Sciencemadness Discussion Board

best way to get low volts high amps?

floggedaway - 19-2-2014 at 00:09

title says all. i want your opinions. looking for <5 volts >50 Amps with an electronic circuit.

ZeroHour - 19-2-2014 at 00:24

Rewire a microwave oven transformer? Try googling, perhaps...?
Not that hard to find.

BromicAcid - 19-2-2014 at 04:24

Certainly results could be found online, but I would figure that floggedaway is looking for someone here with actual experience to chime in on what is the <b>best</b> way to get to that combination of volts and amps. My two cents would be a welding power supply. If you are looking for DC current (i.e., for electrolysis since you did not specify) then an inverter welder supply. I have seen examples online where multiple microwave transformers were rewired and put in series to give a result that could weld metal, but I have also seen other examples where the results are much less reliable. Depending on your end use and reliability needs you might want to plunk down the cash on a pre-constructed device.

testimento - 19-2-2014 at 07:04

This is an easy one. For power up to 1000W get a microwave transformer. Cut off the secondary and rewind it with as thick copper wire as you can fit into for desired voltage. They are usually rated 1:1 for mains voltage, so in europe 220v in you will get about 6v out with 6 loops on secondary. Welding trannies are good to even as high as 3-4 kilowatts.

After transformer you will need a rectifier. They are sold in electric stuff shops. It should be rated high enough or it will burn. I bought myself a series of 2000A rectifiers total, because single one of 500A got too hot to touch. The rectified line is your positive, or anode, and the other wire goes straight from the trannie to the subject. Remember good cooling!

bfesser - 19-2-2014 at 07:22

Quote: Originally posted by testimento  
The rectified line is your positive, or anode, and the other wire goes straight from the trannie to the subject.
That is only the case if you want a junky <a href="http://en.wikipedia.org/wiki/Rectifier#Rectifier_circuits" target="_blank">half-wave rectifier</a> <img src="../scipics/_wiki.png" />. I'd recommend using a full-wave <a href="http://en.wikipedia.org/wiki/Rectifier#Rectifier_circuits" target="_blank">bridge rectifier</a> <img src="../scipics/_wiki.png" />. Of course, appropriately rated capacitors should be added across the output, and overcurrent protection (fuses are easiest) is a must.

testimento - 19-2-2014 at 09:03

Oh yes, diode bridge is always a better option, because the single-rectified wave is practically 50% power from the acual output. That's why one would have to use 4kW tranny to get 2kW of actual power out.

The only major problem with bridge might be the current capacity. At least in my country only monolithic rectifiers are available in 3-number currents. Electronics shops will deal a big range of smaller powered bridged, up to few tens of amps, but when we get up to hundreds of amps, things are getting nasty expensive. I'm certain that there are a lot more powerful ones available. A single 200A rectifier dictates the whole currency of the system, whereas four 200A would give 800A total output. Ebay seems to be offering mostly high voltage rectifiers. Bad thing on big currents is that one will need big chunky components and the mere mass of it will explain the copious costs.

Well there is, 300A 2500V:

http://www.ebay.com/itm/MFQ300A-2500V-Silicon-Controlled-Mod...

[Edited on 19-2-2014 by testimento]

floggedaway - 19-2-2014 at 11:02

what about getting some heavy ferrite cores and building a full bridge converter.

gregxy - 19-2-2014 at 11:08

Here is a single diode: 100A at 100V, about $4

http://www.digikey.com/product-detail/en/VS-100BGQ100/VS-100...

The problem with diodes is they always drop about 1V when conducting. So at 50A they dissipate ~25W each 50A*1V*50% duty cycle. For a bridge it will be 100W since the bridge contains 4 diodes. So you need a hefty heat sink.
Don't place diodes in parallel to increase the current, one will hog all the current and burn out. However you could use multiple cells and use a separate diode/bridge for each cell.

A better choice is to use MOS devices an turn them on and off with logic. 12AX7 published such a design in this forum, but
its fairly complex.

Another choice is use multiple cells each powered by its own PC power supply, 5V on a PC supply should be good for 20A.
Some PC supplies can be made to have a variable output voltage 0-5V, 0-12V etc. if you can find the internal reference and insert a pot between it and the regulator.

Praxichys - 19-2-2014 at 11:35

The easiest way is to get a computer power supply. Most 5V rails today support 40A, and you can find them for $20 brand new. I have a whole pile of them pulled from computers I have found/owned over the years.

There is one on newegg.com for $13 that will do 36A on the 5v rail. Check your local computer store. A lot of times they will sell you an old one from a scrapped computer for $5.

Marvin - 19-2-2014 at 11:40

floggedaway,

You need to tell us more. What sort of regulation requirements, current or voltage? What is it for?

Why do you want to make it yourself? What is "best", power efficiency, cost?

testimento,

For reasons gregxy has explained a full bridge is usually a bad idea at high current. A better choice for that design would probably be a split secondary, but then MOTs are not usually designed to be run at full power for a long time. Also, that ebay listing is not for a bridge rectifier.

testimento - 19-2-2014 at 12:21

Are you sure parallel is not good for multiple rectifiers? I have installed several parallel rectifiers and they work fine with current that I tested only for one and it instantly got very hot.

Metacelsus - 19-2-2014 at 12:37

It depends on the specific type. As a general rule, putting power semiconductors in parallel without protection causes failure.