i'm making a capacitive welder

[mkawa]

i'm just posting this thread in case i die in the next couple of weeks. OK THAT'S ALL DUDES

[SpAmRaY]

i'm just posting this thread in case i die in the next couple of weeks. OK THAT'S ALL DUDES

Would it be a side effect of the welder?

[bueller]

i'm just posting this thread in case i die in the next couple of weeks. OK THAT'S ALL DUDES

BE SAFE DUDE, RUBBERISE EVERYTHING!!!

i'm just posting this thread in case i die in the next couple of weeks. OK THAT'S ALL DUDES

Would it be a side effect of the welder?

More like the massive string of capacitors all storing a charge, can be quite dangerous if not careful.

[mkawa]

there will be a lot of silicone involved.

[mkawa]

actually that sounds dirty

oh well

[rowdy]

First excessive lube, now excessive silicone, what's going on here??

[mkawa]

...

[JPG]

... kk, deleted post, but there was a strong correlation!

[mkawa]

we do not endorse these kinds of chauvinistic things on geekhack sir

[mkawa]

THIS is an elna dynacap DZ-2R5D475T



it's rated at 2.5v 4.7F. ESR is 200mOhm

i have 22 of them.

[mkawa]

so i've worked out the final supercap specs. it's going to be a 14.1F capacitor with ESR of 100mU at 20v. i will be charging it to 18v and using it with at least a 200A thyristor for DC welding. peak current output will be 180A at 18v. it will be foot pedal operated but otherwise direct current. i toyed around with bringing current down then buck driving with a large number of beefy sinked darlingtons, but for efficiency and simplicity will just be putting the capacitor network together so that i can sink them directly into the welding point. (as it's a CD welder it can only weld points. it may have enough reservoir capacity to cut as well, but i don't expect to be able to bead long joints. frankly, it's mostly for making thermocouples and stitching the sides of sheet metal for the ssk plates).

[oTurtlez]

so i've worked out the final supercap specs. it's going to be a 14.1F capacitor with ESR of 100mU at 20v. i will be charging it to 18v and using it with at least a 200A thyristor for DC welding. peak current output will be 180A at 18v. it will be foot pedal operated but otherwise direct current. i toyed around with bringing current down then buck driving with a large number of beefy sinked darlingtons, but for efficiency and simplicity will just be putting the capacitor network together so that i can sink them directly into the welding point. (as it's a CD welder it can only weld points. it may have enough reservoir capacity to cut as well, but i don't expect to be able to bead long joints. frankly, it's mostly for making thermocouples and stitching the sides of sheet metal for the ssk plates).

3240 Watts of POWAH! Good schtuff kawa.

[mkawa]

the thyristors have significant power loss (they need to be sinked big time as well as the cap network) so it won't quite hit optimal power or current. that said, it should be plenty for small stitches. high current battery connector welding is another application

(lifepo4 cells are fascinating me right now. incredibly power density WITHOUT the volatility of li ion or li polymer (which was the initial attempt to keep the lithium in check.. the tradeoff is that they don't deep cycle, period. deep cycles actually temporarily kill the cells (!!!). MCUs to the rescue! etc).

[TheSoulhunter]

Found this clip on a old cellphone...
Thin aluminum sheet -VS- 380A A/C Arc, fun times! )

[mkawa]

SWEET

[mkawa]

i present this without comment:

[Tarzan]

i present this without comment:

Show Image

Good grief!

[RabRhee]

i present this without comment:

Show Image

Are these the capacitives that you want to weld together? :P

[Binge]

My jimmies are scared for Kawa... be safe fine prince

[AKIMbO]

Kawa's crazy.

[TheSoulhunter]

Nice battery!

[MOZ]

i present this without comment:

Show Image

Ca, ca, ca, capacitors!

[tjcaustin]

Did you find your keys before you die?

[Thimplum]

i present this without comment:

Show Image

That's awesome! Will you be providing source files for us?

[damorgue]

i present this without comment:

Show Image

Safe handling guide: After unplugging, let it be for a few years to discharge. Then take long insulated grounded metal rods and poke around with them. Then use a large pair of plastic grippers and throw the entire thing in the river. It should now be discharged and ready to be left there forever.


Farad count on that?

[Game Theory]

Safe handling guide: After unplugging, let it be for a few years to discharge. Then take long insulated grounded metal rods and poke around with them. Then use a large pair of plastic grippers and throw the entire thing in the river. It should now be discharged and ready to be left there forever.


Farad count on that?

)

[MOZ]

i present this without comment:

Show Image

Safe handling guide: After unplugging, let it be for a few years to discharge. Then take long insulated grounded metal rods and poke around with them. Then use a large pair of plastic grippers and throw the entire thing in the river. It should now be discharged and ready to be left there forever.


Farad count on that?

LOL. He is making a battery.

[mkawa]

4.7F * 5 = 23.5F (approximately, too many unknowns to actually measure exactly at the moment)

i'm not getting enough current out of it right now. one of the banks has an issue. so sadly it isn't all that dangerous yet. peak current at the moment is about 10A, so i suspect one bank has a couple cold joints or a bad cap in it. my calculated peak current was over 400A @ 18v. it's also possible that i simply don't have a measurement device that can capture the peak current interval, as you only get peak current for like a ns before it discharges below say 12v and then sits at 5-10A the rest of the way down.

also, i have a 300W 8ohm resistor that i discharge it into.

here's the current experimental setup. it's DC coupled right now but i have darlingtons to buck drive if i have to and a whole bunch of fast power mosfets if i need to go the other direction.

dicharging it through those 18ga clamp wires is fun. HOT STUFF COMIN THROUGH



the big brown thing that looks like a turd is a single resistor. it's a bunch of nichrome wires baked into a 1.5" diameter piece of ceramic. i collect huge passives. it's a hobby.

[mkawa]

oh, the little brown things are 50W 1.2ohm resistors. those are fun too. they can take a lot more than 50W

[mkawa]

crap, i found some wtf wires in the middle of the board, surrounded by capacitors that i can't get to at all.

[MOZ]

Mad scientist at work

[mkawa]

i was trying to figure out why the ESR seemed so high and i realized it was actually all the wires in the network, not the caps themselves.

RESULT:

it pops now when shorted  my clamp meter isn't fast enough to see the transient current, but it is quite high.

[StrikeEagleCC]

mkawa, that pic of your bench made me drool.

[mkawa]

[StrikeEagleCC]

Is that first aid tape?

[mkawa]

the tape is fiberglass weave in tape-form with an ultra-high temp silicone adhesive (260c+). it's a much more effective and tougher insulating electrical tape than the classic rubber crap. i believe there's some kind of resin impregnated in it as well, since i've already heated some to 200C and all that happens is that the resin takes on a glossy appearance, no other ill-effects and it maintains both its strength and dieletric properties. i purchase it from mcmaster, but i'm sure there are cheaper sources. it's made (like everything else) by saint-gobain

http://www.mcmaster.com/#fiberglass-tape/=o9y0ux

[mkawa]

also, i thought about it quite a bit and i decided the other fairchild darlingtons just weren't beefy enough, so bought two of these: http://www.onsemi.com/pub_link/Collateral/MJ11028-D.PDF

max collector amperage continuous is 50A and it has an h_{fe} of 1000x @ 5v V_{be}

MUAHAHAHHAA

[Melvang]

Just want to give you some notes for personal info on actual welder voltages and amp for actual welding

Stick welding with 1/8" 7018 weld rod ~71.2 volts and I run at about 115 to 120 amps unloaded

Wire welding I am a little fuzzier on but I believe amps was set to around 23.5 but not sure where voltage is set to but that was for DC welding.

AC welding for the most part is only used on aluminum due to the way the aluminum reacts to the weld arc and the tungsten.  Also aluminum has a fairly low melting temp (1244F).  It can be melted on top of a charcoal chimney for grills with some forced induction via shop vac with hose plugged into the exhaust port.

[mkawa]

the rectifier in the last picture (which i've tossed and decided to require point to point) will make this a point discharge DC welder. i'm not sure what my transient amperage at full charge is @ the full 18v, and won't until i get the rectifier set up. once i do i'll report and see if i even need to buck convert. basically, if i throw those buck converters on there, i'm going to be getting an initial 1kA transient then a few hundred amps sustained @ 5v. if i just rectify, i'll get the initial pop transient of maybe 500A and then it will actually settle around 20A @ 18vDC

[mkawa]

which reminds me, i'm going to have to pick up more safety gear. i'll probably pick it up next time i'm at IMS grabbing remainder stock

[Melvang]

If you are going to be picking up welding gear you might want to get a cheap welding hood or borrow one.  Extended exposure of your eyes to the massive amounts of UV light that is given off in a weld arc will cause what is called arc flash.  Basically what it amounts to is a very painfull temporary blindness.

[mkawa]

oh yah, just discharging the caps with bare copper wire left an afterglow. the actual welding light at 1000A peak is going to be INSANE. especially switched at high speed to maximize power. i plan on getting a full auto-shading hood from IMS or MCM

[mkawa]

they had a sweet one at IMS with flames on the side (really!). yea or nay?

ims = industrial metal supply, a local awesome metal supplier that serves all of the southwest.

[Melvang]

Personally I don't trust cheap auto tint hoods.  To me that is something that if it is cheap it is crap.  My suggestion is just get a Jackson hood with the large lens and buy a gold tint lens at shade 11 minimum

[mkawa]

ok, i'll run my cart by you before i buy. thanks melvang! going blind would suck. so would burning my hands off

[JPG]

What do you want to meld with it? Your car? Or just for fun?

[mkawa]

thermocouples and SSK plates primarily. anything else i get to fry is just for fun

[mkawa]

melvang: does this fit with what you were saying?

http://www.mcmaster.com/#56405t63/=o9zzcc
http://www.mcmaster.com/#9716t26/=o9zzpg

presumably you can replace the polycarb lens that comes on the hood with the gold-coated one? or is that something you add onto the lens that comes with the hood

[TheSoulhunter]

Personally I don't trust cheap auto tint hoods.  To me that is something that if it is cheap it is crap.

Indeed...
3M Speedglass or bust!

[mkawa]

aaa! i don't want to die. someone tell me what to get

[damorgue]

You won't die, blind maybe, but very much alive

Oh the irony, touch "live", and you won't be anymore...

[mkawa]

ahahahahaahah

[Melvang]

http://store.cyberweld.com/jahswehe1.html is the same hood I have at work and yes you can change out the poly carb lens.  I just recommend gold lens because you might want to see what you are doing and the gold tint gives much greater contrast around the weld than other color tints.  just make sure when/if you buy a different lens it is shade 11 and it is the same size as the hood you are putting it in.

[mkawa]

at least they're cheaper than racing helmets. my god those are pricey

[TheSoulhunter]

Perhaps also get some eye-drops (there are specific ones) in case you get a flash-burn anyway...
They are cheap, they help a lot, and it's a good idea to get em beforehand as you shouldn't drive or walk around with burned eyes, eh?

Also, from my experience:
Mild flash-burn: Eyes are stoned red, they might burn somewhat, they are over sensitive to light, it might feel like you have dirt/sand in your eyes -> Usually goes away after a 1/2 day
Strong flash-burn: Eyes are demon red, you can't even bear room lighting, eyes hurt, strong feeling of having dirt/sand in your eyes, degraded vision -> Go see a doctor, really!

[Melvang]

Perhaps also get some eye-drops (there are specific ones) in case you get a flash-burn anyway...
They are cheap, they help a lot, and it's a good idea to get em beforehand as you shouldn't drive or walk around with burned eyes, eh?

Also, from my experience:
Mild flash-burn: Eyes are stoned red, they might burn somewhat, they are over sensitive to light, it might feel like you have dirt/sand in your eyes -> Usually goes away after a 1/2 day
Strong flash-burn: Eyes are demon red, you can't even bear room lighting, eyes hurt, strong feeling of having dirt/sand in your eyes, degraded vision -> Go see a doctor, really!

This^^

My dad got flash burn at work last summer.  He wasn't the one welding.  But he was the fire watch inside of the stainless steel tank.  The temporary blindness is a very severe flash burn and hurts like nothing else you have ever experienced.  And just an FYI the UV rays can still be damaging at up to 50 feet.

[TheSoulhunter]

Also, when welding for extended time (not just some spots) it's a good idea to cover ALL your skin if possible!
Got my arms burnt (it's like a sunburn) more than once because I worked in a T-shirt next to a welder...
Kinda looked funny because my arms got red or tanned but the hands stayed pale (...gloves)

[Melvang]

Yup and the burn on your skin from welding in short sleeves is more like being in a tanning booth for to long than being in the sun to long.  I think the welding one hurts more.

[mkawa]

presented without comment

hey melvang, what do you think about the passive halo X jackson series helmets? i'm kind of starting to lean toward a 3M speedglass unit with a big lens and respirator. not for just this, but because i want to eventually move towards automotive work (stitch weld crash frame yaaaaaaaaaah), and going blind and/or dying are not on my list of things to do in the near future or really at any point thoughts?

[mkawa]

ok, one comment. the black square is a bridge rectifier that's rated for 500A peaks. and yes, that entire thing on the bottom is an alu extrusion heatsink.

[Melvang]

The 3m speedglass is one of the better brands from what I hear about the auto tint hoods.  I can't say really about that particular model line though, as I haven't decided to take the plunge into a speedglass yet.  But I don't think I have heard any complaints about that brand yet. 

What sort of thermal interface do you have between the heatsink and the components?

[mkawa]

the cap network is fiberglass composite insulated and then JB welded to the heatsink. the fiberglass a massively effective dieletric, but has reasonably thermal conductance. more importantly, the cap network has very large surface area. the rectifier has a non-conductive steel plate over the internals on the down-side which is JB welded to the heatsink. i have thermal epoxy, but JB weld has significantly higher thermal conductance than that crap and much much more shear and tensile strength once cured. the power mosfets and the buck drivers need to be insulated so they'll probably get thermal epoxied at low pressure to maintain the dieletric properties of the thermal epoxy. back-sides and pins will all be potted with various grades of RTV silicone rubber depending on thickness needed etc. i toyed around with the idea of using silicone rubber as the dieletric and then JB welding, but the silicone rubber is a pretty extreme thermal insulator, so i'm going to stick with the thermal epoxy and generous potting.

one thing to note is that everything is very low voltage, so the chance of arcing is minimal. the reason to pot is mostly accidental contact.

[Melvang]

the cap network is fiberglass composite insulated and then JB welded to the heatsink. the fiberglass a massively effective dieletric, but has reasonably thermal conductance. more importantly, the cap network has very large surface area. the rectifier has a non-conductive steel plate over the internals on the down-side which is JB welded to the heatsink. i have thermal epoxy, but JB weld has significantly higher thermal conductance than that crap and much much more shear and tensile strength once cured. the power mosfets and the buck drivers need to be insulated so they'll probably get thermal epoxied at low pressure to maintain the dieletric properties of the thermal epoxy. back-sides and pins will all be potted with various grades of RTV silicone rubber depending on thickness needed etc. i toyed around with the idea of using silicone rubber as the dieletric and then JB welding, but the silicone rubber is a pretty extreme thermal insulator, so i'm going to stick with the thermal epoxy and generous potting.

one thing to note is that everything is very low voltage, so the chance of arcing is minimal. the reason to pot is mostly accidental contact.

Gotcha.  I just wanted to make sure you had something in there to help move that heat around.  I know you being mkawawiki that you probably knew this but just making sure.  Sometimes even the best of us overlook something big like that.

[mkawa]

semiconductor ratings are always functions of temperature at the semi. the datasheet generally gives an approximation of the derating vs temperature. let's just say that if the entire thing hits 100C, things are not good in this circuit.

[mkawa]

cap network, 500A peak rectifierr, 33mfd smoothing cap, 750A of high speed power mosfet capacity. waiting for welding helmet before i make the first spark (have two copper chromium spot welder electrodes waiting for first spark as well). if i'm not seeing enough current at 18v, i will buck boost to 5v with a 200A (input, h_{fe} is about 1000) peak of darlingtons.

[MKULTRA]

(Attachment Link)

cap network, 500A peak rectifierr, 33mfd smoothing cap, 750A of high speed power mosfet capacity. waiting for welding helmet before i make the first spark (have two copper chromium spot welder electrodes waiting for first spark as well). if i'm not seeing enough current at 18v, i will buck boost to 5v with a 200A (input, h_{fe} is about 1000) peak of darlingtons.

So uh is there any chance this will blow up in your face?

[mkawa]

no

[JPG]

Any chance you will record/stream this event?

[Melvang]

record please I doubt I will be able to catch a stream

[mkawa]

IT WILL NOT BLOW UP LA LA LA LA AL LALA LA LAL A LALAL LALA

but i did order a jackson halo x and some pigskin gloves.

but not in case it blows up, because it won't

[JPG]

I am confident that it won't blow. But it could melt, spark, do nothing at all, etc. 

I want to see you first attempt!

[mkawa]

for my own ego^H^H^H^H safety, i will not be recording first spark.

i will be focusing on not dying, however. i've heard that this is quite important.

[Melvang]

if all you are doing is basically spot welding then I wouldn't worry about the leather gloves.

[mkawa]

i got pretty cheap ones just in case. the problem is that i don't have a hands free spot welding setup yet, so i'll be holding the electrodes in some form

[damorgue]

I would get a nice grounded rod with insulation to be able to discharge everything from a distance in case something is wrong. You do not want to have nothing happen and don't know why or what is charged or not.

I wonder how long it would take that charge to diminish naturally. Might as well build a concrete wall around it, add a pond on top and call it a day.

Edit: That sounded overly negative (pun), but I am just messing around. Looking forward to the results.

[engicoder]

When you are looking at building mark II, I am pretty sure I could get you a number of these for next to nothing.

http://www.avx.com/docs/Catalogs/bestcap.pdf

I think they would probably work better for your purpose.

[mkawa]

When you are looking at building mark II, I am pretty sure I could get you a number of these for next to nothing.

http://www.avx.com/docs/Catalogs/bestcap.pdf

I think they would probably work better for your purpose.

MY GOD, YES. PM INCOMING

I would get a nice grounded rod with insulation to be able to discharge everything from a distance in case something is wrong. You do not want to have nothing happen and don't know why or what is charged or not.

I wonder how long it would take that charge to diminish naturally. Might as well build a concrete wall around it, add a pond on top and call it a day.

Edit: That sounded overly negative (pun), but I am just messing around. Looking forward to the results.

did i not take a picture of my collection of huge ceramic resistors? i can discharge the entire network in about a minute.

also the black stuff you see in the other pictures is carbon fiber textile weave. my desk is lined with it. i regularly plop stuff at 500C+ onto it with a completely disregard for flammability (because it's not).

[mkawa]

it works.



muahahahahaah

[Game Theory]

Welding late at night?!

[JPG]

it works.



muahahahahaah

Picture or it didn't happen!

[mkawa]

 

[Photekq]

Just saw this thread. That's real neat-o kawa. Nice work!

[mkawa]

two fun things:

quick spark tests on a case hardened steel rod



a 300w (rated, have put way more through it) 8ohm resistor

[rowdy]

Every time I see this thread I have an image of mkawa doing something like this:

[mkawa]

that's cool and all, but it's huge, impractical because you can't move the electrodes or rectify it, and you have to pull all that amperage from the grid all that once. there's negligible energy storage in a transformer. basically he stole the transformer from a telephone pole.

[mkawa]

this is more related, as it involves energy storage and then a huge quick discharge: http://205.243.100.155/frames/shrinker.html