rectify electricity before transformer?

can you rectify the US mains (120vac, 60hz) as it, and used the pulsed dc to feed into a transformer and get a dc voltage out of the transformer? I'm asking this because it would be much much more efficient for high current situations. Here's my specific situation:

I want to build a capacitive spot welder and a capacitive discharge cutter in an all in one device. To charge the enormous capacitor rapidly (1.5 farads in my case), I need a very high current power supply. I'm using a modified microwave transformer that was rated at 1kw, and it is now modified so that it outputs 12 volts AC at around 80 amps (probably less because of losses). This must be rectified to charge the capacitor.

Assuming a perfect transformer (for arguement), it would have:
120volt, 8.34 amp input
12volt, 83.34 amp output

the following is with full bridge rectifiers, assumed voltage drop: 1.4v
If the power was rectified before the primary, the loss on the bridge rectifier would be: 11.67 watts
If the power was rectified after the secondary, the loss on the bridge rectifier would be: 116 watts

huge difference

so once again, the final question is: Does rectified input of a transformer = rectified output? Does the 120hz pulsed dc (as opposed 60hz ac) matter? Could this setup charge a capacitor?

Comments

[Dr.Bill]

After. Transformer will not work the other way.

[framistan]

What you are talking about is NOT standard practice for building a POWER SUPPLY. But you are building something different.... You may
want to look into using a FERRO-RESONANT transformer.  They put-out
an almost SQUARE WAVE... but it's output needs to be RECTIFIED still.  Also, the
transformer will put out AC not DC if you rectify the input.... but your
output will be 120 Hz not 60Hz.  I suggest you seek advise from someone
who knows electronics as a designer.  You will need some method of
LIMITING the amperes or you will destroy your diode rectifiers without
some method of limiting current.  Perhaps a few 100 watt bulbs in
parallel with EACHOTHER... but in SERIES with the charging capacitor
would do it...  Be carefull working with this could be VERY dangerous. 

[Wesley666]

Get a microwave from the dump and get the big transformer out of it.  Run mains power through BACKWARDS, so mains go to the leads connected to the...I forgot the name but its the part that puts out the microwaves.  The leads that the mains normally plugs into will put out 12v at 60amps which should be enough to weld.  Also go with an older larger microwave as they have heavier duty transformers.

[alexhalford]

Actually, for what it's worth, the transformer will work; transformers work on the current through an inductor changing, it does not necessarily have to reverse. So, full wave rectified, unsmoothed DC will run a transformer, although very inefficiently (losses due to ringing etc).
I just thought I'd mention this for the sake of accuracy. However, I agree entirely with the previous replies, it would be far more efficient and effective if you rectify after the transformer.

[steveastrouk]

No. Err. No. Dissipation in the recitfier is negliglible, the charging time is completely negligible from that supply it will charge in 1/2 cycle. Feed the capacitor with a series resistor of even 1/2 ohm.

[guyfrom7up]

oh, and i guess i should have made it clear that it's charging the capacitor many many times, not just once. It's charging and then discharging the capacitor as rapidly as possible for max power, so it's almost continuously putting a load on the rectifier, not for just half a second.

[guyfrom7up]

people that have similar setups say that the bridge rectifier gets really hot. Like on the verge of smoking hot. That's even with heavy heatsinks and fans. So obviously the rectifier is wasting a lot of heat.

[Dr.Bill]

GAWDDDD!!!