Battery tag welder from China

[robrenz]

Great! very interested to see how they control the triac with an inductive load.  Yes please show waveforms.

[robrenz]

Nice pictures and the unit looks pretty good also.  Makes me very confident that my spot welder will make an awesome tab welder.  My problem may be taming the output down enough for small parts.  More resistance on the secondary does reduce the output current so it might just be using longer runs of #1 welding cable to the actual welding probes.  Since yours is transformer based you dont have to worry about capacitor recharge times but you still have to watch out to not exceed the duty cycle whatever that is.  My spot welder is only 50% duty cycle but that should not be a problem.

I would be very interested in the triac input as to whether they are using zero turn on and random turn off so they can just energize the primary for a certain number of milliseconds duration to control the weld power level. And is the lowest power setting less than half of a line cycle

Also have read about dual pulse where the first pulse is set just low enough power to barely stick the parts together but establishes a more consistent contact resistance for the main pulse that actualy does the final weld. (supposed to give more consistent results) I figure initialy I will just fire the SSR I end up using on the primary of my spot welder with my DG1022 and I can just make arbitrary waveforms to control the dual pulse durations.

If you have a differential probe for your scope or if the secondary is truly isolated you could connect to two points on one side of the secondary acting as a shunt and read the current waveform of the weld pulse.

[mzzj]

Seems to work better than I expected.
How thick Nickel strip You can do?  High power battery packs(ie portable drill, rc-toys ) with 30-50 Amps of current would need some thickness...

[SeanB]

Pity he will not ship to my country.

[firewalker]

 

Are you going to investigate the way they control the transformer to control the current?

Alexander.

[nukie]

Try aluminium or copper.

[philpem]

Here's a pretty simple one (if the SCR is big enough the inductance of the welding cable is enough to limit dI/dt I guess, with an inductor you could use higher voltage electrolytics though).

http://www.philpem.me.uk/elec/welder/
http://www.rcgroups.com/forums/showthread.php?t=763308

Heh, I didn't expect to log in and see someone talking about one of my projects

For the record, the welder works pretty well - I used it to build a NiMH pack for a Psion Workabout and rebuild several laptop LiIon packs. That said, it really needs its own PSU, a case, and dual-pulse welding (which means swapping the SCR for a MOSFET).

The 50RIA20 makes a good SCR, I haven't managed to blow one up yet. All the wiring is car battery power cable (6AWG or something like that, possibly thicker) on copper busbars (thick copper sheet from the local hobby shop, cut into strips).

[robrenz]

Glad it worked out for you

I know you are busy and bought this as a tool not a project. And you have no obligation to satisfy every forum members whim. But if you ever open it up again to investigate and are feeling generous, I would be very interested in the triac input on time relative to the current settings. As a guess it may be the current power number is equal to the number of line cycles of on time on the primary. You mentioned it was dual pulse so it would be interesting to see what the ratio of first pulse to second pulse on times are.

Even if you never get to the above, thanks for the thread because I now know I can mod my spot welder to do the same.

[robrenz]

Thanks, and obviously no rush

[Spawn]

LOL Aurora He wants the Fluke CNX3000 

Ontopic again, I still have a lot batteries to weld, but hobby budget of the month went in to another things so I am trying to avoid this topic  But it is looking good, thanks for the photo’s Aurora

[ConKbot]

Hmm, this thread is relevant to my interests... and timely too.


Though obviously I'm going for the CD route instead of using a step-down transformer.

[entoni]

LOL Aurora He wants the Fluke CNX3000 

Ontopic again, I still have a lot batteries to weld, but hobby budget of the month went in to another things so I am trying to avoid this topic  But it is looking good, thanks for the photo’s Aurora

 

[nukie]

Very nice to know there are people who makes custom tools in small quantities these days. I have a manual lathe, the time and material cost definitely more than 2 pounds here in ozland. Looking at the part, it will require at least 3 tool changes and soft sticky nature of copper makes it hard to machine.

[SeanB]

The fuse blowing is most likely because of inrush current, and using a standard fast blow fuse ( common on plastic sealers, which use the same style of construction. They should use a slow blow fuse of 8A, but it is more expensive. 1kVA is common as inrush power, big transformer and low resistance windings. On the plastic sealers the transformer is 56V unloaded, around 2V when heating the strip, though this is incredibly sensitive to the temperature of the sealing strip. I modify the controllers by adding a 100k resistor across the pot outer connections, so that when ( it always will be) turned up to max nuke it is only half the time. 8 seconds of heat is enough to melt a 5mm plastic strip, never mind a 100 micron plastic baggie.

[robrenz]

@ Aurora,

Thanks for taking the time to do those measurements. I had a preconcieved idea that the power would be controled by the portion of/number of powerline cycles fed to the primary of the transformer.  I believe that is how the sunstone units control the power levels. But I dont see any significant difference in the voltage pulse length on your traces but the amplitude changes and they all are just one line cycle (first pulse on the top of the sine wave and second on the bottom of the sine wave)  . I think the high resistance of the scope connection (= no load) is keeping us from seeing the real weld pulses. When I get to experiment with this I plan to use two differential amplifiers connected to portions of the input cabling and output cabling acting as shunts. That will let me see the actual input/output current waveforms including whatever inductive kicks are happening.

Thanks again, robrenz

[bgeery]

Any update on this welder?  I was looking to buy one ASAP, so the info would be timely and helpful? 

[Fraser]

Still working well. No complaints.

I have welded tabs onto all manner of cells and not a single failure in service to date.

The only problem I have has is that if I use a current setting above 50 I trip the whole house RCCD ! The welder passes the tests on my Seaward 2000 PAT tester so it is not faulty.....it just upsets the sensitive house breaker at high instantaneous currents. I normally weld at level 25 to 30 so all works fine.

I purchased the correct thin zinc battery tab material and highly recommend you do the same rather than trying to improvise. The come in strips and individual tabs in various widths and I bought all mine from China very cheaply.

Would I recommend this welder to a friend? YES. For the money, and my lightweight application, it is excellent.

[turbo!]

I recall seeing a "spot welder" at Harbor Freight.  You could also make one yourself with a set of capacitors and control the available energy by setting the voltage you charge the capacitors to.

[macboy]

I wonder if the issue if finding caps with good enough discharge characteristics could be solved simply by using a  large number of low-cost electrolytics in parallel, using their lead resistance to control the peak current out of each one.

One concern about putting many caps in parallel is the failure mode.
You will be subjecting these caps to very high stress and failure will happen eventually. Consider what will happen when a capacitor fails. You have say 10 or even 100 caps in parallel. One fails, short circuit. What happens? The other 99 instantly discharge through the shorted one, and KABOOM! You will see an explosive failure which could not occur with just one large capacitor.