Resistance-No Resistance wire welder

[LucentShadow]

I built my welder according to the orig. specs, and have been using it as many others here have: 35v with 1000uf. I have been getting good welds at 35v with 30 nic to 30 kanthal with just a couple of breaks when coiling. Using 32 gauge half of the welds break. I used the joule calculator from this link : Electronics 2000 | Capacitor Charge / Energy Calculator . It's telling me that I am getting just over half a joule and its telling me I need 45v just to get a little over a joule with my 1000 cap.

Has it been working for my 30 gauge, even though I am not close to 1 joule, because of a balancing act between voltage and energy?
To get better welds with the 32 gauge, should I go down or up a little in voltage (my LMBoard can go up to 37V)?

Interesting. I'd probably concur with the above suggestion of trying lower voltage first, as it seems that it may be too much energy when a smaller wire fails at settings that work with a bigger wire.

I did not build mine to the original specs, at all, so it may be that mine needs more energy because of the design differences.

There are some things at play that are not easy to quantify without proper equipment, such as the equivalent series resistance (ESR) of the capacitors. I used ones bought from Radio Shack, and they were probably not high quality, plus I have 3 of them in parallel. I suspect that mine may have higher internal resistance than some, causing them to dump their energy slightly more slowly. This may work in my favor, making for a slightly more controlled weld. Not sure.

From my experimenting, adding more capacitance worked better than increasing voltage, if more power is needed. I suspected that might be the case when I built it, so I used dual banana plugs for the output, and planned on using those to add capacitance.

 

[Scope666]

around 80v makes a perfect weld with .25 nickel and .20 kanthal

Nice, looks great!!!

 

[MagicPanda]

I was going to just follow the tutorial and add some capacitance (2000uF) but the fuse: 507-1032-ND is out of stock.

I can't really find any other fuses that are 2A, 125V, and slow acting. I'm not sure how important any of those ratings are though, I would guess a lot of other fuses would work fine.

I'm pretty confident I can build this box, I've done a lot of other small electronics projects in the past. So I already have most of what you need on my work bench. (iron, multimeter, various solder, wire, hot glue gun, superglue, helping hands, flux, electric tape, solder sucker, etc...)

 

[Scope666]

I was going to just follow the tutorial and add some capacitance (2000uF) but the fuse: 507-1032-ND is out of stock.

I can't really find any other fuses that are 2A, 125V, and slow acting. I'm not sure how important any of those ratings are though, I would guess a lot of other fuses would work fine.

I'm pretty confident I can build this box, I've done a lot of other small electronics projects in the past. So I already have most of what you need on my work bench. (iron, multimeter, various solder, wire, hot glue gun, superglue, helping hands, flux, electric tape, solder sucker, etc...)

My suggestions: definitely get a couple of capacitors, so you can experiment with more or less total capacitance. Also I'm not a big fan of the alligator clips in the parts list, I would get something a bit larger with tighter jaws. Also, some people prefer to have one of the clips mounted on the box, with one clip on a wire, instead of 2 wires ... just something to keep in mind when ordering parts.

 

[MagicPanda]

Yea, I was considering getting different jaws and mounting 1 of the clips on the box.

I'm going to look around for different capacitors and get a few.

I did find this fuse: 283-2966-ND on digikey which seems to be the same other than it's time delay instead of slow.

I would think that the amps would be what matters on the fuse rather than the voltage, in which case pretty much any slow acting 2 amp fuse would work alright. Though I don't think anything would really draw that much from the battery so a fast acting fuse would probably be ok as well, the one test I read seemed to indicate it was drawing 0.3 amps under load so I'm not really sure why a slow acting fuse is used in the first place... (i'm not an electrical engineer though I just know how to build things from directions and use a multimeter for the most part )

 

[Scope666]

Yeah, for what it's worth, I've never had to replace the original fuse I put in. I think it's really only there if something goes horribly wrong, like one of your components comes loose and you get a dead short.

 

[Lentulusbatiatus]

It is much easier to use with one fixed post rather than trying to line up two on cables.
I prefer to solder the ends of the clips but there is better flat clips out there than the mini ones, should find some if you search for "fine wire croc clips"

 

[Joe Tang]

hi there,
i noobie here. need some advise for building my own welder. can i use this board? DC 3 35V to 3 5 35V Booster Step Up Converter Regulated Power Supply Volt Meter | eBay

it come with the volt meter and the input and output is more or less..

 

[yo han]

I'm also going to build one of these (with LM2577) but the information in this topic is spread out over 119 pages at the moment. Would it be possible to do a recap please?
I've been reading posts where people are replacing the capacitor with double the value of the initial specs. So 2000uF is better?
Also, I want to add a voltmeter and found this one on eBay: click It has 3 wires so it looks like it runs on external power (wouldn't drain the capacitor).
If I want a button to lower the voltage, can I just have a switched resistor over the capacitor and operate it with a second push button? (Like the Zapper v2) If it's that simple, what value resistor would I need?

 

[awsum140]

I can't believe I read the WHOLE thing!

I'm late to the party as usual and have been fooling with CD welding a little bit. One thing that hasn't been discussed is the variable that is introduced by the size and length of the leads used. Being "Mr. Overdo" I've been using "zip" cord, two 18ga stranded conductors, for all the wiring from the capacitors to the banana jacks on the box and for the lead to the moveable, positive lead, alligator clip. The negative lead has the alligator clip soldered directly to it. I can vaporize wire, 32ga or 33ga, Kanthal or copper at the energy levels being talked about here. I attribute that to the lack of loss in my setup. The losses of normal resistance seem to multiply, dramatically, when short, high current, burst energy is used rather than continuous voltage. Micro resistances can become very problematic not to mention the inductive effect when that current pulses through the lead.

I am using an LM317T adjustable regulator on a 25.2 volt transformer with a bridge rectifier so it yields about 36vdc at max setting for the 317T. After reading through this thread, almost like War and Peace, I have a whole new perspective on CD welding and a ton of tricks to try.

So, just what are you guys using for leads, gauge and length?

More importantly, what energy levels (Joules) are working best for what combinations of wire types and gauges?

 

[Berner]

Just finished building my zapper according to the Instructables article. I used the parts as listed and did not add additional capacitance. Haven't tried to weld any wire yet but this thing generates quite a spark and can weld the alligator clips together at 20 volts!

 

[Lentulusbatiatus]

For my Zappers I use one static post and one lead of 18" length, Test lead cable - Extra Flexible 55/0.1mm plain copper. I also use the same cable for internal wiring for capacitors, meter and sockets.

 

[awsum140]

I just "tweaked" my welder a bit. I filled the alligator clips with solder then used a Dremel to grind everything flush. A little cheating with that by using a cutoff wheel as I held the tension, carefully, on the jaws while grinding with the wheel between the jaws. It produced a good, continuous, flush match of the jaws.

Some results of spending an hour or so "practice" welding some 33 and 34 gauge Kanthal to 30ga copper. The copper was always in the negative post with the Kanthal in the positive lead. Easiest weld is an overlap of about 1mm, my eyes ain't what they used to be even with a binocular magnifier. I did get a few good .... welds, but have opted for the overlap since it's far easier for me to control.

33ga Kanthal to 30ga Copper, 19 volts, 4000mfd, .722 Joules for either a .... or overlap joint
34ga Kanthal to 30ga Copper, 17.5 volts, 4000mfd, .613 Joules for an overlap joint. I couldn't get a good .... weld, 34ga is pretty thin.

 

[tarnishedhalo]

Great thread everyone. It was a long read and didn't quite 'get' it until I got parts and built it myself . I used the instructable found here Small Welder for joining Nichrome and Nickel Wires.

I got it to work somewhat well with 35v and 1000uF. The welds are a bit weak using 28g nickel and 28g kanthal. The problem I'm having is the unit will weld a couple wires just fine then it stops working. After letting it rest for a while and trying again it works as before, but again only for a couple wires. I checked all the solder joints as they were made and I *think* there's no cold joints and everything is routed properly. The multimeter says it goes up to 35v but its a tiny little pop sometimes. Or pressing the switch won't weld anything at all no little pop nothing. Based on these observations does anyone know where I could start to troubleshoot?

Because the weld for 35v and 1000uF isn't very strong for me I'm ordering a couple 4700uF caps and going down to 27v as was suggested in another post. Before I do that I want to know where I went wrong with the current set up! Any help would be appreciated.

 

[awsum140]

A surge limiting resistor in series with between the power supply and the capacitor may help, sounds like overheating or overtaxing the power supply. As others have said, the 3.6 ohm resistor really isn't needed, the fuse handles that protection, but a surge resistor, to limit initial current draw, would be a good idea. It is also possible that your LM2577 module is not quite working properly, especially under the high current demand of initial charge.

1000mfd at 35 volts will yield .6125 Joules which doesn't sound anywhere near high enough to produce a good, solid, weld. I'm just guessing but for 28ga wire I'd say you're going to need .8 Jules or higher. If you added a second 1000mfd capacitor and maintained 35 volts, you'd be at 1.225 Joules which would probably vaporize the wire. You will need to experiment with voltage to develop a good weld.

 

[Mad Scientist]

A surge limiting resistor in series with between the power supply and the capacitor may help, sounds like overheating or overtaxing the power supply. As others have said, the 3.6 ohm resistor really isn't needed, the fuse handles that protection, but a surge resistor, to limit initial current draw, would be a good idea. It is also possible that your LM2577 module is not quite working properly, especially under the high current demand of initial charge.

I agree 100%. I never understood why the people who post these designs didn't include a relatively high value resistor between the supply and the cap / weld section to prevent a short on the supply and / or the supply continuing to send current through the weld, in addition to inrush issues charging a discharged cap. Do these supplies generally shut down on a short circuit or is it just sloppy design?

 

[tarnishedhalo]

Thanks for the ideas awsom and mad. Would either of you have suggestions of values for the surge limiting resistor? I can draw, follow, and calculate a circuit diagram but far as selecting components I'm clueless.

I do have a spare LM2577 so may just start from scratch tomorrow and see if I can get everything working correctly with the current (even if shoddy?) design.

 

[awsum140]

I'm using 4000mfd with a supply that goes up to about 38 volts. I have a 220 ohm resistor to control the initial surge a little and it works well for me. I'm using a simple normally open push button to "tap" the voltage up to whatever I need and can tweak it in about quarter volt increments with no problem.

To check the LM2577 a little further, see if the potentiometer provides a full range of voltage, from zero to 35. While not under load and with a modest load of, say, 250 milliamps or so. A 150 ohm resistor would be fine for that, but make sure it's rated at a couple of watts, like maybe five, or it'll get overheated very quickly.

 

[Ta2ooz]

Whew. That's a lot of info and a big post, not to mention back and forth between breaktru's site and here. I decided to stick with the camera mod, went to my local electrical supplier and picked up a bunch of this and that. Will post a pic when it's done. Thanks to everyone for all the info and schematics and ideas, awesome thread.

 

[yo han]

I'm using 4000mfd with a supply that goes up to about 38 volts. I have a 220 ohm resistor to control the initial surge a little and it works well for me. I'm using a simple normally open push button to "tap" the voltage up to whatever I need and can tweak it in about quarter volt increments with no problem.

Thanks for all the valuable info awsum.
My knowledge about electronics is pretty limited but I can read simple schematics. Since there have been many people here suggesting improvements to the initial design, would it be possible to post an updated schematic so inexperienced guys like me can use that to build our own welder? (preferably one with a voltage display)