...you can do the same thing stick a few supercaps or something in parallel with the outputs of the powersupply...but be careful...and investigate it first. Make sure votlage ratings and capacity are enough and that the powersupply can handle the initial inrush current.

First off, how would converting from 50Hz to 1Hz and back be easier than converting DC to AC and back?

Second, why would storing energy at 1Hz be easier than storing it at 50Hz?

Third, do you think that the storage elements (tank circuit) would actually be SMALLER than batteries with comparable capacity?? Just as a quick reality check, in a resonant circuit, the energy goes back and forth between voltage on the capacitor, and current through the inductor - so at one point in the cycle, all the energy is contained in the capacitor. Go look at the energy capacity of a big capacitor, and then compare it to the capacity of a similar-sized (physically) battery. Add in the huge inductor that's going to be the other half of the resonant circuit, and the battery is going to very easily win for capacity vs. size - plus, again you have the issues of the significant losses in the resonant circuit making energy storage impractical for the periods of time you're looking for, even if you didn't care how big and expensive your storage elements were.

i supose you could charge one battery with the positive half of the cycle and another with the negative half. then just use transistors in a push pull configuration and output a sine wave with a signal generator. and as far as computer ups's are concerned. wouldnt it be simpler to just store the 12, 5 and what is it 3.3 volts seperatly and just feed the voltage dirctly to the computer during a power outage?

Gaston said:

i supose you could charge one battery with the positive half of the cycle and another with the negative half. then just use transistors in a push pull configuration and output a sine wave with a signal generator. and as far as computer ups's are concerned. wouldnt it be simpler to just store the 12, 5 and what is it 3.3 volts seperatly and just feed the voltage dirctly to the computer during a power outage?

Click to expand...

but, how is that an improvement over using a full-wave rectifier and storing BOTH halves of the cycle in ONE battery?
The second part of what you said is exactly the job that a regular, cheap, commercial inverter will handle for you.
Storing 12v, 5v, 3.3v, -5v, and -12v in separate batteries would also not be a great plan. It would make more sense to store a single voltage and run a DC-DC converter to power the computer from there full-time.

But, all of these ideas are exactly what you can already achieve by just buying a (usually inexpensive) commercial UPS.

i'm not saying its better, i was just talking hypotheticaly that it could be done that way. as for the ups. the only part that doesn't make sense to me is to convert from ac to dc. then from dc to ac, then fron ac to dc again. just seems like it should be from ac to dc one time

Gaston said:

i'm not saying its better, i was just talking hypotheticaly that it could be done that way. as for the ups. the only part that doesn't make sense to me is to convert from ac to dc. then from dc to ac, then fron ac to dc again. just seems like it should be from ac to dc one time

Click to expand...

I see what you're saying, and that might seem more efficient, but unless you make some major modifications to the computer power supply, it is expecting its input in AC. It's far cheaper just to have the AC-DC to charge the battery, then the DC-AC to go to the computer power supply.

what i am sugesting with the ups is. to convert it to dc one time and have the power supply able to run the computer and charge the battery or batteries at the same time basicaly in parallel with the devices in the computer. when the ac power goes off there will be power from the battery ready to go.

Gaston said:

what i am sugesting with the ups is. to convert it to dc one time and have the power supply able to run the computer and charge the battery or batteries at the same time basicaly in parallel with the devices in the computer. when the ac power goes off there will be power from the battery ready to go.

Click to expand...

as I said, you could do this quite easily by just using an ATX DC-DC power supply, which is just like a normal PC power supply but can run with 12V DC input. These power supplies are quite common in the carPC world.

However, because they're not produced in such huge quantities as inverters and AC ATX power supplies, it's going to end up costing more than inverter+AC ATX PSU. It should be a lot more efficient though - many DC-DC converters manage efficiencies of 90% or more, whereas the efficiency of the combination of inverter+AC ATX would be much, much lower.

I didn't know they made a 12 volt atx power supply. that would make it easy.

Stupid question: why not using a 24 V battery plus a rectifier, and then use a couple of 7812s to have 0, 12V and 24 V levels. With these level you:
use the 12 V as the 0 V for your computer.
your 24 V level, when the reference is 12V, becomes 12V
use a 7805 connected to 24 V as Vdd (+) and 12V as reference and there you have your 5V
an LM 338 connected between 24V and 12 V (with appropriate resistances) to get 3,3V
Your 0V on the battery becomes your -12V.

If you feed all these directly to the motherboard and peripherics (bypassing the PSU) you have the most efficient, cheap and simple way of ensuring that your computer will have power even when the lights go out.

yeah, except that you can only get a few watts out of those kinds of linear regulators, and they're horrendously inefficient for dropout voltages that high (12v dropout from 24v is already 50% efficiency at best...). PC's generally require 100-200W or more. You do the math.

This is why switching regulators are used for high power loads - which is precisely what the DC-DC ATX power supplies I suggested do.

Also, you probably could not get away with the "virtual ground" scheme you have proposed - the positive 12v rail of a PC is going to carry much more current, and thus your 12v regulator acting as virtual ground would have to SINK current - which is not how these linear regulators operate.

And on top of all that, ATX supplies need to do more than just provide voltages. They also have to deal with some control lines that let the computer control the power. Not a big deal, but just one more reason why it would make no sense to reinvent the wheel in such a terribly inefficient way, when existing DC-DC supplies can do the job better, much more efficiently, and be smaller and cheaper.

evendude said:

I believe I once read an article about using superconductors to make a resonant tank circuit which could store energy much more efficiently - though you had to keep it immersed in liquid nitrogen or something to make it superconducting...

Click to expand...

In practice it wouldn't last for ever because there will be losses in the form of electromagnetic radiation.

Second, why would storing energy at 1Hz be easier than storing it at 50Hz?

Click to expand...

Well the losses will be lower at a lower frequency.

Anyway what's normally done when AC power needs to be stored is it's converted to DC (using a transformer and rectifier) then converted back to AC (using an inverter). Un fortunately the overall effenciency is quite low, it can be less than 60% once the losses in the battery, rectifiers and inverter are taken into account.