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

So how do you know what sort of weld energys and delays to start with for any given combination of metal, thickness, etc. Or is it a case of being a "proper" welder to know ?
 
You just have to test it on the materials that you are welding. Experience helps...

I have a weld test function on my welder. You set the desired starting energy level and make a weld. If the weld sticks, you're done and it backs off the setting. Otherwise it increases the energy setting and you make another weld. Lather, rinse, repeat.

That sets the first pulse energy. Typically the second pulse is done at around 4-5 times that level. Try delay time values 1-2 times the first pulse width. There is very little literature that talks about a good pulse delay. It should be low enough that the oxides that you blew away with the first pulse don't have time to reform before the second pulse, but long enough for the probe pressure to cause the materials to seat themselves together.

My welder has an auto mode where you only need to set the first pulse. It uses those rules of thumb for the delay and second pulse.
 
When you say molybdenum, is that an alloy of, or 100% rod? I saw some nice thicker rods of 100% on ebay that would fit my application very well.

Another question for now, is if anyone has tried on any welds on silver, sterling in particular.

I am guessing that I will do an inital build with scr's. Then go over to fet control when I have time and understanding.

Thanks again for all the great information and opinion.
 
As always Texas words worth a fortune. Would you also care to share your weld material - to - weld probe material rundown ?

Cam.
 
Thanks Texas. I have all parts ordered and I will be starting my build at the start of next month. I will be sure to keep the forum updated on my progress.
 
You rock Pyro.

I just made my 100v spot welder.
 
diaz said:
As always Texas words worth a fortune. Would you also care to share your weld material - to - weld probe material rundown ?

I use copper probes (#4 bare copper ground wire) for most stuff. The molybdenum for low resistance materials like copper.

Due to the Peltier/Seebeck effect, the '+' probe gets hotter than the '-' probe. You want it on the thicker material/lower resistance side of the weld.

Also probe pressure is VERY important to getting good welds. You must have a way of applying the proper amount of pressure CONSISTENTLY. Too little pressure and you will blow out the weld. Too much pressure and the weld resistance drops to the point where you can't generate enough heat to do the weld.

If you are welding up 1000 cells in a pack, about the only way of doing this is with some kind of pressure controlled weld head. And because the weld temperature depends upon the probe polarity, you need to be able to adjust the pressure independently for each probe.

My welder supports the use of pressure dependent resistors to monitor the weld pressure. You can configure it to trigger the weld only when the probe pressures are within a set range.
 
liveforphysics said:
I just made my 100v spot welder.

Higher voltages are usually not a good thing. Weld energy goes up with the square of the voltage, so it becomes very difficult to control the weld. Also the weld pulses for a given energy level get very short. This tends to make for poor heating profiles and thermal effects in the material... again leading to weld blowouts. Some fancy welders ramp up the weld voltage, hold it, then ramp it down in order to avoid thermal shock effects.
 
texaspyro said:
My welder supports the use of pressure dependent resistors to monitor the weld pressure. You can configure it to trigger the weld only when the probe pressures are within a set range.

Any chance of a link to such sensors ?

Also would you care to spill a few beans about the cutting mode your welder has ? Do you just PWM the weld fets ? Presumably at a rate which the power supply can keep on top of ? What sort of frequency is used ?

I know always questions :oops:

Cam..
 
diaz said:
Any chance of a link to such sensors ?

http://www.sparkfun.com/products/9376 (they also sell smaller ones)

On the cutting mode, you specify the pulse rate and the width or energy. I think the defaults are 500 Hz, 100 microsecond pulses. You need to set the values so that the caps can remain charged. Doing cuts in energy mode is best, so the caps can drain some and still keep the pulse energy constant. With the charger supply that I have, you are limited to rather low duty cycles if you want to cut continuously.

Also, you need to use a tungsten electrode for cutting. I use a 1/16" diam cerium doped electrode... avoid thoriated electrodes... thorium is rather radioactive.

All the pulses in my welder are generated off a 10,000 Hz timer tick. This limits the pulse on/off time resolution to 100 microseconds. The capacitor voltage is measured every timer tick and the energy flow out of it is calculated in real time.
 
texaspyro said:
http://www.sparkfun.com/products/9376 (they also sell smaller ones)

Ah those funny looking things, think I even have a couple lying about somewhere, was never quite sure what they were until a couple of years back. So I have to ask what do your probes look like, any chance of a pic. I'm interested to see how you managed to cram it all it, I'm guessing from you comments you also have changable tips.

Cam.
 
Right now, I'm back to manual probes. I am rebuilding my kludged up automatic weld head...

For the manual probes, I used two pairs of car audio cable connectors. (they look like http://cgi.ebay.com/NEW-STINGER-SPT820-CAR-AUDIO-0-GAUGE-RING-TERMINAL-/310314404061?pt=LH_DefaultDomain_0&hash=item48402e08dd#ht_2940wt_941)

I join two of the terminals together at a right angle using a bolt through the hole. The cable goes in one terminal. The probe gets clamped in the other using the set screws.
 
Cerium doped tungsten ordered. Pressure sensors ordered. Head scratching in progress as to how to make neat weld probes with replaceable tips and pressure sensors :?
 
I've started circuit design for something akin to what Texaspyro has done.

I'll be adapting the charger / discharge circuitry from Ultrakeet& Jiri, can anyone comment on these?

http://ultrakeet.com.au/uploads/pseudo_complete_v1-2-1.pdf
http://www.pittnerovi.com/jiri/hobby/electronics/welder/schematics1.pdf

Anyone have suggestions for gate drive circuitry? Is there any advantage to using dedicated gate drive circuitry, or should I just use another MOSFET like everyone else seems to do? How important is topology here? I'd worry, looking at Texaspyro's example, that a driver at the end of the board connected to a common "gate" bus might not result in even turn on?

Been planning on using the IRFP4468 over the IRFP2907; lower Rds(on) (33% less resistance), Qg (faster gate charge), and higher power dissipation(157%). Runs at 100V too, so potentially usable for other projects. Diaz used 10 FET's, Texaspyro 16... I was planning on 8, because I already feel like this is overkill. ;) I was planning on heatsinking them.

Is there anything to a voltage divider other than an ADC tapping the center of two (high value) resistors?

I just went through Texaspyro's posting history... I could've sworn he has a MOSFET blow out at some point, but I cannot find the post. :( I really was hoping to relive a cautionary tale. I could really use recommendations for dealing with transient voltages / inductive kick back. Thyristors, somewhere, or something? I just have no idea.

Diaz, where did you get the _awesome_ bus bars? Those things look incredibly nice, way better than the 3oz copper route I was slowly moving towards. Hrm, I suppose I could ask around and see who can fab something like that for me.
 
The bus bars started life as 3mm copper (Obtained from good old eBay), then attacked with my milling machine to make them look as per the picture. Not the easiest things to solder. I got em warm with a hot plate then soldered each FET in with a big-ish 75w iron, warm work.
 
The only FETs that I blew were in my charger circuit. I was playing with some code that required me to disable some of my protection systems and did something stupid.

The firing FETs don't seem to need any heatsinking at all. I can detect no heat rise even after welding 1000 thermocouples in one sitting. My charger FETs definitely need heatsinking, but nothing serious. Each pair of FETs in the charger share one heatsink (a medium sized TO-220 heatsink). With that, I get no heat rise. The only thing that gets warm is the 1 ohm capacitor discharge resistor when I drain the caps when shutting down the unit... there is 800+ watt seconds stored in those caps. I can drain them about three times in a row before triggering my thermal shutdown warning.

I drive my firing FET bank gates with a cheesy/crappy/slow/way obsolete DS0026 driver from one end. I have been meaning to replace it with a better driver connected to the middle of the row of FETs, but have never gotten around to it. It seems to work just fine. I am using a larger value series resistor to the gates than I originally designed. I was getting over 2 BILLION amps/second of dI/dT and thought it best to tame that some.

The charger FETs (P-channel) are driven from a 7406 open collector hex inverter.

Oh, and I have 18 IRFP2907 FETs on the firing circuit, not 16.
 
I just built a battery tab welder from plans at ledhacks.com. It uses two 2.2 farad capacitors firing through a thrystor. I just tried it for the first time and it doesn't have enough power to weld the tab. It will just barely stick. The capacitors are car stereo capacitors Did I make a mistake getting these capacitors or could something else be the problem??
 
I have 3 Monster brand 1 farad caps (two from Texaspyro), and they do go PHWAP and weld nickel shim no problemo. They all measure over 1 farad.


But I blew my SCR. Hopefully only half of it is blown, it's two-in-one.


I saw a DIY welder on hackaday, where the guy just whacked two big copper blocks together. I'm thinking a refinement of that may do me. Otherwise it's lots of FETs or a very expensive SCR. My $60 SCR just couldn't hack it - it had a 2000 amp peak rating, but it still died..

I did have a flashy 6 farad cap, but it measured 0.8 farads and was useless.
 
dhwort said:
I just built a battery tab welder from plans at ledhacks.com. It uses two 2.2 farad capacitors firing through a thrystor. I just tried it for the first time and it doesn't have enough power to weld the tab. It will just barely stick. The capacitors are car stereo capacitors Did I make a mistake getting these capacitors or could something else be the problem??

99% of card audio caps are fradulent crapassators... Their true values are usually around 10-20% of what they are labeled as, and they have high ESRs. The only ones that I can recommend are Monster Cable and Rockford 1 farad caps. Also the Scosche 1F cap measures OK, but I have not used them much in welding. The Volfenhag ZX-3 and ZX-4 are usable. The ZX-3's are only 1.0F . The ZX-4's are 1.5F (not 4F) (avoid the other Volfenhags).
 
Here are my test results of various car audio caps (the good ones are at the end):


Model XCP-1 by XXX (a division of NIPPONAMERICA.COM):
1195 N.W. 97th Ave
Miami FL 33172
Advertised as: 1.5F, less than 2 milliohms ESR.
Measures:
@20 Hz 0.097F 6.450 miilohms
@120 Hz 0.097F 5.066 milliohms
@DC 0.106F
(note: a second sample had an ESR in the 3 milliohm range)



Model PC1.5F by PowerAcoustik.com
Power Acoustik Electronics
1550 S. Maple Ave.
Montebello, CA 90640 USA
Advertised as: 1.5F, ESR not specified
Measures:
@20 Hz 0.199F 4.048 milliohms
@120 Hz 0.200F 3.470 milliohms
@DC 0.212F



Model CAP300DBL by PyramidCarAudio.com
Pyramid Car Audio
1600 63rd. Street
Brooklyn N.Y. 11204
Advertised as: 3.0F +/- 5%, ESR less than 1.6 milliohms
Measures:
@20 Hz 0.415F 2.87 milliohms
@120 Hz 0.390F 2.40 milliohms
DC 0.457F



Model CAP3.5B by BOSS Audio Systems (bossaudio.com)
3451 Lunar Court
Oxnard CA 93030
Advertised as: 3.5F +/- 10%, ESR less than 1.95 milliohms
Measures:
@20 Hz 0.520F 1.5 milliohms
@120 Hz 0.540F 1.4 milliohms
DC 0.563F




Model SQCAP2M by SoundQuest (sqelectronics.com)
a division of AAMP of America
13160 56th Court
Clearwater FL 33760
Advertised as: 2.0F, ESR not specified
Measures:
@20 Hz 0.204F 3.390 milliohms
@120 Hz 0.202F 2.223 milliohms
DC 0.233F


Stinger HPM model SC201MSCPT
By: Stinger Electronics (same as Squndquest)
13160 56th Court
Clearwater FL 33760
Advertised as 1F +/-3%, 1.6 milliohms, 16V/20V surge(?)
Note: this cap is not totally bogus like the Soundquest cap,
but it is still very far from its advertised specs.
Measures:
@20Hz 0.674F 2.33 milliohms
@120Hz 0.710F 2.20 milliohms
DC 0.736F



Model IB-CAP1F by TheInstallBay.com
Metra Electronics
460 Walker St
Holly FL 32117
Advertised as 1.0F, ESR not specified
Measures:
@20 Hz 0.211F 2.64 milliohms
@120 Hz 0.210F 1.90 milliohms
@DC 0.231F


Model Power Pumper by Tsunami
Note Tsunami is distributed by Metra (same as Installbay)
Advertised as 1F
(note: This cap was purchased used... may be defective.
Additional samples are needed)
Measures:
@20 Hz 0.295F 8.32 milliohms
@120 Hz 0.298F 7.87 milliohms
@DC 0.302F


Model PP815-SCAP by Tsunami
Advertised as 1.5F +/-10%, < 1.2 milliohms ESR @120 Hz
19V working, 21V surge
Measures:
@20 Hz 0.473F 2.03 milliohms
@120 Hz 0.504F 1.78 milliohms
@DC 0.540F

Tsunami PP912UM-CAP
Advertised as 1.2F +/- 10%. Package marked 24V working, 30V surge. Cap marked 20V.
Current catalog says 24V surge.
Measures:
@20Hz 0.691F 1.81 milliohms
@120Hz 0.730F 1.99 milliomms
@DC 0.790F



Model 1.0 Farad (?) by MAS Power Road Rage
Advertised as 1.0F, ESR not specified
Measures:
@20 Hz 0.620F 2.47 milliohms
@120 Hz 0.680F 2.30 milliohms
@DC 0.687F


Model ZX-2.2 by Volfenhag Audio
4480 Pacific Blvd
Vernon CA 90058
Advertised as 2.0F, 1.95 milliohms @ 120 Hz
(note: fine print on back of package says capacitance value is
appoximate and may vary "substantially". Their web site says
+/- 5%)
Measures:
@20 Hz 0.422F 2.81 milliohms
@120 Hz 0.420F 2.10 milliohms
@DC 0.480F


Model ZX-3 by Volfenhag
Advertised as 3.0F 1.95 milliohms
(note: fine print on package says value may vary by 1.2F!)
Measures:
@20 Hz 0.904F 1.7 milliohms
@120 Hz 0.920F 1.4 milliohms
@DC 1.053F


Model ZX-4 by Volfenhag
Advertised as 4.0F 1.95 milliohms
(note: fine print on package says value may vary by 1.2F!)
Measures:
@20 Hz 1.31F 1.150 milliohms
@120 Hz 1.44F 0.955 milliohms
@DC 1.532F


Model LQ19CAP by Lanzar
1600 63rd St
Brooklyn, NY 11204
Advertised as 1.9F +/- 5%, 1.6 miiliohms, 20V-24V surge
Measures:
@20 Hz .218F 3.60 milliohms
@120Hz .219F 3.35 milliohms
@DC .239F



The following capacitors met or exceeded their advertised ratings:

Model ECAP1 E2 by Scosche
Advertised as 1.0F +/- 0.2F ESR not specified
(note: wholesale price was quoted to be "over twice what the XCP-1 was")
Measures:
@20 Hz 0.815F 1.52 milliohms
@120 Hz 0.904F 1.37 milliohms
@DC 0.932F


Model RFC1 by Rockford Fosgate
Advertised as 1.0F +/- 10%, ESR less than 1.95 milliohms
(note: wholesale price is around $90)
Measures:
@20 Hz 0.912F 1.17 milliohms
@120 Hz 0.966F 0.91 milliohms
@DC 1.019F

Model Punch by Rockford Fosgate:
Same specs and performance as the RFC1 and RFC1D


Monster Cap by Monster Cable
(Note: this unit had a label with a white background, no model number)
Advertised as 1F, 20V, ESR not specified
Measures:
@20Hz 0.860F 1.16 milliohms
@120Hz 0.947F 1.09 milliohms
@DC 1.015F

Intellicap by Monster Cable
(Note: this unit had a label with a silver background and a pattern that looks
like a circuit board, model number thought to be MCAP-1FD S) Unknown if it
is the same as the current model MPC P300 MCAP-1F model, but likely is)
Advertised as 1F, 20V, ESR not specified
Measures:
@20Hz 1.019F 0.84 milliohms
@120Hz 1.300F 0.67 milliohms
@DC 1.192F

Advertises as 0.5F 20V
@20Hz 0.47F 1.7 milliohms
@120Hz 0.48F 1.3 milliohms
@DC 0.51F
 
Anyone one know where I could buy elkonite rod at reasonable price (a place that ships in Canada ) ?

I have no success with copper electrodes ...
 
Hello!

New guy here who was very interested in building my own tab welder but aside from part sourcing, matching and trial and error in finding what works and what doesn't and the actual build time..... I succumbed to the idea of finding something already built......in China. I know, I know but I did find something affordable and now very curious if anyone here has seen these already or possibly even owns one????

At roughly $200 delivered, I'm seriously interested/contemplating in giving it a try but figured I'd check here first. I'd be willing to be the test dummy...lol. Here's a link, what do y'all think??

http://www.aliexpress.com/product-fm/288884131-MCU-Spot-Soldering-Station-POWER-787A-welder-wholesalers.html

Within the ad, they include a link to watch a video of it in use, their English isn't so good tho. After comparing this specific model against a few others on the site, this one seems the most promising, IMHO. Due to thicker leads/pins and the option of using a foot pedal. If any Chinese members can translate other features that'd be great too. :)
 
Back
Top