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

[texaspyro]

You know you have a good weld when the tab leave parts of itself behind... You know your power is too high when your electrodes leave parts of themselves behind... probably unavoidable with those tabs and copper electrodes.

I have had miserable luck with tungsten electrodes. Lots of POW! and the tip blows off. I have some 3/16" moly rod coming. BTW, be careful grinding/cutting moly. It's a bit on the nasty side. Work outside, use a good nose condom...

Where did you get those tabs? They look like the ones in the Sunstone video.

With that design of tab you want to weld with the electrodes on the dimples (aka weldments). The split in the tab forces the current to flow in the tab, between the weldments.

 

[The Mighty Volt]

You know you have a good weld when the tab leave parts of itself behind... You know your power is too high when your electrodes leave parts of themselves behind... probably unavoidable with those tabs and copper electrodes.

I have had miserable luck with tungsten electrodes. Lots of POW! and the tip blows off. I have some 3/16" moly rod coming. BTW, be careful grinding/cutting moly. It's a bit on the nasty side. Work outside, use a good nose condom...

Where did you get those tabs? They look like the ones in the Sunstone video.

With that design of tab you want to weld with the electrodes on the dimples (aka weldments). The split in the tab forces the current to flow in the tab, between the weldments.

Hi Texas...I pulled those tabs off the A123 in the DeWalt packs I bought off DocBass.

I was desperate for something to practice on.
elds
One other factor: the pressure you do or don't have down on the electrodes: I got great welds when I leaned gently, and shoddy welds when I either leaned too heavily or not at all heavily on the electrodes.

It's a pain in the ass but there you go, yet another intangible for us to work on.

And yes, I heard Tungsten blows apart in these circumstances: I think I read it on one of the sites dedicated to these home-made welders {LED Hacks or such like}

Thanks again for your help- you reckon I am onto a winner here?

 

[texaspyro]

One other factor: the pressure you do or don't have down on the electrodes: I got great welds when I leaned gently, and shoddy welds when I either leaned too heavily or not at all heavily on the electrodes.

So far it's looking pretty good. I'd like to find a source of those pre-made tabs...

Weld pressure is a VERY important parameter. Too much and the weld contact resistance goes down, power into the weld decreases, temp goes down, weld fails. Too little and POW! weld blows out.

Also weld polarity is very important. Positive electrode is much hotter than the negative one. Usually it should go on the most conductive or thicker material. Go to http://www.muc.miyachi.com/ Under App Notes, Tech Briefs, Articles, select Fine Spot Resistance Welding. Read everything... then read everything again.

My welder supports the use of force sensitive resistors (like http://www.sparkfun.com/commerce/product_info.php?products_id=9375) for measuring electrode pressure. I have two ADC channels with op amp converters dedicated to reading them. Haven't written the code to make use of them yet. One idea is to specify + and - electrode pressure ranges and then automatically fire the weld when both electrodes are in range.

 

[lookingelectric]

I've tried building one of these with ~4.5F of capacitors and an 18V power supply. The problem that I have is that I seem to vaporize all of the metal (theres a big region of missing metal where I apply the electrode) instead of melting it.

Right now, I'm not trying to do battery tab attachment. Rather, I'm trying to get things right by putting two strips of metal together. Whenever I do it, though, I seem to remove the metal from the region where I've been working. Sometimes I get a tiny bit of a bond between the pieces, but the majority of the energy seems to go to vaporizing metal rather than melting it.

How can I adjust this?

 

[The Mighty Volt]

I've tried building one of these with ~4.5F of capacitors and an 18V power supply. The problem that I have is that I seem to vaporize all of the metal (theres a big region of missing metal where I apply the electrode) instead of melting it.

Right now, I'm not trying to do battery tab attachment. Rather, I'm trying to get things right by putting two strips of metal together. Whenever I do it, though, I seem to remove the metal from the region where I've been working. Sometimes I get a tiny bit of a bond between the pieces, but the majority of the energy seems to go to vaporizing metal rather than melting it.

How can I adjust this?

If your 4.5f is really 4.5f and not just nominally so, then it is possible that you have too much current.

Also, I find, at a nominal 3.0f {probably less} that 18v is simply too much.

My guess you are simply have too much juice.

Jeremy Harris recommends that 2.0f of true capacitance is in-or-around where you need to be.

 

[The Mighty Volt]

One other factor: the pressure you do or don't have down on the electrodes: I got great welds when I leaned gently, and shoddy welds when I either leaned too heavily or not at all heavily on the electrodes.

So far it's looking pretty good. I'd like to find a source of those pre-made tabs...

Weld pressure is a VERY important parameter. Too much and the weld contact resistance goes down, power into the weld decreases, temp goes down, weld fails. Too little and POW! weld blows out.

Also weld polarity is very important. Positive electrode is much hotter than the negative one. Usually it should go on the most conductive or thicker material. Go to http://www.muc.miyachi.com/ Under App Notes, Tech Briefs, Articles, select Fine Spot Resistance Welding. Read everything... then read everything again.

My welder supports the use of force sensitive resistors (like http://www.sparkfun.com/commerce/product_info.php?products_id=9375) for measuring electrode pressure. I have two ADC channels with op amp converters dedicated to reading them. Haven't written the code to make use of them yet. One idea is to specify + and - electrode pressure ranges and then automatically fire the weld when both electrodes are in range.

Thanks for the advice and the reading materials. Great to see I might finally be making some progress.

 

[texaspyro]

The problem that I have is that I seem to vaporize all of the metal (theres a big region of missing metal where I apply the electrode) instead of melting it. How can I adjust this?

The problem with an SCR welder is it dumps all the energy in the cap into the weld. The only way to control the energy is to adjust the voltage on the cap. The energy is proportional to the voltage squared... a little voltage change can make a big change in the delivered energy. You need a variable power supply.

Also note that depending upon the power supply design, you may not be able to reduce the voltage on the cap when you reduce the supply voltage. You may need to fire a dummy weld to drain the cap... or wait ages for it to self-discharge... or put a bleeder resistor across the cap (which for values that wont load your charge supply may still take ages to drain the cap)

Also, unless you have a way of testing your capacitors, you don't know what their true values are. Most car audio capacitor ratings are fraudulent. The can easily be off (low) by a factor of 4 to over 20. I have tested over a dozen brands and only three met their published capacitance specs (Monster, Rockford, and Scosche). And avoid "hybrid" capacitors that fudge their ratings by paralleling low ESR aluminum (good) caps and high ESR carbon (bad) caps. All audio caps over 1 farad that I tested were either fraudulent and/or hybrid caps.

 

[The Mighty Volt]

Texas is right.

You need to be operating from around 15-17v, provided the capacitance is already sufficient.

Even 15v might be too much if you are playing with really narrow thin stuff.

 

[texaspyro]

Even 15v might be too much if you are playing with really narrow thin stuff.

With 3-5 mil thick nickel on regular steel cased batteries, I use 12V or less. Since my welder is FET based, I can control the pulse width. I get very good welds with pulses in the couple of millisecond range (I usually operate in energy based mode where the pulse width is determined by the energy flowing out of the capacitor rather than a fixed time interval. I get excellent welds with 30-100 joules).

An SCR welder will deliver much longer pulses (maybe 30 milliseconds long, but typically less than 30% of the pulse width contains meaningful energy). My welder probably has much less ESR and internal resistance losses than the typical welder. The latest config can be downright scary in what it can deliver...

 

[texaspyro]

cell_man in China posted that the A123 pos terminal is nickel 200 alloy (what most battery tab material is), neg terminal is nickel plated steel. Cells should be quite weldable.

I got in my 3/16 moly rod. Did a couple of test welds with the 5 mil copper onto a nicad "D" cell. Seems to weld pretty well, but that copper is so soft the welds tear off rather easily (tears hole in the copper sheet, leaves a small disk of copper behind).

I used one electrode on the copper and one on the battery. I have the transient suppression diodes off my welder right now, so I don't want to crank up the volume and try some thicker copper.

 

[texaspyro]

Here is the A123 cell info from cell_man (MANY thanks!):

The welding recommendations/guidelines include and are likely required due to the quite thick nictel tabs that are generally used:

Pressure: 7kg force per electrode
Squeeze: 250ms
Weld Time: 5ms
Weld Power: 5kA
Hold: 250ms

Positive terminal is 0.2mm Ni-200 laminated to 1.0mm Aluminium
Negative terminal is 0.8mm Nickel plated steel

When soldering leads onto the cell, the terminals must be kept below 150degC and soldering must take no more than 5 seconds.

The cells vent from the positive end cap and not the negative as it would appear by looking at the negative terminal. At least 1mm gap at the cell end is recommended to allow the cells to vent. No more than 40lbs of force axially down on the terminals.

The recommended welders are Miyachi Unitech models.

 

[The Mighty Volt]

cell_man in China posted that the A123 pos terminal is nickel 200 alloy (what most battery tab material is), neg terminal is nickel plated steel. Cells should be quite weldable.

I got in my 3/16 moly rod. Did a couple of test welds with the 5 mil copper onto a nicad "D" cell. Seems to weld pretty well, but that copper is so soft the welds tear off rather easily (tears hole in the copper sheet, leaves a small disk of copper behind).

I used one electrode on the copper and one on the battery. I have the transient suppression diodes off my welder right now, so I don't want to crank up the volume and try some thicker copper.

Yep- I was so encouraged by initial tests that I took the plunge and ordered some 0.05 nickel from MTO.
I ordered both tab and roll form so I can make custom tabs as I see fit.

I was having real results with the SCR and apparently I am not the only one. Somebody on a hobby forum reported similar results.

Do you mind my asking where you got the molybdenum rods?

Thanks
MV

 

[texaspyro]

Do you mind my asking where you got the molybdenum rods?

From http://www.torreyhillstech.com via Ebay item 390108410612... $19 for a foot for polished rod. They also have it unpolished for less and in 3 foot lengths. And they ship to Euroland. They also have copper/tungsten and silver/tungsten alloy rods... aka Elkonite


I got the welder back together. I was playing with the inverted DC weld mode at high power levels and managed to blow out the TVS diodes across the probes (gotta watch that repetitive pulse rating). I was using 24V, 1.5kw diodes since I had them... that was just too much pulse to clip off. I am now using 56V bi-directional diodes. The TVS diodes are not really needed, but I'm a belts and suspenders kind of guy...

So I started playing with some other TVS configurations and put one across the FETs. I had blown up several in the past doing that (with the hand wired 1F prototype with the crappy FETs) but that was at higher power levels than what I was testing at. But the new welder has 3.5 farads of REALLY low ESR caps and FETs. It was a 24V diode and I was firing 20V pulses... the inductive kickback from the first pulse shorted and smoked it.

Because I had a the probes bolted to a 1 milliohm meter shunt at the time, the charger tried to charge that through the TVS carcass... I couldn't get the probes off in time and that wound up letting the magic smoke out of the charger fets. Normally my shorted output detector would not have let that happen, but I was testing another circuit on that pin at the time... the Welder Gods got a real laugh out of that one... Unsoldering FETs from double sided 3 oz copper is a royal PITA even with a $6000 vacuum desoldering system.

 

[texaspyro]

Yep- I was so encouraged by initial tests that I took the plunge and ordered some 0.05 nickel from MTO.
I ordered both tab and roll form so I can make custom tabs as I see fit.

Who is MTO? Do they have those nifty pre-made split tabs with the weldments?

 

[The Mighty Volt]

Hi Texas.

MTO are a company who specialize in battery rebuilding.

As a sideshow , they sell nickel tabs and nickel roll.

http://mtobattery.com/store/page7.html

Unfortunately, I see no mention of the special DeWalt type tabs.

Thanks for the heads-up on the Molybdenum!

Cheers.

 

[texaspyro]

Thanks for the info...

Forgot to mention that when I had the welder apart, I changed that discharge resistor to a 1.080 ohm, 60 watt device. You can dump 3.5 farads (at 12V) into it over half a dozen times before the temp sensor hits 120F. It also cools down much faster than the old 25 watt unit.

 

[The Mighty Volt]

Thanks for the info...

Forgot to mention that when I had the welder apart, I changed that discharge resistor to a 1.080 ohm, 60 watt device. You can dump 3.5 farads (at 12V) into it over half a dozen times before the temp sensor hits 120F. It also cools down much faster than the old 25 watt unit.

Those FET Welders seem the way to go. Next $200 I get I'll get me some plans and boards.

For the time being my SCR will do nicely for the A123 packs. In fact I may consider trying to build one with the Molybdenum and a pedal-trigger. I am nearly sure Jeremy is left handed!

 

[texaspyro]

I finally got around to testing those force sensitive resistors on the welder. Seems to work pretty well. I tested it with several scale calibration weights (from 200g to 5700g). The output is much better than expected. They say that they are not suitable for use as a scale, but I am seeing under +/- 3% error!

The main problem with those sensors is the reading is rather dependent upon the area that the weight is spread over. For best results you need to arrange it so that the pressure is transmitted to the sensor via something some like a smooth, fixed diameter metal disk covering as much of the sensor area as possible... A 1kg weight sitting on a dime reads around 20% off from one sitting a a quarter (and coins do not have a smooth surface for making contact). The sensors that I bought on Sparkfun's last sale are around 2x2 inches.

Absolute accuracy is not really an issue as long as your pressure transfer system is consistent. You want to know "I get good welds when the number is around xxx". You really don't need to know that xxx is 2250 grams. The welder supports two sensors if you can arrange it so each probe has its own sensor. Now I need to integrate the sensors into the weld firing code...

 

[texaspyro]

The megadonky LCD touchscreen controller used on my welder has just undergone rigorous live animal (for now) testing and has received CBTT certification! It is now certified Cat Barf Tolerant Technology... takes a barfin', keeps on workin'... at least when used in CO2 laser cutters...

 

[amberwolf]

Some of my technology has been tested that way by the dogs. :lol: Sometimes it's from the other end, though; I'm not sure which is more disgusting (they're equally hard to clean up).

Still, it's better than when Hachi attempts to assimilate it.

 

[The Mighty Volt]

Wow. Sorry to see that your pooch did that.

I was just wondering....in terms of using a pliers to peel a soldered tab off, how much resistance should my soldered tab be putting up?

Whilst I acknowledge that using my forearm and a pliers basically constitutes a giant lever, should my tab really be expected to stand up to being peeled back?

One thing I do notice, and I think this is good....when I peel the tab back, it puts up quite a bit of resistance, and leaves behind a little piece of tab all the time.

Is this what I should be looking for?

I will post photos tonight, for this evening I finally glued 10 A123 together and welded them in series using the MTO rolled tab and the Jeremy Harris welder with 3 Farad of capacitor.....which I found have improved hugely since I swapped over to bus-bars.

 

[texaspyro]

The more oomphala it takes to get the tab off, the better. A very good sign of a good weld is when it leaves part of itself behind. If it doesn't, the weld is not good. If it does, the weld was stronger than the tab material.

When I remove tabs from a cell, I grab the tab with needle nose pliers close to the cell and rotate the pliers so that it rolls the tab around them and peels off the cell. This minimizes the stress on the cell and maximizes the leverage getting the tab off cleanly.

This is how I would weld the tabs on using straight tab material (image from epbuddy.com):



Use two tab strips with a gap between them (probably a bit bigger than those shown so the cell vents are not totally blocked). Place one electrode on each strip. This forces the current to flow through the tab, across the cell, and out the other tab. Welding thick tab material without a notch or gap between the electrodes does not work well... too much current gets shunted.


Another useful thing to do is to emboss dimples (weldments) onto the tabs where you will place the electrodes. This focuses the weld current even more. That is what A123 does with their tabs. See http://www.sunstonespotwelders.com/video-wh-battery-welding.php for an example/explanation.

I have made nice dimpled tabs using a drill press. Tightly chuck an upside down drill bit in it (round off the edge of the bottom of the bit with a file or dremel). Put a block of hardwood on the bed of the drill press and adjust the throw of the bit to the depth needed for your dimples. Drink beer, dimple the night away...


I need to make a good cell holder for welding. Vices and wood working clamps are a no-no. Cells are slippery, wobbly little beasts and nickel strips might as well be made of greased teflon. With hand held electrodes you really need four hands and a foot to make good welds. What I REALLY need is a real weld head... that doesn't cost a few kilobucks.

 

[The Mighty Volt]

Wow. Top post there Texaspyro!! It hadn't ocurred to me to make my own dimples. As it happens I have the very thing for that- a pillar drill, I can just use the lever to raise and lower the bit at will. Should make for fast work.

The tab I got from MTO is 0.375" wide. That means I would have to cut it down the middle to make the strips.

For some reason, the individual tabs are far shinier and slippier than the roll. Don't know why that is.

I agree on the need for something to hold the cells in. I had an idea for what I call "templates", basically a block of wood routered out with a high speed router to a depth of one-third of the cell, and to a width of just 1mm short of the actual width of a cell. This way it's a tight fit and this provides the gripping action required.

Is it absolutely vital to not obscure the vent hole? I have in my first sub-pack.

Sorry for the lack of photos....as usual I am surrounded by very expensive A123 cells and would give my right arm for two AA Ni-Mh cells for a camera

Thanks for that most informative post. Tomorrow I should have some pics and updates. Cheers.

 

[texaspyro]

The tab area on the batteries is 1/2" diam. A123's tab arms are 0.2" and 0.25" wide with a 0.025" slot between them and a 0.14" hole in the slot around the vents. The tabs are 0.010" thick. The dimples look to be quite shallow, maybe 0.003"-0.005". Once the tabs are on, you can't see under them. On a couple of my cells, the welder electrode missed a dimple by a little bit.

You could split your 0.375" tabs, but would be a little better to cut wider ones from sheet.

If you never abuse the battery then vents are not needed... :wink: The Killacycle people, etc block the vents and tightly confine the ends of the cells... something A123 says is a no-no. Also something that is alluded to as a problem in the Infamous Slagged Prius report. I suspect that if the ends of the cells are not confined, the end caps would eventually blow off.

A friend made a dimpling tool out of a small block of metal with eight pieces of drill wire in it. A stud on the other end chucks it into the drill press. The wires were ground even in a lathe. Makes 8 nice even dimples in a whack.

I would go with a cell holder at least 2/3 as deep as the cell is long... my CR123A tabber is as deep as the cell. Works very well since the tab rests flush with the top of the block of wood. It is just a single cell holder. I tabbed a couple hundred cells in an hour or so... paid for the welder a few times over. They replaced a $20+ unobtainium battery with a $0.99 CR123A and diode and saved a company from scrapping about $200,000 of equipment.

Fritz's blog has a couple of photos that show some cell holders:

http://1.bp.blogspot.com/_GmuiJkAUlZw/S9iBtvnLpXI/AAAAAAAAAJQ/JolmyGnrWp0/s1600/Sid.JPG

http://4.bp.blogspot.com/_GmuiJkAUlZw/TA235mUyBaI/AAAAAAAAAK4/-ajdGNy5F4E/s1600/fuji+camers+shots+113.jpg

 

[texaspyro]

I found another use for the welder... It slices! it dices! It vaporizes! It makes Julienne fries! It annoys small furry animals! But wait! There's more! It even anodizes aluminum! I fire a 100 millisecond pulse, then recharge. At 3F and 15A you get about 1/2 volt drop over the pulse. It was just a silly beer-induced experiment that worked rather well. I need to play with different pulse lengths, etc. Maybe add an output filter cap. Should also be able to do plating with it.

My welder design has hardware to prevent turning on the charge fets and firing fets at the same time. I could connect the tank directly to to the caps and bypass the firing fets, but drawing the power off the welding electrode connections is so much easier... plus you can control switch the current on and off and control the time.

Now it's time to go and make some turtleburgers... damn they're good... mine have jalapenos... http://aht.seriouseats.com/archives/2010/02/photo-of-the-day-bacon-cheese-turtleburgers.html