PDA

View Full Version : Sput welder?



Scatterplot
09-16-2007, 03:15 PM
So I came into a pretty large capacitor bank the other day. 9x 300 Farad, 7.5 volt monsters that were gonna be thrown away. I was trying to think of something to do with it, and I think I would like to make a sput welder. Here's the thing, I don't know the first thing about welders. I own one, but that's about it. Has anybody done something like this? I feel like any one of those caps could make me a good one. The voltage might not be quite right (too low) but who knows. I also don't know how to properly charge a cap or even check it's level of charge. I figure all I'd need to do would be to use a transformer to get the AC down to a reasonable level (I just got one of those too, outputs like 16 volts, IIRC, at 1.5 amps AC, I could grab a few high-power diodes to rough it out to get DC to charge the caps). Anyway I figure all I gotta do is short out the cap with a wire to something grounded and boom I have a wire. Any thoughts?

BTW I have a crapload of tools to derust and really enjoy powdercoating, so that's where this idea came from.

Any other large coke-can sized capacitor fun projects I would be interested in as well :)

jacampb2
09-16-2007, 03:34 PM
I have a really hard time believing a coke can sized cap is 300 farads, maybe 300 uF... I also have no idea what a sput welder is, so I probably can't help you, but an overcharged capacitor can explode, also hooking them up in reverse polarity, as well as running them at more than their rated voltage.

Whatever you do, be careful, but I think you may find them to be nearly useless w/ that low of a voltage rating.

Later,
Jason

Scatterplot
09-16-2007, 06:23 PM
It says 300,000 mf on the side. It's only 7.5 volts though- that's why I'm thinking it definitely is 300 Farads. I have no idea what the original application was, obviously some sort of huge power supply (could be a filter of some kind, this came from a University so they might have used it to power some benchtop power supplies. Then again it might have just been "sitting around")
http://img.photobucket.com/albums/v240/bertmcmahan/SSPX0455.jpg
http://img.photobucket.com/albums/v240/bertmcmahan/SSPX0456.jpg
http://img.photobucket.com/albums/v240/bertmcmahan/SSPX0457.jpg

A sput welder temporarily welds a piece of wire to a part that conducts, used for plating, powdercoating, anything requiring a conductive temporary joint that you don't want to clamp. Video:

http://video.google.com/videoplay?docid=-156416373727477390

ligito
09-16-2007, 06:24 PM
Sput is what happens when you try to spit and it dribbles down your chin, instead.;)

TGTool
09-16-2007, 06:42 PM
I thought that was redneck past perfect tense as in:

... Aww &%*# I done sput bacca juice on ma dawg!!!

Mac1
09-16-2007, 07:30 PM
Not 300 Farads.
300,000 micro farads. MFD

Your Old Dog
09-16-2007, 08:15 PM
Don't know if this is relevant or not.

Harbor Freight has a close out on "spud welder tips" Looked to be about 1/8" diameter and about 1/2 to 5/8 in length with a domed top on one end and flat cut on the other. They had them for half price. Looked to be something lilke 100 in package for something like $2.50. I don't know if this applies to this thread or not?

Lew Hartswick
09-16-2007, 08:21 PM
As Mac1 pointed out 300,000 MFD is only 0.3 Farads.
Now that is still a good size cap. but the low voltage wont do much
for a welder.
That must have been some power supply they were used on.
They are undoubtly electrolytic so be carefull with polarity.
...lew...

Scatterplot
09-16-2007, 11:08 PM
The prefix for micro is the letter mu, it looks like a funny lowercase u.

The prefix for milli is m.

Millimeters = mm

Perhaps on capacitors they do it differently, although I can't see why. I'm not arguing that the value seems large, I could be mistaken, but everywhere else in the world M means milli. Like I said though I could be wrong, but that's where I'm getting my 300F.



Anywho, upon more thought, I could just make some kind of what I think is called a "flyback" transformer, I think they can put out a crapload of voltage from a small amount. The transient cap drain should do well in creating a single burst of high voltage through a step-up transformer.

Scatterplot
09-16-2007, 11:41 PM
Don't know if this is relevant or not.

Harbor Freight has a close out on "spud welder tips" Looked to be about 1/8" diameter and about 1/2 to 5/8 in length with a domed top on one end and flat cut on the other. They had them for half price. Looked to be something lilke 100 in package for something like $2.50. I don't know if this applies to this thread or not?


Thanks for the tip YOD, I think I know what you're talking about, but I don't think thats it. Those are (I think) for body panel work, weld them on and then use a slide hammer to pull out dents. The thing I'm looking for just welds a small wire to something to hang it in an acid bath/powdercoat booth/anything that needs a conductive mount.

Forrest Addy
09-16-2007, 11:52 PM
.3 farads is pretty impressive but at 7.5 volts don't expect much in watt-seconds charge. Also that impressive bussing is mostly for appearance. The tiny screw and the little bitty post on the cap shows me it's strictly for power supply hold-up while a back-up power supply can be switched in.

Work that cap to major energy rates or even major ripple current and you'll most likely blow something. OTH working the time constant and L/C network formulas for one farad caps leads to some interesting numbers.

BCtech
09-17-2007, 02:33 PM
http://img.photobucket.com/albums/v240/bertmcmahan/SSPX0457.jpg


According to http://www.techwr-l.com/archives/9706/techwhirl-9706-00366.html

M means Mega or "10^6: mega M" so apparently,,,, what you've got there is a 300,000 Mega farad or .3 Tera farad capacitor.

Hmm, should be able to spot weld billets with that thing.

... but I'm betting it's actually 0.3 farad.

Scatterplot
09-17-2007, 04:18 PM
Lol. I figured it being an actual big M was out, so little m seemed closer. Is there any way to measure that easily?

Fasttrack
09-17-2007, 06:41 PM
Yes - some multimeters actually have a setting for measuring capacitance, but this is how they do it so you can do it with a normal one and a power source

I(t) = Io (e^-t/RC) where Io is the inital current given by IR=E (ohms law)

The circuit should be constructed with a resistor of known resistance in series with the capacitor. You need to measure the current flow and the time. Start the timer when the power source is turned on and watch for the current to drop to zero and then solve for C.


Basically all you need to do is measure the amount of charge being stored in the capacitor, but since your current changes with time this is hard to do without a computer to log a bunch of data points to simulate the current graph. The definition of capacitance is C=Q/V where Q is the charge stored for some potential. So if you hold the voltage constant and log the current measurements every 1/60 of a second or something, you can then integrate the function of best fit for the curve. Or just use simpson's method to approximate if you dont have a program to find best fit.


But it is .3 Farad. A farad is a pretty big unit when it comes to capacitance - most capacitors are nanofarads or microfarads. Very rarely do you see one large enough to be .3 farad so thats still really good for a free capacitor.

Scatterplot
09-18-2007, 03:47 AM
My experience with capacitors is sadly limited. Thanks for the math though, I can definitely do that. Actually there should be a multimeter nearby that will do it easier, lol.

Thanks for the info.

On one last point, I have seen a few 1 Farad capacitors that were the size of a handful of quarters. Now they were like .06 volts but they were 1 Farad :D


So any more thoughts on the sput welder idea? What kind of voltage do you think I would need to spark a wire onto something like in the video?

Swarf&Sparks
09-18-2007, 08:11 AM
You don't need a capacitor for that, a car battery will do just fine :D

Lew Hartswick
09-18-2007, 10:46 AM
Fastrack:The circuit should be constructed with a resistor of known resistance in series with the capacitor. You need to measure the current flow and the time. Start the timer when the power source is turned on and watch for the current to drop to zero and then solve for C.

Just how many time constants are you going to use to get to "zero"?
:-)
...lew... (electrical engineer)

Dawai
09-18-2007, 12:21 PM
Everyone else accumulates parts for future projects? right?

I had a pair of large 10hp 480 volt inverter drives I let get messed up, then gave them to the junk man. The capacitors in them 280 volt dc bus would have been perfect for that edm machine I have wanted to build for years.

Save them capacitors.. you can double voltage by putting them in series. Wrong polarity thou and they make a loud noise and look like a toilet papering at Halloween.

Weston Bye
09-18-2007, 12:57 PM
Some capacitors are not made for rapid discharge. I just went out and checked our magnetizer. Type HES capacitors are used to rapidly discharge through a pair of coils to make a magnetic field to saturate magnet material. Other types may work for a while...

Fasttrack
09-18-2007, 01:30 PM
Lew -

As many as i dang well want :D :D

Yeah that equation ... you can't solve it for 0; don't know what i was thinking. But you can solve it for some measured value at a known time.

i.e. .2 amps and 3 seconds or some such. So basically you need to be able to look at two places at once ... :o

I know theres a way to do this but its been a couple of years ... you all should be impressed i even remembered that dang thing

Fasttrack
09-18-2007, 01:41 PM
I guess, if i did this correctly, after about five time constants elapse your current should be reading zero or fairly small compared to the initial current so you could take the time it took to drop to a small flow or to zero and divide by five and set that equal to RC...

Swarf&Sparks
09-18-2007, 02:44 PM
I'll still go with the car battery.
Electronic or mechanical timing.

We usta do it mechanically with a "rat trap" blade.

I'd do it with silicon now. Just don't have that many high power transistors in the the junk box :(

Evan
09-18-2007, 03:59 PM
What Wes said. If it doesn't say pulse rated on them they won't last long at all, maybe a few hundred discharges at most before they fail due to excess internal currents.

And, those aren't pulse rated. They are standard computer caps for linear power supplies. They also may be very old in which case do not bring them up to ull voltage right away. Start off at a couple of volts and leave it there for an hour or two then slowly do the same a volt at a time up to seven volts. The insulating layer needs to be reformed to full thickness. However, if you use them for pulse discharge it won't matter much anyway.

I have a collection of those myself.

http://vts.bc.ca/pics/caps1.jpg