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

oh and do you think that you would be able to use this cap with a ac transformer that can dub down to 24 volts ac insteed of a dc one?
YodaKrawler said:
oh and do you think that you would be able to use this cap with a ac transformer that can dub down to 24 volts ac insteed of a dc one?


Unless you are using a much higher voltage capacitor than one of the car audio types, you will blow it up on AC or DC at 24V.

Use DC only (or at least put a diode or better, a bridge rectifier between your transformer and the Cap.), Observe proper polarity, and stay well under the capacitor's rated voltage.
thanks i was just wondering would never try any thing like that with out consoltiong people that know first. thanks
i got it today and the number on it is

T6001215 57 2095 8635

1200 volt 150 amp.

it does have a red lead that goes to the main big wire and then a white wire that goes to the white cap. so will this alow me to use the red and the white lead wire to activate the switch. thanks for any info or any site that have the info. i have tried to look.


oh and i found this about it
I was hoping Fechter would reply before me, he knows his electronics a whole lot better than I do;

Until someone smarter chimes in, here is what I'd try:
Looks like the gate operates at 3 volts instead of the 1.5 that all of mine do, so you are going to need to use two alkaline (probably AA for simplicity) batteries in series instead of one (Or you could use a single 123 or CR5 lithium camera/flashlight battery) between the switch and the SCR gate.

Hook the positive from the battery to the white wire , have the switch anywhere in the circuit, then hook the red wire to the negative end of your battery(ies).

If you have a multi-meter, put one lead on one end of the SCR, and the other on the other end, activate the gate, and see if if the resistance goes from 'infinity' to close to zero. If it does, you got it right the first try.

If you don't have a multi-meter (and you really should if you are going to be messing with this stuff), I'll tell you another way to test with low power, if you ask.

If it doesn't go to zero, then reverse the polarity of the battery(ies) and try again. As long as you don't apply more than 3 volts to the gate, you won't hurt it if the polarity is wrong.
thanks for the info and yes i have a few volt meters. i use then for my toyota and the garage doors operators that i fix. so then the two leads the red and the white i can controll the gate from that is cool. i was wondering cause in the plans it shows the wire on the opposite end of the scr. and not the negtive power side. the nut on this things i hugh i think that the shaft diameter in like on inch. do you think that i can drill the bolt end and tap it to attach my leads? thanks again guys for helping a gadget noob. i am working to make a ballon popper laser to. later
i was thinking or using a box to put everything in so i can make it a little easier to put away or move and then connect my switch to my leads so i can activate by hand. do you think this will be okay?
YodaKrawler said: you think that i can drill the bolt end and tap it to attach my leads?
Yes, that's what I did on several of mine. Makes things a lot easier to hook up. Just don't go too deep and drill into the semiconductor stuff. Most of that 'bolt' is solid copper, so you should have plenty of room to work with.
I was thinking or using a box to put everything in so i can make it a little easier to put away or move .....
Yes indeed. Much safer that way too. You may need to put a cooling fan inside the box if you will be doing a lot of welds in a short period of time., especially if your power supply is going inside too.

I'll probably put my CD welder AND the mini AC spot welder from an old microwave oven transformer, inside an old computer mini-tower box I have laying around when I get through experimenting. (plus a high current 5VDC power supply, a small variac and some other stuff, so it will be ready for all kinds of projects, repairs and testing.)

.......then connect my switch to my leads so i can activate by hand. do you think this will be okay?......
As far as the switch.... I thought about doing it that way, but unless you make a fancy handpiece (I plan to talk about different handpiece designs once I get some more experimenting done, and have time to write it up.) and mount the switch on that; PLUS, have a jig for your batteries and an extra finger (mechanical or human) to hold down your nickel strip prior to the first weld, I think you will probably find the foot switch to be worth the extra hassle and wiring.

Also, there might be a problem with induction causing false triggering or too much reverse voltage and/or current, damaging your SCR gate if you mount your switch and run the control wires along with your welding cables. I don't know that for a fact; but there IS so much current flooding through your wires during the welding pulse that the welding cables will 'jump', and so will any magnetic materials within a few inches.... So the possibility does exist
Once the SCR turns on, nothing is going to turn it off until the cap discharges. That one is a real honker! 3,000 amps peak - should work great.

Below is a schematic of what I was thinking.

The switch could be a relay.

When you fire, the power supply is disconnected. As soon as you let off the foot pedal, the caps get recharged.

There might be a way to re-arrange the parts to make it more idiot proof, but I think it should work OK like this. If you let off the pedal before the main caps discharge, the SCR could still be on. I think you'd have to be pretty fast on the pedal to do this, but it could happen.

Depending on the power supply, it might be necessary to fuse the connection and/or put a resistor in line with the power supply. A current limited power supply would be best. An inductor in series with the power supply may be needed if the supply has its own capacitor. If you use a resistor, you won't need the inductor.
You could simply use a bunch of wire coiled up to do both.

The idea is to keep the switch contacts from welding and getting stuck.


  • CD welder.jpg
    CD welder.jpg
    12.4 KB · Views: 8,602
Hmm... That seems like kind of an atypical approach. Or more likely, MY approach was the odd one.

It sure would simplify things, and cut out the need for a separate power source to activate the gate.

I wish I could have opened up one of those 'industrial-strength' Unitek CD welders when I was around them bac in the early 90s, to see how they do things, rather than just peeking in through the ventilation holes and trying to guess.... Or even one of the Sunstone engineering ones...

... Actually, I just found one of a different brand:
Much more complex though... and they use higher voltage (60-150V) and much lower capacity Caps.

Oh, and here is some info on the Unitek:
They have very small capacity Caps too.

Guess there is lots more to be learned on this subject than I thought.
Hmm... I have no clue what a typical approach is. I just made that up. It's pretty much the simplest thing I could think of that would do all the required things. I'd still like to add something to prevent the caps from getting recharged if the SCR is still on. If the switch was a relay, it could have a time delay.

Most of the monster SCRs we find surplus can handle much more than 12v.

Since the energy in a capacitor is 1/2 C V^2, going to a higher voltage gives a lot more energy.

I don't know how the voltage affects weld quality.

If you look at the available capacitors in range of 15 to 60v, it may be more cost effective to go up a bit in voltage. I'd shy away from anything over 48v just for electrical safety. Ideally, I'd suggest finding a source of surplus capacitors to keep cost down.

The Sunstone dual pulse thing was interesting. With two banks of caps and two SCR's, it wouldn't be to hard to rig up a sequential trigger. I guess the first bank could be a lot smaller.

I'd like to see a pic of your electrode setup. That's the part I'm really clueless about.
fechter said:
If you look at the available capacitors in range of 15 to 60v, it may be more cost effective to go up a bit in voltage. I'd shy away from anything over 48v just for electrical safety. Ideally, I'd suggest finding a source of surplus capacitors to keep cost down.
I found twelve nice Mallory 36,000uF @ 35V (45V surge) ones on eBay last night at a decent price, and bought them. Plan to try that out and see what happens. That will give me nearly half a Farad and hopefully optimum ESR. The one I built ages ago ran at 40V if memory serves and had a lot more capacity, but it had other design errors, so it's mediocre (at best) performance doesn't give me anything to judge by now.

Plus I FINALLY got my '5F' car audio cap in today, that may actually be capable to 20V /24V surge. (no idea if the E.S.R will be good though).

Will experiment with that late tonight and report.
The Sunstone dual pulse thing was interesting. With two banks of caps and two SCR's, it wouldn't be to hard to rig up a sequential trigger. I guess the first bank could be a lot smaller.
I was 'noodling' that idea around myself last night. four low tech possibilities I could think of, using two Caps or cap banks and two SCRs:

A: 2 stage mechanical switch manipulated somehow to give a very fraction second delay (or three stage, using your idea to cut off the power to the caps just before and during the weld pulse).

B: a Delay Line between the gates of two separate SCRs on separate caps. (But I have no idea of how to wire a Delay Line, whether it requires additional components / circuits or what).

C: A few wraps of of wire around the welding cable ought to have enough induction to trigger gate 2 with a few thousandths of a second delay.

D: Starting to get a little high tech and adding a bit of programmability: A Timer relay like:

I'd like to see a pic of your electrode setup. That's the part I'm really clueless about.
Very simple right now. Just some #4 copper ground wire, with a conical point, and some insulation.
I have fiddled with some other ideas, but haven't really made anything worth showing yet. I need some time down in the workshop, that I just haven't been able to do.[/quote]

(Edited to update some of the numbers.... DOH!!!!)
The two stage thing could be triggered by a 555 timer chip. When the first pulse is triggered by the pedal, it starts the timer. When the timing cycle competes, it triggers the second one. The time delay could be adjusted to any value.

An even simpler approach might be done with just a resistor and capacitor.
The gate on the second SCR would have a capacitor across it, and fed through a resistor from the first gate. When the first one triggers, the capacitor will slow down the voltage rise on the second one.

I suspect you only need a few milliseconds of delay.
Had been to the hobbyspot welder site before, but somehow missed that page.
I found this to be of particular interest; explains something that has been causing me an occasional problem:
Uneven or inadequate pressure can damage electrodes and leave holes in the work piece.

I got to experiment with the new capacitor late last night. It is a Power Acoustik 5Farad ( which I suspect is really somewhere between 2 and 3 Farad)

That place seems the cheapest but their customer service is lousy, so I can't fully recommend them. And definitely stay away from their Pyramid brand Caps.

But this Power Acoustic one is decent for welding. somewhat better than the Volfenhag 1.5F I have been using. I'll do a more detailed report later.
I built one of these a while back. I use it more than I ever realized I would... mostly at work... we spot-weld heat-seal ribbons together... put connection tabs on the ribbons.

With a 24 volt transformer and enough capacitance you have plenty of power for welding tabs. Mine is very rudimentary and I have always wanted to make it better. I basically use a meter across the caps to let me know what the charge is and then disconnect the supply. It serves me but when you have a lot of welds to do it would be so much nicer to have it charge back up unattended.

I'm no electrical genius but I've thought about using the a/d channel on a small pic processor to do this for me but I think it could probably be done a whole lot simpler with a small opamp circuit comparing the charge voltage with a potentiometer setting. You don't really need to know the voltage... you just need to find the pot setting that allows you to make descent welds.

I like the switch arrangement in the earlier diagram. I agree that the circuit needs a resistor to protect the switch contacts but even more so to protect the supply transformer. Those caps look like a dead short when the transformer is innitially connected. I blew a small variac I was using in conjunction with a 24v transformer because of this. "Watch this folks... Boom!"

My resistor is just a coil of stainless wire inbetwen the bridge rectifier and the caps. It works fine.

Still... what I have works great! I'll try to get some pictures of it and post.

I'm a newbie here... this is my first post. My second will be a bike pusher I am currently building. Wish me luck!

Love the forums!

Welcome Ross.

If the power supply has an adjustable voltage regulator and a current limiter, it should solve both problems. These are not too expensive from surplus places.

I'd love to see a pic of the electrode setup.

As I mentioned before, I have upgraded several of the components and refined the design of the original subject of this thread.

It will cost a little more to build, but it is somewhat neater and works a LOT better.

Rather than adding all the info to this thread and making it take forever to load due to the photos, I put all the info up on a new page at my website; ii.htm

Be glad to answer any questions I can here though.
Thank you Fetcher,

Yep that'd do it alright. I'll post pics soon. Forgot to bring in my camera. Will try to do it tomorrow.

This video shows how to tab weld A123

From the video I suspect that it is really not aluminium-cu weld but nickel-cu weld. On A123 there is very thin sheet of something round soldered with regular solder and the original nickel strips are welded to this metal rather than directly to aluminium body.
Why not to reuse it too!
I love this video and will try to imitate this kind of welding on A123 with my future home tab welder that will be inspired by this forum.

I'm thinking about 16V, 8 FARAD cap welder setup.

Notice the width of the cu strip on video.. I think that at the end it will be bended in half alongside upwards to provide even more curent handling capacity.
Is there any way anyone has tried to modify a welder like this one:

to work on a battery? Is there any hope for this to be possible?
This looks dangerous! I doubt that you can regulate output down so much.
On the other hand I like replaceable coper tips.

Since I'm working on 8 FARAD version of the CD welder inspired by this thread I need a small help.
I'm thinking of getting KOLE 8 FARAD audio cap like one here:
Price seems to be favorable. DO You think that inside there are good q. caps? Carbon-foil? I searched on the internet for an hour and found no data. nothing. Even a maker doesn't give any clues. Only the low ESR and no value, no type of caps.
I don't want to get burned and purchase something not suitable.
Also there are Hybrid 20 and 25F caps. What hybrid means I don't know and do we want this? Hope it is not some cheap electronic trick how to get good (high)FARAD numbers and not really that many farads in actual caps.
All I want is to tab weld my 80 A123 to a coper :)
One of the factory built ones I was looking at only had 2 Farads.
When making a weld, the current will need to be near 1000 amps, so heavy wire is needed and the terminals on the caps need to be very heavy. The ESR is very important.
In general, several smaller caps in parallel will work better than one big one. You get a lower ESR this way.
I got my higher voltage welder put together last night... twelve 36,000 uF 35V (45V surge) Mallory high quality electronics capacitors. connected in parallel to give 432,000uF.... pretty close to half a farad.

Using a 30V - 3A bench power supply to charge them. Other than that, same basic setup as before.

Still have a lot of experimentation to do to figure out what will work best, but these are my initial findings:

1: The difference in the ESR between 'real' capacitors and even the better quality audio ones is very noticeable Some of the audio capacitors are really lousy, and since you can't trust the marketing hype or even the manufacturer's specifications on the car Caps, getting a good one is a gamble.

3: As expected, higher volts allows you to use much less capacitance;

3: 30 volts with all 12 caps is too 'hot' for battery tab welding. I burned through several nickel tabs and a couple of battery cases. Almost as bad as the M.O.T. AC spot welder, although it doesn't heat up the battery internals as much. I'm going to try it with fewer capacitors in the near future, to see if lower capacitance at higher voltage will give better results.

4: At 20-22 volts this setup produces battery tab welds about equal to the original '1.5F' car cap at 14V. Not quite as good as the '5F' car cap at 14V.

5: At these higher voltages you have to use more pressure on the electrodes and keep it pretty equal. Light or uneven pressure causes arcs, sparks and burn throughs. Several times I ended up with about 1/32 - 1/16 th of an inch of my electrodes burned off and stuck to whatever I was trying to weld. Maybe thick tungsten electrodes would do better.... (Although the thinner ones I used on the welder I built back in the early 90s were not very satisfactory.) OR I have seen some copper / graphite electrodes somewhere before... That might do if it doesn't introduce too much resistance.

6: Still can't weld to aluminum with any acceptable results. :(

7: There was a seven, but it slipped my mind :oops:

So far, I am happiest with the '5F' capacitor at 14V. and even the 1.5F at 12-14V is better than I have been able to do with this higher voltage setup.

I can see that using 'real' capacitors at higher volts has a lot of potential (literally and figuratively ;) ) but it will take a more refined technique and/or more electronic controls to produce consistent, safe welds.

If I do decide to put much more 'serious' work into this particular effort, the 3A PS is going to have to be replaced with something with higher amperage. Either a rewound microwave oven transformer (if I can discover an 'ideal' single voltage) or spend the big bucks on a more powerfull bench supply..... or something.

However, as long as you can find a car audio capacitor with decent ESR you should be able to produce excellent welds with much less effort with the lower voltage stuff as described previously..