Endless-sphere.com

Been promising this for a couple of weeks. Finally got it written up and put on one of my own sites. Very extensive description and photos of a very crude, bare-bones , yet very functional Capacitive Discharge welder for welding tabs for homebrew battery packs.

You can build your own for very close to $100, even if you have to purchase everything for it. Most of us reading this are probably enough of a 'packrat' and/or 'scrounger' to be able to build it without having to spend much money at all.

It will be improved and refined over the next few weeks, and of course by spending a little more money we can make it more powerful, efficient, prettier and safer.

I'll try to excerpt the article suitable for posting here on the forum late tonight, and try to answer any questions that may come up. But I gotta get some sleep now.

So, I'll leave you with a teaser photo:


and a link to the page on my LEDHACKS.COM site with the initial writeup on it.
http://www.ledhacks.com/power/battery_tab_welder.htm

I haven't even fully proofread it yet, so there are bound to be errors and omissions, so if you see anything that doesn't seem right, let me know.

More to come!
(Thanks, Eric, for letting me hijack your Project #2 thread occasionally before. I'll try not to do that anymore )

That's a good news!

Since i have 320 Kinion cells to link in 20s 16p.. That should be helpfull !

I already have all the stuff to built the spotwelder.. i will try that for sure!

Doc

Hey doc.. got extra capacitors handy ?

Thanks Doc.

Oh, one important thing I forgot to mention:

The way it is set up now, the SCR will keep draining power after the pulse, even if you release the foot switch. So you need to break contact of one of the electrodes very quickly after the pulse.

Not a real big problem, just something to be aware of. Might be kind of hard on your power supply if you maintain full contact for more than a few seconds.

Wouldn't be a problem if we were dealing with AC, but under DC, the SCR works like a latching relay.

Anyone got any ideas for a cheap simple circuit that will fix this problem? I have a couple of ideas for getting around it, but they either complicate the manual process or cost a fairly significant amount of money (compared to the cheapness of the components for rest of the basic project.)

Ypedal wrote:Hey doc.. got extra capacitors handy ?

Hey, yeah, I'll bet a flux capacitor at 1.21 gigawatts would work great.

How 'bout sharing some, Doc.?

RLT wrote:
Anyone got any ideas for a cheap simple circuit that will fix this problem? I have a couple of ideas for getting around it, but they either complicate the manual process or cost a fairly significant amount of money (compared to the cheapness of the components for rest of the basic project.)

In the "old days" dc-ac inverters used SCRs. There was some kind of commutation circuit that used an inductor to give a short reverse current pulse to turn off the SCR. I'll have to dig to see how that one worked.

What about just charging a capacitor bank to the desired voltage, then disconnecting the charging supply. Fire the caps through a SCR and let the caps drain completely.

RLT wrote:

Ypedal wrote:Hey doc.. got extra capacitors handy ?

Hey, yeah, I'll bet a flux capacitor at 1.21 gigawatts would work great.

How 'bout sharing some, Doc.?

I tried to sale the last one i had on ebay last week, but the military guy knoked at my door and told me to destroy its because previously they detected some large EMP pulse around my city caused by my previous experiments with it that caused some crash in their...radar tracking system...


Sorry !

Doc

What about just charging a capacitor bank to the desired voltage, then disconnecting the charging supply. Fire the caps through a SCR and let the caps drain completely

Yeah, that's one of my less than elegant solutions. I thought of using two foot switches, one to power the charger... Or maybe rig some sort of two stage single pole double throw switch. in one foot control.

Gotta be a better (but still cheap and easy) way.

Doctorbass wrote:I tried to sale the last one i had on ebay last week, but the military guy knoked at my door and told me to destroy its because previously they detected some large EMP pulse around my city caused by my previous experiments with it that caused some crash in their...radar tracking system...

I guess you better not show them your rail-gun.

TylerDurden wrote:

Doctorbass wrote:I tried to sale the last one i had on ebay last week, but the military guy knoked at my door and told me to destroy its because previously they detected some large EMP pulse around my city caused by my previous experiments with it that caused some crash in their...radar tracking system...

I guess you better not show them your rail-gun.

Doc

A few new things, not unexected, but confirmed after more experimenting;

The 8ga wire makes a noticable difference over 10ga. And the super flexible stuff makes manipulation of the electrodes so much easier and more pleasant.

The more amps your SCR can handle, the better. The lot of new SCRs I bought on eBay came in. There were four in there that the seller thought were probably 225A.... I'm guessing those huge things are at least 300-400A, but I can't find any info on the internet using the parts number to confirm. But they make a noticeable improvement in the welds over the original 130A SCR.

Also tried paralleling a couple of 110A SCRs. That works very well too; So, if you have a few lower amperage ones laying around, rather than searching for a reasonably priced big one, you can probably get by by paralleling them. (Although I think ...Guess...that it is probably important that they be all the same part number, or at least the same specifications in all respects.

Mounting / connecting the SCR on the capacitor rather than on the electrode end of the wire is better.

Drilling and tapping a 1/4in bolt hole in the end of the stud, if you are using a large stud mount SCR, makes it much easier, neater (and probably more electrically efficient) to make connections.

With all of the above improvements I was able to tack weld two pieces of 1/4 inch thick titanium together, with just the 1.5F cap with 14.2V charge. It wasn't a 'structural' weld by any means, but it was strong enough to survive modest handling and a slight amount of torque. Two 1/8 " thick steel plates tack welded together much better than I would have imagined.

Battery tab welds were just about perfect, (at least by MY criteria).

Along the lines of Fechter's comment, and one of my earlier 'crude' ideas to overcome to the 'latching' action of the SCR under DC:
I think the problem might possibly resolve itself, although I haven't actually tried it yet:
My commercially made foot switch has a SPDT switch, with the normal position being N.C. and the actuated position being N.O....
If I am figuring this right;... If I wire one side of the wiring between the of the PS and the capacitor (probably have to use a relay or solid state relay switch) to the NC pole of the foot switch, and (of course) wire the SCR gate actuator part to the NO side of the foot switch, the cap will charge normally until I press the foot switch, then cut off a few milliseconds before the gate is actuated, then resume charging the cap when I release the foot switch.

I guess it depends on whether or not the cap retains enough residual voltage after the pulse to keep the SCR latched.


I also learned that I can't quickly draw a simple circuit diagram that wouldn't be an embarrassment to make public. I'll have to put that off until I can spend some time on it.

You could have a relay that disconnects the charger from the capacitor bank at the same time it triggers the SCR. When you let off the footpedal, the relay reconnects the charger (through a resistor maybe).

Putting some inductance in series with the discharge circuit might help commutate the SCR. A few turns of your 8ga wire in a coil.

I have one of those big honking SCRs around somewhere. They are fairly common on eBay.

fechter wrote:....
Putting some inductance in series with the discharge circuit might help commutate the SCR. A few turns of your 8ga wire in a coil.
...

Wouldn't that rob some of the welding power? maybe even a significant amount?

I've seen those reactance coils for welders made out of high frequency aircraft generators
(like this:

which I assume is sort of the same idea as you mention.

I dunno.... We are getting into stuff that is a step or two above my brain being able to grasp easily.

RLT wrote:
Wouldn't that rob some of the welding power? maybe even a significant amount?

Good point. From what I've read, you need to get the wiring resistance as low as possible.

Is there any data on what the optimum starting voltage for the caps is?

Seems like you could use a higher voltage and a smaller capacitance to get the same energy. Too far toward either extreme will run into problems. I just wonder what the best range is.

You could use one of those solid state relays on the input to the charger to kill the output when you discharge so the SCR can unlatch. With the right timing, you'd be able to zap a weld and recharge almost immediately depending on the current output of the charger.

Thanks RLT:

That's really excellent information. It would be great if someone made this setup and helped others with their batteries. I'd do it myself except I don't have batteries to weld now. However, I do have a good power supply.