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

RLT said:
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
Click to expand...

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.
Click to expand...

i've done a fair amount of experimenting with this and i have found that charging to a voltage is very inaccurate. what is much better is to charge the cap using a CC/CV power supply for a controlled time period. since the scr drains the cap completely when it fires the timer controls the amount of energy in the cap very precisely.

i use a cheap 16V 5A laptop power supply. i have a simple 555 timer circuit that i can use to control the charge time, the results are very repeatable and i get very little wear on the tips

rick
 
This seems to be the biggest DIY spot welder thread, so...stumbled across a 3-minute video of a dirt-cheap DIY spot-welder made from a junked microwave ovens transformer...seems too easy to be true. Anyone with some electrical smarts please take a quick look and tell me if this is a waste of time for welding 18650's onto metal strips?...

http://www.youtube.com/watch?v=hTaGa93lOGU

[youtube]hTaGa93lOGU[/youtube]
 
Actually, it is that easy. Just basic electric stuff. http://en.wikipedia.org/wiki/Transformer#Induction_law

Here's more detailed video. He is actually using it for welding tabs on NiCd battery cells.

[youtube]al45MXMp7e0[/youtube]

If you need more voltage, just make more turns of secondary 4 GA wire. He's got 2.73 volts on 3 turns.

BTW, spinningmagnets, I was glad to come across you on electrotransport.ru

You're brave person. If you need help with translation, let me know, my Russian is still fluent :)
 
Commercial welder. Exactly same idea, for DIY one, you just need to experiment with material for tips.

[youtube]GGTGIlT6JvM[/youtube]
 
I'm in progress of making mine. Just found microwave 1450w on scrapyard. I am going with arduino setup, double pulse with adjustable second pulse duration. Just need to order couple components that I could not find in mine electronics scrap yard :)
Hardest part would be making electrode arms to apply even pressure on the tab. Should be fun
 
I'm thinking of making boards for people for 20 unit bank of fets.. similar to my previous setup. Possibly populated, if there was enough interest. The board will be the height of one rockford cap.. and will screw on vertically. Thinking $100 for populated board sans fets. my fet of choice is the IRF3703PbF. Let me know if anyone is interested
 
I would like to thank all the folks who made this thread as informative as it is.

Im in the process of buiding a tab welder. I took the mosfet route. 44 of the irfp3710 kind, salvaged from a bunch of aps ups boxes.

I already got three capacitors. I got the dual pulse and delay all programmed into a microcontroller. Im currently in the process of constructing the gate drivers.

My question is, do i need a gate resistor for each of the 44 fets?, and will pulse width modulating the weld pulse successfully control the current, without toasting the fets?

The fellas using fets, how is your power stage hooked up? Freewheel diodes? Tvs?

Thanks
 
Assuming the same pressure is applied on the welding tips, and all other variables are the same, what would cause a large variance in the result spot weld? Most of the time, the welds are very small and don't hold. Sometimes I get a big spark and blow holes in the tabbing. What is causing this? Even the large spark welds don't hold most of the time. My setup is like the one described by the OP, on ledhacks.
 
So its not a good idea??
35F?? I read 350F...

What's the best ESR of audio caps usually used?
 
I am not experienced with spot-welding strips onto cells (I've never done it), but I'd like to share some info I have found during my research.

Some of the professional battery packs made from 18650's have a split in the center of the strip, just where the end of the cell touches it, and right where the spot-welds occur. I had my theories, but recently I had someone in the industry respond to me and he indicated that the split creates a situation where the shortest distance for the electricity to travel across the electrodes during the spot-welding process...is across the battery tip, as opposed to most of the current traveling through the strip, which would leave a weak weld. Doing this results in a more consistent quality of spot-weld.

PackBuild29.JPG


A second item I want to share is that I have come to believe that if we solder thick wires onto the part of the strip in-between two cells, the mass of the strip and the distance to the tip of the cell means that it would be very unlikely that any heat damage would occur to the electrolyte in the cell right near the tip. That potential heat-damage to the electrolyte is the only reason I have been very reluctant to trust a soldered batch of cells. I got the idea from Doc Bass, who mentioned that this is the way he solders to his reclaimed cells. He cuts them out of the cordless-drill packs in pairs, so that the stock spot-welds to pairs of cells are still intact. A "bus wire" connects the paired cell strips to the pack (of course he has use a different method on the other ends):

file.php
 
i'm not a skilled welder, so i'm stuck with soldering.
soldering is a skill too, but i have experience, so it is easier for me.
Luckily i found a 110w iron at a yard sale for $2 years ago. It is the MINIMUM wattage imo. The trick is to have an iron so big, it can do spot soldering.
so if i were to do lots of this work, i'd get as big an iron as they make, like 300w if available. don't heat much of the cell, and cool it fast, so the result is like spot welding.
Spot welding is far superior for factory mass production where time is money. But pro soldering gives the same result at a lower cost in SMALL runs.
UPDATE: for pack repair i now use JB Weld. With replacement cells, they may not have tabs, in that case i JB them, which also acts as a thermal fuse. So NO SOLDERING or spot welding needed. Takes overnight to cure, so u may not want to build a large pack with JB if you are in a hurry.
 
I like the idea of soldering to the centre of the interconnecting strips, however, any concerns about melting the cell shrink wrap directly underneath? I understand this is already a potential trouble-spot for vibration/chaffing. Perhaps those little cardboard discs would address both issues?
 
hi, i see suggestion on using varistor rather than tvs diode to protect the mosfet. I have never use a varistor before and have no idea on the connection to the mosfet. Can someone explain or draw out a simple schematic? can i use V20E17P or V14E20P or V20E23P for irfp2907? thanks in advance. I assume is connected between drain and source or between electrode?
 
texaspyro said:
hillzofvalp said:
K. Care to elaborate? What do you recommend for the ratio between pulses and the width? I'm having better luck right now at 440 microseconds first pulse and 2000 microseconds second pulse. You want the first pulse to be say around 800 microseconds and the second pulse to be like 3000 microseconds?

BTW I just obtained a 20V 1F monster cap to try.. didn't you say these were the best?
Click to expand...

I seldom weld by time anymore. My welder lets you specify the energy to put into the pulses. This helps compensate for things like inconsistent connections. As a general rule you start doing single pulses of increasing energy until the material starts sticking and then back off some. Set the delay to the main pulse to the length of the first pulse, and the time/energy of the main pulse to 4-5 times the first pulse.

Monster Cable and Rockford 1F caps work very well. Avoid the Rockford 2F and larger caps... they are "hybrid" caps that don't work for welding.
Click to expand...

I just rediscovered this post. I am going to try to increase the power a bit more. splattttt

I'm a bit skeptical of a 440microsecond first pulse being short. I have been using 170 now. I verified with a scope, and the material seems to stick right aorund 250. Maybe yours are longer because you are regulating the pulse with a controlled curve.

I've used shorter connections and twice the amount of fets. I now use a 1mS main pulse on the negatives. Positives I'm still learning. I get some splatter still even with a nice head. WHY!? Let's assume it's not the electrodes. I'm running 5.5-6.5 setting on the pressure knobs. (18650s, 1/8" electrodes with like .04-.05" tips)

I couldn't find the post on ES regarding weld pressure on positive vs negative electrode. I'm going through about 1/16" of electrode length for every 1000 welds or so.

Advice appreciated
 
spinningmagnets said:
I am not experienced with spot-welding strips onto cells (I've never done it), but I'd like to share some info I have found during my research.

Some of the professional battery packs made from 18650's have a split in the center of the strip, just where the end of the cell touches it, and right where the spot-welds occur. I had my theories, but recently I had someone in the industry respond to me and he indicated that the split creates a situation where the shortest distance for the electricity to travel across the electrodes during the spot-welding process...is across the battery tip, as opposed to most of the current traveling through the strip, which would leave a weak weld. Doing this results in a more consistent quality of spot-weld.

PackBuild29.JPG
Click to expand...

Very basic questions:

1. When you say "Some of the professional battery packs made from 18650's have a split in the center of the strip, just where the end of the cell touches it", you are referring to the empty semi-circles, right?
2. I assume (but I am not 100% sure) that the metallic tabs are spot welded to the cells (the four small dots in each cell). But then, it is necessary to have a nickel wire linking all the cell pairs (those united under a single tab). And from your comment "and right where the spot-welds occur", I assume this second spot weld is done just next to the semi-circle. Is this interpretation correct?
 
I wonder where are actually RLT and Jeremy Harris ?

I hope no one of them is in prison or hospital , and 100% still operative.

RLT , the creator of this thread made the simplicity of his design an art. He could make thousands of spots welds in batteries and tabs were strongly holded. He opened a path and I admire him for that.

Jeremy Harris , improved the path with a simple schematic that also worked , and showed how generous an old british guy can be.

Unfortunatly , none of them could avoid exceeding the 100$ limit for their bill of materials , but they were not too far away above it.

Audio Car Capacitors and SCRs were their weapons.

nemo brougth mosfet into battle , but perhaps he sunk his Nautilus with fet explosions :)

texaspyro seems to have dominated the mosfet , but the overall cost of his fabulous welder is VERY far above 100$.

At this point I remember a sentence of a teacher I had -> "the best engineer is not the one who can design a device to produce an action -> the best ingeneer is the one who can make THE CHEAPEST device able to make the action"

I remember a famous program in US tv where they take a millionaire and they make him annonimous for a week , where he has to live that week with 75$ of limited money.

I wonder if texaspyro was put in a reality program with 100$ money EXCLUSIVE , and producers checked if he was able to create a valid TAB battery spot welder with that $ limitation. I think reality program would be very interesting to see.

At this point , any there has a basic bill of materials to create the cheap 100$ battery spot welder ?

I have an idea , but I didn't study in Harward.
 
But to be honest with RLT , I think in USA following his design the objective of not passing the $100 limit is NOW possible (at least in 2014).

Although he choosed the BUY NEW option for the critical expensive part (car audio capacitor) , the BUY SURPLUS option is still there.

And following expert Texaspyro sentences , instead of buying an exagerated 5F overdescribed capacitor , it would be better to buy 3 Capacitors of almost real 500.000uF and putting them in parallel.

This way overall ESR would be 1/3 of the single 0.5F and overall capacity 1.5F.

Recently I have seen some of this surplus capacitors in eBay around 16$ per piece:

http://www.ebay.com/itm/331370112139

So in USA a poor home builder I think could actually build the welder with just $100 and RLT design.

RLT should be mentioned now in 2014 for a honour medal in USA , because he made cheap DIY tab battery welding possible.

My old teacher would even say in fact that RLT is better engineer than Texaspyro because he made THE CHEAPER welder able to weld a Ni TAB to a 18650 battery.

Jeremy Harris , you have a lot of merit too , but not doing schematics :D , please take a look at this to argue my idea of a good one (about a related subject -> graphics talk):

http://www.lucianomarroccu.com/app/download/4313506451/ESR.pdf?t=1307722460

Watching that wonder makes sense the sentence "only a page of Nouva Elettronica magazine is better than a full year of Elektor's magazine"

But Jeremy , perfection is hard , intermediate level of schematic is this one (with some personal dedicated contribution) :



In spite of all Jeremy your schematic is provably better than the ones RLT still has in his web after 5 years passed (more revenges than schematics should be called :oops: ) . Any way bronce medal for him instead of gold for such a rudimentary graphical explanation of such a fantastic solution.

But RLT , sorry fot speaking with my heart , I saw a post in this almost 1000 with such a catastrophic shame schematic :oops: :oops: :oops: as to put in prison to the author -> or make him a politician in my african country :lol: -> look "this" :

http://endless-sphere.com/forums/viewtopic.php?f=2&t=2633&start=850#p511125

Turning back to serious mode Jeremy , I saw an important missing wire dot in the one you posted here several times , and to give colour and life to it I promise to post a colour "doted" version to help novices to build the FINAL 100USD CHEAP SURPLUS PARTS DIY TAB WELDER VERSION reached almost after 1000 posts.

Unfortunatly outside USA shipping cost would break limit again and finding surplus parts of this type of caps is difficult , so there is still a path to investigate.
 
A promise is a promise.

Here is the coloured schematic I evolved from Jeremy Harris design.

Please Jeremy don't ask me for copyrights because I am not winnig a penny with this :? , it is just an adition of colour and some slight changes , including the estimated cost per part to reach the 100$ without passing the barrier.

The overall big parts cost reaches 90$ , with still 10 bucks for Cu wire conductors and miniparts.

Here is the result :



Undoubdtedly , something virtually any DIY enthusiast can build.
 
Thank you Ricbauer for the kind words.

Mostly I get told that I'm full of crap and no way my simple design would work.

Too bad that surplus 'real' capacitors have all but disappeared in the world. The first time I tried to build a CD welder I could get 65,000 mF 35 or 40V Mallory or Sprauge capacitors for $3.00 each at a salvage place in Arizona.

Well, good luck with your project.
 
Welcome back to your topic RLT , it wasn't the same without your presence.

Please don't feel disturbed with my previous critic to your schematics , provably you hate them as I hate doing flow diagrams of a new self created PC program ( I know they make the program easy to understand after 6 months , but STILL HATE THEM ). Any way , even without a flow diagram a program can be excelent.

Old genious , from previous paragraph I hope you understand the difference -> if you builded the circuit for your exclusive personal use , you can use the schematic you better like or a NULL one , BUT if you share it to be available to any virtual home builder , the better the schematic , the easier work for him to build . I promise to make a flow chart of every program I share from this moment (-> I will share next in 2040 :D )

Your creation HAS BEEN PROBED AS EXCELENT by the REAL WELDING PRACTICE , and that makes the critic of many ignorants a VOID critic. Just remember 2 sentences of 2 great men:
+ Don't put me in front 1000 men saying I'm wrong , put me just 1 who demonstrates I'm wrong.
+ Those who can do , those who can't teach.

I regret in USA in 2014 it can still be difficult to find real good surplus audio capacitors.

In Europe it really is even more difficult , so I will not be able to personally test black posted schematic because of the unavailability of surplus car audio capacitors.

So your sentence has pushed me to mention now 2 ideas I had :

1 -> Replace the audio car capacitors with a bank of NEW real electrolitic capacitors in parallel.
2 -> Replace the audio car capacitors with 2 surplus car batteries in parallel.

My idea 1 is not extrictly new in this topic , because Jeremy Harris has probed it with success using 97.000uF real electrolitic capacitors -> 15 in parallel.

The problem with this particular model is that is almost more difficult to find than audio capacitors , and if found 15 pieces are more expensive than a new audio capacitor.

My implementation of idea 1 would be going smaller in size and higher in number.

I found the practical rated voltage 25V to keep price low.

Only 1 data left to choose the small real electrolitic capacitor -> we have to achieve 1.500.000 uF

I found that between 2200uf and 6800uF the 1.500.000 uF could be got under $100 of capacitors cost (shipping cost included worldwide).

The good news with this small NEW caps are :
+ You can buy them from China by eBay or AliExpress in packs of 100pcs or more worldwide shipping included.
+ Although the ESR of a single one is not as low as TexasPyro has recommended , if you put them in parallel , say 300 , the ESR final is divided between the number , say 300 , resulting an equivalent ESR close to the 0.001 ohm of TexasPyro.
+ If you need to build a bank of 300 in parallel you can divide the bank in 10 rows of 30 , to make the bank shape and size similar to the space occupied by a huge car audio capacitor.
+ As Jeremy Harris probed , this bank could also be fired with a single Thyristor , but with this NEW rows configuration you could even use 2 Thyristors this way:
-> + Thyristor 1 to fire 2 rows for cleaning shot
-> + Thyristor 2 to fire 8 rows for welding shot in delayed pulse.

This caps are not expensive in 2014 if bought NEW from China.

In fact I have some hundreds in bound to my african country at the moment.

So let us leave idea 1 in transit at this moment too and pass to idea 2.

For implementing idea 2 the first important thing to mention is -> continous permanent power here , so in schematic thyristor should be replaced by a device able to cut the current flow after the welding pulse time -> DON'T TRY TO USE MY PREVIOUS POSTED SCHEMATIC WITH 2 CAR BATTERIES IN PARALLEL !!!

But before implementing deeper the idea 2 , I think the critical surplus part , the car battery should be discussed deeper , and the first and most vital factor now is THE PRICE.

So RLT , if you have read till here , I need you for idea 2 , please .

I can accept from your experience that in USA it is difficult to find good surplus audio capacitors.

But my friend , don't tell me it is also difficult fo find a surplus car battery in a car scrapyard.

In my african country a 12V 65AH lead acid used car battery in good working order can be found between 20€ and 25€ (this is 22$ to 28$)

Please RLT , my question is :

How much does it cost in a car scrapyard in USA a typical 12V 65AmpHour surplus car battery in good working order?

I need your answer :)
 
You must log in or register to reply here.
Back
Top