alternative way of charging 330V caps?

[Saibot]

Hi, currently I need to charge... 11x 160 uF capacitors (330V) with a COMPUTER power supply.

It outputs 3.3v, 5v, or 12volts all DC.

Unfortunately, they have an overcurrent protection limit of ~5-18 A.

This is how i am currently charging them:

I used a simplified representation of a photoflash circuit . The transformer in that diagram is the actual circuit for charging the camera caps.

Anyways, i attached four of these transformers to my primitive voltage divider (to input 1.5V, as a normal battery would) in an attempt to charge faster. Unfortunately... it takes about 10 minutes for all my caps to be fully charged, while my resistors get extremely hot.

Are there any other ways of somehow getting ~300V out of a computer power supply, that is MORE efficient?

 

[dknguyen]

Why are you dividing a voltage rather than feeding it directly? You want to step it up...why step it down first?

Why are you using a transformer to step-up DC? Transformers don't work with DC and capacitors do not charge up to a DC voltage when fed AC- they stay at zero average voltage.

You need a boost converter. Or a charge pump. Both circuits could be suppliued power by a PC power supply...or any other power supply for that matter.

 

[HarveyH42]

Why are you dividing a voltage rather than feeding it directly? You want to step it up...why step it down first?

Why are you using a transformer to step-up DC? Transformers don't work with DC and capacitors do not charge up to a DC voltage when fed AC- they stay at zero average voltage.

You need a boost converter. Or a charge pump. Both circuits could be suppliued power by a PC power supply...or any other power supply for that matter.

The disposable camera flash circuit fries at about 2.5 volts (normally uses a single AA battery).

Saw some place (instructables I think), where the output of several flash boards were wired in parallel to charge the capacitors.

I've modified some as strobes, and had 4 of them wired in series, so I could use a 6 volt SLA to power them. Made a pretty good lightning effect, as each flashed at a different rate.

Why not just build something to charge the capacitors. Find the right transformer, couple of diodes... Plug it in the wall.

 

[Saibot]

Why are you dividing a voltage rather than feeding it directly? You want to step it up...why step it down first?

Why are you using a transformer to step-up DC? Transformers don't work with DC and capacitors do not charge up to a DC voltage when fed AC- they stay at zero average voltage.

You need a boost converter. Or a charge pump. Both circuits could be suppliued power by a PC power supply...or any other power supply for that matter.

i dont exactly know how to make a circuit diagram of a photo flash circuit... so i represented it with a transformer :X

How.. exactly would i build a not too complex charge pump or boost converter? I'm on a very very low budget as i have already spent alot of my money on these stupid flash boards and caps.

 

[HarveyH42]

i dont exactly know how to make a circuit diagram of a photo flash circuit... so i represented it with a transformer :X

How.. exactly would i build a not too complex charge pump or boost converter? I'm on a very very low budget as i have already spent alot of my money on these stupid flash boards and caps.

You can get a schematic off the web for 'Kodak Max', maybe others, but seem about the same.

Can't believe you bought the flash boards... I got 17 free from the second store I asked, and still have quite a few I haven't even cracked open (yeah, a little work removing the case).

Mine draw about 700 mA when they start to charge. Anyway, try this link;

**broken link removed**

His boards are 3 volts, mine would fry. Put two boards in series, and the outputs (to the capacitor) in parallel, you get twice the current, and waste nothing on resistors and heat. You can pair up as many as you want, and hook them up to the 3.3 volts of your PSU.

 

[Blueteeth]

Charge pumps are very low current, and boost converter design is really tough for 12V - 330v. (those that use a single inductor anyway). This one struggles for just 180v:

https://www.elecfree.com/electronic...ost-converter-using-the-555-as-controller.gif

If 'budget' is your main concern rather than efficiency, then I would suggest a mains transformer, wired backwards. The real trouble with that is that transformers like that are designed for relatively low current on the secondary, limited how much current you can put into it, and therefore, limiting your charging current (therefore, charging time). Not only that but they're designed for 50/60Hz...

Most camera flash transformers have a turns ratio of 1:12, using the primary as a boost converter, so the primary voltage is about 25V. Which is then stepped up by the transformer to 330 ish. The ones in small cameras are for low current (one or two AA powered) so you'll need a lot to get any half decent charge times.

So I guess winding your own transformers is out. And using any special IC's for the job isn't too useful either (linear techology have some camera charging IC's).

I guess the best way would be the mains transformer, wired backwards, with a 555 timer circuit driving a power transistor/mosfet to chop DC for the transformer. There are some clever tricks you can do with the 555 to increase efficiency, like using its as a pseudo switched mode power supply, which can boost the voltage on the primary winding, increasing the voltage on the secondary. You can get up to a 1000v this way using standard household transformers. Alas they are very bulky, and as mentioned, not ideal for this app.

Blueteeth

Edit Obligatory warning! 11 160uF caps in parallel, charged to 330v can provide up to 96 Joules of energy storage. If you discharge them in 10ms, thats 9600Watts, or 9.6kW. At 330V, enough current can flow through your body to kill you. more likely it will just burn, and throw your arm away. But sitll, exercise caution when dealing with high voltage cap storage. High voltage DC is bad because its continuous, but cap storage can provide a hell of a lot more power, in a shorter time.

 

[Hero999]

I don't think it's possible to do as Harvey said because they are active converters so they probably won't like being connected in series and the 0V side of the output is often connected to the 0V side of the input.

It's dangerous but not as dangerous as 230VAC mains because it's DC and the current won't flow for as long.

 

[dknguyen]

I would think the easiest way is to step-up mains AC and then rectify it since you are working from wall power anyways. But that's kind of dangerous depending on your level of experience.

 

[Ubergeek63]

How much were you paying for flash boards?

At any rate, a simple inductor will not work. You need to transfer the required energy to the capacitor through the inductor.

J=½IL and J=½VC,
J=11*160µF*330V/2=0.58J
Shall we say 100µH at 1A peak?
J=100µH*1A/2=50µJ
So you would need 10,000 pulses to charge the caps in a perfect world. In the real world you will have to compensate for circuit losses, capacitor leakage, and the time that it takes to get the inductor to 1A at your supply voltage.

In most cases a simple inductor will not work, it is impractical to use a simple boost inductor to go more than 10X voltage gain. It is for that reason that you find transformers charging the photoflash capacitors.

 

[Saibot]

You can get a schematic off the web for 'Kodak Max', maybe others, but seem about the same.

Can't believe you bought the flash boards... I got 17 free from the second store I asked, and still have quite a few I haven't even cracked open (yeah, a little work removing the case).

Mine draw about 700 mA when they start to charge. Anyway, try this link;

**broken link removed**

His boards are 3 volts, mine would fry. Put two boards in series, and the outputs (to the capacitor) in parallel, you get twice the current, and waste nothing on resistors and heat. You can pair up as many as you want, and hook them up to the 3.3 volts of your PSU.

i did exactly what he did... just with more caps and without putting the boards in series (actually, thats a good idea. i'm gona try that)

The boards i bought For $1.50 seem to fry at 3.3 volts. They just like.. die and never work again. thats why i divided the voltage from the PSU so that the input is 1.5V.

it just takes wayyy to long to charge..

edit: i see a problem with wiring two boards in paralelle. the boards seem to put out 200* the voltage that you put in. If i put two in paralelle on the 3.3V, each board is getting 1.65V... that means the output is 330V exactly... which means its at the limit of my caps.. which is... 330V.

Should i be concerned about this?

 

[Hero999]

If you're worried about that then stick a Schottky diode in series to drop the input voltage to 3V.

 

[Ubergeek63]

I would think the easiest way is to step-up mains AC and then rectify it since you are working from wall power anyways. But that's kind of dangerous depending on your level of experience.

Very true...I went back and reviewed the OP and wondered why he is doing it that way.

And a half a joule is a lot of energy to have ready to abuse!

 

[Saibot]

Very true...I went back and reviewed the OP and wondered why he is doing it that way.

And a half a joule is a lot of energy to have ready to abuse!

for this particular project, i'm not allowed to create anything home made to run off wall power, as it is considered dangerous. Its either i use batteries, or tap into wall power with something premade.

 

[Ubergeek63]

how fast do you need the caps to charge?

 

[Saibot]

hopefully under 2 minutes.

 

[Ubergeek63]

well from the standpoint of readily available parts I would say to experiment with small power transformers and see what you could get out of it at 10KHz supplied from 24V supply into a 5V winding.

It is hard to say exactly since they are speced as power transformers they get away without really telling you much about them.

 

[Blueteeth]

The boards i bought For $1.50 seem to fry at 3.3 volts. They just like.. die and never work again. thats why i divided the voltage from the PSU so that the input is 1.5V.

it just takes wayyy to long to charge..

edit: i see a problem with wiring two boards in paralelle. the boards seem to put out 200* the voltage that you put in. If i put two in paralelle on the 3.3V, each board is getting 1.65V... that means the output is 330V exactly... which means its at the limit of my caps.. which is... 330V.

Should i be concerned about this?

If the cameras they are from are 3v, then you'll want to give them no more than that, if its 1.5V, then you'll have to find a better way to step it down rather than using a votlage divider. I would suggest a voltage regulator, such as the LM317. As already suggested a shotky diode in series with the 3.3V power supply would also lower the voltage. Diodes drop voltage across them, and shottky diodes tend to drop about 0.3-0.4V. (normal silicon diodes drop 0.6-0.7V). So a shottky would lower your 3.3V from your PSU to 3.0 to 2.9. If you need 1.5V, then a dirty way of doing things would be to use several diodes in series to drop the voltage. for 1.5v to your boards, from 3.3v form your PSU...3.3-1.5 = 1.8V. Thats 3 (maybe four) silicon diodes. Make sure you use ones rated for 1A. 1N4001's are so common its scary.

So, with your diode in place, wire the power input to the flash boards in parallel. The voltage will be the same for every board, but they may draw different currents.

Then wire the outputs of these boards in parallel too. Then the caps in parallel, then connect it all up. The outputs of the chargers will have diodes in series with them, so current cannot flow 'back' intop the charging circuit. This means that if one charger is putting out 280V, and another is putting out 310V, the 310V charger won't force current back into the lower power charger. Make sense?

Each charging circuit will produce 280-330V at a small current. The voltage is at the maximum, but for lower charging times, you need current. That is why you wire them in parallel, not series. If each charger outputs 3mA (at 330V), then 5 will output 15mA, reduce you charge time to one 5th. You can just keep adding more because you computer power supply has more than enough current capabilty.

Don't worry about 'over charging' the caps. Yes, over their rated 330v will make them fail, but as long as you stick to what the charging circuits are designed for you should be ok. They wouldn't sell those cheap flashes if they failed after one flash, as long as you can emutlate their power supply without going over it. (1.5V if it took a single AA, 3v for two). Measuring the output voltage of these little chargers won't be much use unless its charging a cap. Open circuit (no capcitor connected) its voltage will soar, and possibly destroy your circuit. It was never designed not to have a cap connected.

If you need a diagram I can draw a simple one up for you.

Blueteeth

 

[Saibot]

wow. thats reely good blueteeth

so many things i didnt consider such as putting diodes on the outputs of each board..

i was kind of afraid that one of the boards was dead... and it was simply stealing current from the other ones.. shoulda thought of this lol.

also, i think the shottyky diode with 2x two paralelle boards in series would work? what would be the better method, connecting two boards in paralelle, which is then in series with another two boards in paralelle, OR the 4 diodes in series with 4 boards in paralelle?

 

[HarveyH42]

**broken link removed**

The neon consumes about 20 mA, might try removing it...

 

[Saibot]

**broken link removed**

The neon consumes about 20 mA, might try removing it...

thx..
neon should only light when caps are charged