Build your own CD battery tab welder for about $100.00+-

fechter said:
You might try going to a metal recycling place to find used copper buss bars. ....

The cheapest way to make buss bars from easily acquired material is to get some copper tubing from a local plumbing supply. Depending upon how you are trying to hook things up, you an use it 'as-is' or flatten it. I don't think you need 1/4 inch thickness nor an inch wide. That would be enough to carry half a million watts continuously*.

(* not a real number, I just pulled that out of my @$$, but you get the idea ;) )
 
I finally updated my site with some info on my experiments with higher voltage/ lower Capacitance (But with 'REAL' capacitors). Not much different than what I posted here a couple of weeks ago, but if you want to take a look:
http://www.ledhacks.com/power/higher%20V%20welder.htm
 
Holy crap, that thing blows nice holes in metal!

I wonder what effect the electrode pressure has on the weld?
It looks like the SCR handles the discharge fine.

What is the charging time at 3 amps?

Ever try discharging into a coil of copper wire and stick a magnet near it? :twisted:
I've seen big setups that could crush an aluminum can with a coil of wire wrapped around it.
 
I wonder what effect the electrode pressure has on the weld?

Well, if my experiments are any guide, the higher the voltage, the more pressure you need. Can't really estimate, but I was probably putting about 15 pounds of pressure on each of my electrodes at 30V to keep from 'blowing the tip off the electrode" and leaving it in the middle of the weld. Any less pressure and you might as well be arc welding.

What is the charging time at 3 amps?
About 10 seconds on average.... + 5 seconds, it seemed to vary somewhat.
Ever try discharging into a coil of copper wire and stick a magnet near it? :twisted:

Nope. But even under normal use, any ferrous items within about a foot 'jump' a bit when I weld. Even a loose nickel tab within a few inches will move, since Nickel is slightly magnetic.

In fact, I'm going to try using a reed switch on/near one of the electrodes (with a battery in series) to see if I can trigger a second pulse from a parallel Cap bank /SCR array and maybe avoid all the more complex electronics that you guys are playing with to generate dual pulses. (I tried doing the same thing with a scrap torriodal inductor around the electrode, and while I was getting some sort of pulse, it wasn't enough to activate a gate, or to actually measure on either a decent digital or a cheap analog multimeter.) I have no idea if it will work, but it is worth a try.
 
HA! It works!
The natural inductance of a CD welding pulse WILL activate a reed switch to trigger the gate of the SCR in two parallel Capacitor/ SCR banks.

What I did, was set up my first CD welder using the '1.5F' audio cap and the cheap 12V battery charger in parallel with the '5F' audio cap and 14V 14A power supply;
Taped a cheap little glass encapsulated reed switch to the negative welding cable just above the -electrode. Used the regular foot switch to activate the SCR to the '1.5F' cap...
And a fraction of a second later the '5F' cap fired.

I'd estimate that I had about a quarter-second delay between pulses. (I imagine this would vary quite a bit between different reed switches... The stiffness of the electrodes inside the switch and the inductance of the pulse along the cable would be the determining factors, I guess.) While the Pro dual pulse CD welders delay may be in the high nanoseconds to mid milliseconds range, this somewhat slower 'mechanical' delay would seem to have potential for ease of implementation and low cost for those of us who aren't competent enough to design and build more complicated electronic delay circuits.

With the dual pulse, I was able to tack nickel tabs to both copper and aluminum at 14V for the first time, and it gave great, solid welds from nickel to steel. The strength of the tack welds to Cu & Al were not good enough for making battery packs or anything, but were good enough to be encouraging.

I'm going to try this again with my 'real' capacitors at higher voltage when I get a chance. I'll try using a bank of four of the 36,000uF caps at about 24V in the initial pulse, then the other eight at about 30V in the second pulse.

Here is a crude drawing of the setup; I'd post a photo of the real thing, but there are so many wires coming and going, it would only confuse everyone.



Oh--- You may have to 'play' with the orientation of the reed switch.... In some positions (rotation) the reeds won't make contact.... I think that the flat plane of the reeds needs to be parallel to the inductance field....
 
This forum is simply the best! I'll pursue my quest to precise variable pulse width even harder now. I really want to be the first who welds copper to nickel/steel so I can use it for my A123.( as from the video form youtube I posted couple of posts back that shows a pro welder doing just that !.) I also wonder if it will be possible at the end to weld to coper with full length pulses and SCR's successfully.I wish my 8F cap had arrived yesterday as my A123s did! I'll have to be patient I suppose. And I'm not very good at that !
 
Wow!
I love the reed switch setup. Nothing could be simpler.
You might be able to enhance the reed switch performance by using a piece of ferrite around the wire. I've seen this setup on some hall current sensors. Something that resembles a torroid with a gap cut into it that's large enough for the reed switch to fit into. When the current pulse goes through the torroid, a high magnetic field will be present in the gap. This would make the reed switch more sensitive (work at lower currents). Even half of one of those split core noise filter things you see on computer cables would probably help out quite a bit.

I don't know how critical the delay is, but if your setup is working, then it sounds like it's in the right range.

Cool!
 
.....enhance the reed switch performance by using a piece of ferrite around the wire. I've seen this setup on some hall current sensors. Something that resembles a torroid with a gap cut into it that's large enough for the reed switch to fit into.......

Great Idea. I may even have something like that laying around in one of the parts boxes. If not, I KNOW I have a torroid of the perfect size that I can cut with a cutoff wheel. That would help protect the glass envelope as well.

Hmm.... I wonder... Would a complete gap be better, or just grind an indentation or drill a hole just big enough for the switch to fit in be better.. (I'm even worse at physics than I am at electronics :? )
 
I had an easier idea if you haven't already discovered how hard ferrite can be....

Since we don't really care about efficient energy transfer to the reed switch, you could use ferrous metal to make a magnetic circuit.

You could take a piece of steel coat hanger material or soft steel wire or sheet metal and bend it into a U shape to fit around the conductor. The top ends of the U could be bent toward each other, but leave a gap of (1/4"?). The contact area of the reed switch would go in the middle of the gap.

With either ferrite or steel, the thing is going to have a tendency to jump when you discharge. Some kind of glue or duct tape to hold it might work.
 
No problem at all cutting the ferrite with an Aluminum oxide cutoff disk... 'Like butter'. I didn't even have to use SiC or diamond.

I used hot melt glue the first time to secure the reed and torroid to the cable. But I broke the reed switch ... Not because it was jumping around, but because I burned the heck out of my finger; dropped everything and it hit a sharp angle at just the wrong place. THIS time I'm encapsulating it with Epoxy-ceramic putty (love that stuff).

Before I broke it, it was working very well. I thought I was going to beat Nemo at his goal to be the first to weld copper with a homebuilt, but I haven't quite found the trick yet...... Although if my Chromium-copper electrode material order was here yet, I might have. This dual pulse - higher voltage stuff is just completely too hot for regular copper electrodes. I had to regrind the tips after just about every weld. May have to even go to zirconium copper or even tungsten electrodes before I'm finished.
 
that is really, really KEWL!!!!!

i couldn't find a second SCR So I just connected up an LED and a battery. I could not get a reliable pulse with just a reed switch. worked sometimes and sometimes it didn't. so i cut a slot in a ferrite bead i salvaged from an old computer monitor (conveniently left by someone in the gutter) and put the reed inside the slot. works a charm.

what a creative use. i am using a few feet of 4AWG speaker wire for the electrodes. And every time i trigger the SCR these cables would literally jump. i knew that this was from the intense magnetic field generated, but it never occurred to me to use this to trigger the next stage.

rick
 
Happy New Years to everyone!

What a way to spend New Years Eve.... Most people would consider the way I spent the last five hours a waste of a New Years Eve, but I'm pretty satisfied, nonetheless.

Sorry Nemo, but I beat you to welding copper, aluminum and several other metals, (including a few kind of exotic ones) with a homebuilt Capacitive Discharge welder.
DPweldsS1.jpg


Ok, Here's the setup I used:
The Tab Welder II ( http://www.ledhacks.com/power/tab_welder%20ii.htm ) '5F audio cap at 14.5V
in parallel with the higher voltage welder ( http://www.ledhacks.com/power/higher%20V%20welder.htm )
with the reed switch near the negative electrode (as described up above a few posts) to activate the SCR of the 30V 432,000uF 'real' caps.

DPweldsS2.jpg
DPwelds3S.jpg

DPwelds4s.jpg
DPwelds5S.jpg

DPwelds6S.jpg


All the welds above are what I'd call 'serviceable' ; some aren't quite as good as I would like, but should hold up under normal use.
On the ones that look like copper on 'non'-copper metals; That is residue from the electrodes that 'burned off' the tips.
It is just too hot for 'pure' copper electrodes... I think with Cr-Cu or Zr-Cu or maybe Tungsten electrodes, things will work a lot better.

If you are planning to weld batteries using copper tabbing material with a homebuilt welder, it would be wise to use .008 to .012 inch thick strip. I had no luck at all welding .024 thick copper to anything; The .016 inch thick stuff produces marginal strength welds at 30V with the 'pure' copper electrodes (although at 40V with better electrodes it would probably work, OK, I'm guessing.) and the .004 copper is pretty flimsy, although it will weld nicely to steel at 22V, rather than 30V.

Watch the pressure on the tips of your electrodes!! The higher the voltage, the more pressure you need. 15 to 20 pounds on each electrode seems to be about right for around 30V.
I darn near burned a hole all the way through the copper coin above when I got a little careless and only had a couple of pounds pressure on the electrode.

Another thing I noticed... Using Fechter's idea of the ferrite 'C' around the reed switch makes the second pulse happen lots faster than just taping or gluing the reed switch to the welding cable. I can no longer actually feel the second pulse.. (That's probably a GOOD thing... More in line with the milliseconds delay on Pro dual pulse welders.
 
Good stuff here! Good pictures too!
Happy new year!and I admit...... You won!
What was your trick to weld the coper?
Higher voltage? Double pulse? Your welder is huge!
I'll use 0.012 inch coper to steel and nickel on A123s.
My cap is still not here. I'll still try to use single 8F audio cap bank.
I'm going to actually simulate the circuitry in one nice software before prototyping it.
I don't have all the parts yet . I can at least play in a virtual world.
I will approach the low voltage way as I saw it working in a pro welders.
As you reported the electrodes with higher voltage don't last, but still it is a great success for not much money.
Does your SCRs get warm? Caps? How is your audio cap doing?
 
That thing kicks butt! Cool.

Tungsten has a pretty high resistance, so you might be able to take copper rods and drill a hole in the end and insert a piece of tungsten so only a short length sticks out of the end. I have some nice tungsten TIG welder electrodes, which are available from the local welding supply place. Hard as hell. I wonder if carbon (graphite) would work? At least it won't stick to anything.
 
What was your trick to weld the coper?
Higher voltage? Double pulse?
A combination of the two, I believe.
That plus the fact that I used thinner gauge copper on a medium gauge aluminum than I had tried previously.
On thinner aluminum I was just blowing holes through it, and the thicker aluminum sucked up all the heat too fast.
Same with the thicker copper; It just sucks up all the heat, and since this is 'resistance welding' and copper doesn't have much resistance, it is hard to do.

I'm guessing that with .08 or .10 copper even a lower voltage, single pulse setup might work. Not sure if it will do .12 really solidly, but good luck... if your '8F' cap has 50% more power than my so called '5F' and a decent ESR, you might be able to do it.

Tungsten has a pretty high resistance, so you might be able to take copper rods and drill a hole in the end and insert a piece of tungsten so only a short length sticks out of the end.
That's what I planned to do. I figure that is why the old one I built back in the early 90s didn't work too well. I was using tungsten electrodes sticking about 2 1/2 inches from brass holders. If I do try tungsten, it will be more like a sharpened pencil; with copper instead of wood and only about 3/32" of tungsten sticking out, instead of the 'lead' (graphite/clay, actually)

I wonder if carbon (graphite) would work?
It might... A couple of weeks ago, I actually dug out a few pieces of graphite electrodes from my days of running an EDM machine and have them sitting close to the welder setup; I was going to try them sooner, but haven't gotten around to making a holder for them. If I do, it will be the same as I described above. And if It doesn't weld, at least I'll have a fancy copper bodied pencil that I can use to take notes. ;)
 
I imagine you were joking, Tyler;
No way that the tiny amount of platinum in even the best of sparkplugs would handle the current.
Although platinum plating the tips of the copper electrodes might make them last longer.
 
Platinum is definitely interesting stuff; Very tough and durable, high melting point (although only half as high as tungsten, though)
But the platinum sparkplugs (Incuding the ones that claim to have a solid platinum core) that I've destroyed.... Uh, 'reverse engineered'... don't have enough platinum in them to do anything any more interesting than what they are designed for.
It would take about a thousand of them to salvage enough platinum to make a really thin ring for your girlfriend.

I do want to try some 'pure' molybdenum electrodes if I can find some moly rod cheap enough though.
 
I finally got the 8F cap. It weights slightly over 1.3kg ! The building can start.CAP 8 FARAD max 24V.JPG
 
Just a heads up:

Had a 36V LiIon battery pack fail 'on the bench' last night.

Took it apart, (Tubular battery column of ten 32650 5AH cells, inside a tube) to find that one battery tab melted through, (at a sharp 180 degree fold, if that is significant) and two other spots on different batteries were intact but singed the plastic insert around the + nipple of one of the other cells.

Was using the 1/4" X .005 Sunstone Engineering nickel tabbing strip. I don't think I pumped more than about 700 watts through it, and then, only for a couple of seconds.

So, for those of us building Lithium or NiMH battery packs for e-Bikes, scooters etc. it would probably be wise to either use double thickness of the 1/4 X .005 strip, move up to the 1/2 inch wide .005 strip, or try to find a source for .007 or thicker tabbing material.

Fortunately the batteries all seem fine, and all the welds held perfectly.

That reminds me, I never did get an answer from Sunstone on their recommended limits on their tabbing material. Seems like they would WANT that information to be available, for the sake of their liability insurance if nothing else.
 
Back
Top