kWeld - "Next level" DIY battery spot welder

[parasole]

The calculations result from my experiments, and findings that I came to with a customer here in Germany who was using 5 x 1F (huge) electrolytic audio capacitors. Even they don't provide enough energy for one 50 joulse pulse to weld 0.3mm nickel.

If you use 6AWG and no fuse, then this helps improving the efficiency and explains why you can do 0.15mm nickel.

Please don't get me wrong, I don't want to critizise your impressive design, just want to share my experience with smaller capacitors.

Don't worry, I do not take it as critics, and sharing mutual experience I beleive is beneficial for the community... And your work and effort is no way to be underestimated.
Using audio caps may be the answer for poor performance, unfortunatelly almost all "modern" audio caps are just piece of junk and very few brands do provide the declared values. Before I built my caps bank I dig out as much as possible on this subject and got pretty good insight of audio caps quality from endless-sphere thread devoted to an SCR variant spot welder: https://endless-sphere.com/forums/viewtopic.php?f=2&t=2633. And on page 34 is an impressive report by Toorbough ULL-Zeveigh about their real quality: https://endless-sphere.com/forums/viewtopic.php?f=2&t=2633&start=825
Initially I bought 3pcs 0.49F caps, just to find out they are about 0.2F with quite high leakage, still not bad but quite not as declared for their cost :-) I even did try to open one, however found it to be fully embedded in paraffin without an aluminium can (and no brick inside , so decided better I will live it as is and use it considering measured capacity.
Any way, check out on page 40 green666 solution, his implementation nail down my decision to go by this route, real 1F (or more who wish for it) with very low ESR and leakage is proven to be an effective solution, altough it take more work to build it up.

[branadic]

Just wanted to let you know that the capacitor module is now available for pre-orders at https://www.keenlab.de/index.php/product/kweld-ultracapacitor-module/! I expect to be able to deliver by beginning of June. The production batch size for this first round is not too big, and I expect not to be able to build up regular stock yet.

Question: The module available is a redesign of the pcb shown or the version with bodge wires? How do you charge the capacitor bank? Are there plans to offer a charger in near future as well as a complete welding system?

-branadic-

[tatus1969]

Using audio caps may be the answer for poor performance

My customer had been measuring them, and their ESR and capacitance was as expected. The initial current was >1000A. In his case, it is just the limited amount of capacitance that is keeping him from welding thicker material.

impressive report by Toorbough ULL-Zeveigh about their real quality: https://endless-sphere.com/forums/viewtopic.php?f=2&t=2633&start=825

Yes, they are very creative when it comes to alienate their customers: https://www.eevblog.com/forum/projects/kcube-led-cube-design-revisited/msg1482458/#msg1482458

Just imagine them putting that much effort and creativity into designing real products. That would be a real threat to the Western industry!

Question: The module available is a redesign of the pcb shown or the version with bodge wires? How do you charge the capacitor bank? Are there plans to offer a charger in near future as well as a complete welding system?

Yes of course :-) You'll only be able to buy professionally made and tested stuff from me (guaranteed without bricks or sand).

I am actually working on a charger (https://www.eevblog.com/forum/projects/guesses-on-what-i-am-attempting-here/msg1314337/#msg1314337) and have a prototype running under full load. But there is one problem with it. Its specifications are matched to allow full load operation of welder and capacitor module (one weld of 50J every second), and therefore it is capable of delivering 80A at 8.1V. This comes at a cost and I wouldn't be able to sell the charger module for less thanj 70€. But as many will be using the welder for casual operation and would rather spend some time waiting for recharge instead of spending that much extra money, this project is currently not at the top of my list.

My current low-cost proposal is to take an old AXT power supply, and feed the 12V into an adjustable buck converter. There are many of these like https://www.amazon.com/Aideepen-Constant-Current-Adjustable-Converter/dp/B0747QDRW9

[HDRW]

Hi, newbie to this place (but not to electronics - learned which end of a soldering-iron gets hot at the age of 8!).

I want to "repair" cordless-drill battery packs, so Sub-C NiMH rather than the 18650s that a lot of people seem to be doing, and this project looks great for this.  I am very impressed with Frank's engineering and experimental approach.  I wish I was as conscientious! 

I see the kits are out of stock and I have signed up to be notified when the board-only kits are back in (I hope - haven't seen an acknowledgement).

Is there an ETA for the next batch of boards?

I'm on a tight budget (so the €100 capacitor kit is too much) and I'm not fond of mistreating Lithium batteries, but I have quite a selection of Sealed Lead Acid Batteries (SLAB) from my other hobby of Amateur Radio, but I haven't seen anyone mentioning this much, such as size of battery, using them in parallel, etc.

I have choices from about 7Ah to 80Ah (all 12V) but am mindful of how impressive shorting one of these can be!  Would it make more sense to mount the 300A fuse directly onto a battery terminal in this case, and connect the cable from it to the "far" side of where it would usually be?  That way a short anywhere, apart from directly across the battery terminals, would be protected.

Thoughts?

Cheers,
Howard

[carracing111]

Perfect, thanks for the quick response.

I will order two boards without condensers plus a full board.
I could try 4S3P with my capacitors. I guess I can make some kind of adapter, so that the pins of my capacitors match.

I attach a couple of photos of my current capacitor configuration and with the welding head that will inherit my new Kweld!

If someone is interested in a head like mine, I can send detailed photos, explain how it works and offer a good price.

Regards

https://preview.ibb.co/hZaJ4x/20180330_122252.png

Where did you get the pneumatic head? And price?

Hi.

The machine, I get directly from the manufacturer Will Best. It is a brand that has been in the construction of welding machines for 19 years and it seems that the quality is quite good. I am very happy with my welding head.

They have offered me to distribute their machines in Europe, it's not something I'm going to take seriously...
The price that I offer is exactly the price they give me without earning anything. The idea is to get a discount on other machines that I plan to buy, such as a nickel cutter and ultrasonic welding machine.

The price of the welding head with double pneumatic system is of 880 dollars and the shipment by DHL (3 or 4 days) costs 121 dollars.
Maybe you can get a lower price with the system by springs. I think about 50-80 dollars less, like the one on the photos on my machine. I have both systems and soon I will install the double pneumatic system.

If you want detailed photos of any part or information do not hesitate to ask!

I add some pictures of the machine with the double pneumatic system

[carracing111]

I bought two HP power supplies - DPS-800GB - 1000W and I am interested in buying two controller cards like the ones in the photo! Is this possible, when can they be available?

Thank you very much !

[tatus1969]

I see the kits are out of stock and I have signed up to be notified when the board-only kits are back in (I hope - haven't seen an acknowledgement).

I think that the shop doesn't send an acknowlegement email after registering to the waitlist. But you will definitely be notified immediately after I have been addding stock for pre-orders.

Is there an ETA for the next batch of boards?

My orders are all out, the only reason why I haven't started offering them yet is that I don't have definite lead times from the electronics assembly and cable assembly firms yet. They had promised to deliver that this week, but that's life...

I'm on a tight budget (so the €100 capacitor kit is too much) and I'm not fond of mistreating Lithium batteries, but I have quite a selection of Sealed Lead Acid Batteries (SLAB) from my other hobby of Amateur Radio, but I haven't seen anyone mentioning this much, such as size of battery, using them in parallel, etc.

I have choices from about 7Ah to 80Ah (all 12V) but am mindful of how impressive shorting one of these can be!  Would it make more sense to mount the 300A fuse directly onto a battery terminal in this case, and connect the cable from it to the "far" side of where it would usually be?  That way a short anywhere, apart from directly across the battery terminals, would be protected.

I haven't personally tested, but my feedback from a number of customers is that any decent and new 65AH lead acid car starter battery would work fine.

I bought two HP power supplies - DPS-800GB - 1000W and I am interested in buying two controller cards like the ones in the photo! Is this possible, when can they be available?

I am sorry that this is going slow, but it will come soon. At the moment I am working on sourcing kWeld components, building kCap capacitor modules, and designing five new products (two of them battery related including this charger). All of these are aside of my main work, which is desigining custom electronics for two different customers. I'm pretty busy ;-)

[carracing111]

I'll wait for them to be available! As soon as you let me know, I make the payment for the preorder!

Regards

[branadic]

I am sorry that this is going slow, but it will come soon. At the moment I am working on sourcing kWeld components, building kCap capacitor modules, and designing five new products (two of them battery related including this charger). All of these are aside of my main work, which is desigining custom electronics for two different customers. I'm pretty busy ;-)

Can you tell us more about that five products? Why I ask is, that I'd like to know what is coming next. If I today order the kWeld ultracapacitor module, is it compatible with products coming soon or is it an isolated application?

-branadic-

[tatus1969]

Can you tell us more about that five products? Why I ask is, that I'd like to know what is coming next. If I today order the kWeld ultracapacitor module, is it compatible with products coming soon or is it an isolated application?

One is the mentioned 1kW CC/CV charge controller.

The second I have to keep secret for now. It is something that does not exist on the market at all as far as I know, and it is a bit larger in scale (effort and time). Besides a hardware/firmware/USB design it involves a Bachelor thesis and a complex website. I hope that it will be a large throw, and I don't want to give this idea out of my hands and into our far East friends...

Next is something totally useless, but hopefully very beautiful: https://www.eevblog.com/forum/projects/kcube-led-cube-design-revisited/msg1482458/

Then something of which I hope that it will be very useful, while being very simple. I just ordered samples of 1.5mm x 5mm x 6mm tinned copper platelets in standard SMD reels, because I was frustrated of not having an elegant - and cheap - solution to high current PCB tracks. These platelets cost almost nothing and can be fed into standard SMD pick&place machines. You can literally pave your high current PCB tracks with them and then use standard 35um FR4 material. A 50mm x 6mm x 35um PCB track can carry 15.7A when dissipating 1W. The same PCB track, reinforced with 9 of these platelets can carry 109.6A (!!). Besides of using them for my high current projects (I will test if kWeld can work completely without bus bars), I hope to sell these later on.

The last project didn't get enough interest *and* it was not new, so I stopped it. It was a special filter arrangement for inspection microscopes that would drastically improve readability of otherwise sometimes impossible-to-read laser-marked chips: https://www.eevblog.com/forum/reviews/rework-microscope-how-to-make-laser-marked-chips-readable/msg1464441/

[branadic]

Thanks for your answer. So if I get you right the only product on kweld from this five projects is the charger and this will be compatible with the ultracapacitor module? The rest is more or less another application and has nothing to do with the welder.

-branadic-

[tatus1969]

Thanks for your answer. So if I get you right the only product on kweld from this five projects is the charger and this will be compatible with the ultracapacitor module? The rest is more or less another application and has nothing to do with the welder.

[/quote
 

[JohnG]

Then something of which I hope that it will be very useful, while being very simple. I just ordered samples of 1.5mm x 5mm x 6mm tinned copper platelets in standard SMD reels, because I was frustrated of not having an elegant - and cheap - solution to high current PCB tracks. These platelets cost almost nothing and can be fed into standard SMD pick&place machines. You can literally pave your high current PCB tracks with them and then use standard 35um FR4 material. A 50mm x 6mm x 35um PCB track can carry 15.7A when dissipating 1W. The same PCB track, reinforced with 9 of these platelets can carry 109.6A (!!). Besides of using them for my high current projects (I will test if kWeld can work completely without bus bars), I hope to sell these later on.

Brilliant! I would also consider using them as heat spreaders on a PCB.

John

[tatus1969]

Brilliant! I would also consider using them as heat spreaders on a PCB.

They are tinned and thus have a high reflectivity, which doesn't help for heat dissipation. But there won't be much heat to dissipate (at least for the PCB traces) ;-)

This is how they look like, I'll receive the first sample reel in a few days:

[tatus1969]

I have just finished the laser-cut housing for the capacitor module and ordered a few copies. If someone is interested, I have put them in the shop here:

https://www.keenlab.de/index.php/product/kcap-laser-cut-housing-kit/

The DXF files are available here for download:

https://www.keenlab.de/index.php/product/kcap-laser-cut-housing-dxf-model/

I'm becoming a fan of acrylic sheet material laser-cutting, this design works without glueing a single piece 

[biot]

I have just finished the laser-cut housing for the capacitor module and ordered a few copies.

Any particular reason this is not a full enclosure? Do these caps heat up perhaps?

[tautech]

I have just finished the laser-cut housing for the capacitor module and ordered a few copies. If someone is interested, I have put them in the shop here:

https://www.keenlab.de/index.php/product/kcap-laser-cut-housing-kit/

The DXF files are available here for download:

https://www.keenlab.de/index.php/product/kcap-laser-cut-housing-dxf-model/

I'm becoming a fan of acrylic sheet material laser-cutting, this design works without glueing a single piece 

FYI:
https://www.eevblog.com/forum/chat/personal-laser-cutter/

[tatus1969]

FYI:
https://www.eevblog.com/forum/chat/personal-laser-cutter/

OMG, that's a dream. But at that price, it will remain that for quite a while. I just made another dream true and bought a FLIR One Pro, these have finally come down to price tag that I can afford 

Any particular reason this is not a full enclosure? Do these caps heat up perhaps?

There are a few. Less material and cost, simpler construction (building the housing works without glue, just four additional 15mm standoffs needed), better heat dissipation capabilities (yes, the capacitors heat up, that's why I added 120mm fan support for heavy duty operation), and last but not least I think that it looks beautiful. Reminds me of a vintage tube amplifier 
But thanks for rasising the question! I have just added six adhesive stickers for the capacitor caps, because they have + potential and the housing would not provide full protection otherwise. They will be included in the kit.

[mzzj]

Sooo.. had to try also. 18pcs  NTP75N06 mosfets (why such a lousy mosfet and 18 pcs...thats what I had in drawers) in parallel, soldered to 2 mm thick copper sheet, FR4 board between as a separator and source leads bent and soldered to another ~1mm copper sheet.

My quick and dirty mosfet switch under test:

650mV over the 600A/60mV shunt resistor so that makes it 6.5kA! 

Measured voltage over the mosfet RDS-on during switching = 3 volts, 8 mOhm typical rds-on -> 0.44mOhm typical and 6.8kA measured that way. Scary amount of avalanche energy but the mosfet pack seem to handle it fine. (final use is going to be at less current than 6.5kA)

[mzzj]

Instead of car battery or BCAP's I might go to the multitude of 16 volts electrolytics.
https://www.digikey.fi/product-detail/en/nichicon/LLS1C473MELB/493-7253-ND/2548837

50 pcs of those would be ~130 euros (assuming the 100pcs volume discounts), would provide 0.4mOhm ESR and 40kA short-circuit current 

Capacitor losses would be  neglible compared to  6x BCAP 350F ultracaps with combined 18mOhm internal resistance and you could limit the current with more suitable part than caps itself. ( piece of rebar for example)

[Kjelt]

Instead of car battery or BCAP's I might go ....

Check the datasheet first, won't surprise me if they can't uphold this torture test for long, and keep the temperature monitored.

[mzzj]

Instead of car battery or BCAP's I might go ....

Check the datasheet first, won't surprise me if they can't uphold this torture test for long, and keep the temperature monitored.

As usual, datasheet is not much use in this sort of use. I could run just a one or two roughly similar caps in "accelerated" torture test but really don't expect to see anything interesting there.. 
3s2p 350F BCAP pack with  4.5mOhm ESR is going to have 4.5kW internal losses at 1kA, whereas 50x electrolytic pack would have ~400W or less than 1/10th compared to 2p3s  at 1kA.
So judging by ESR the heating is not going to be much  of a problem. 

[mzzj]

Took less energy than I expected to weld 0.10 and 0.13mm tabs and I was able to get sound welds with only 0.132F total capacitance.
0.10mm tabs were welding best at about 21 volts =  29 Joules
0.13mm tabs were ok at 25 volts = 40 Joules.

Weld pulse is pretty damn short and if the contact pressure is not sufficient I get loud bang and blackened terminals.
Need to re-write the program to actually use dual-pulse functionality, now the pre-pulse is already doing all the welding and draining the caps. 

[tatus1969]

50 pcs of those would be ~130 euros (assuming the 100pcs volume discounts), would provide 0.4mOhm ESR and 40kA short-circuit current 

That would be C=2.35F. With V_before_pulse=16V and V_after_pulse=8V, these can deliver 225 joules. That should be enough if you can maintain a high system efficiency, but that would mean that you need to switch a current of 5000A or more.

Capacitor losses would be  neglible compared to  6x BCAP 350F ultracaps with combined 18mOhm internal resistance and you could limit the current with more suitable part than caps itself. ( piece of rebar for example)

kCap uses 3x2 BCAP0310-P270-T10 and has a theoretical ESR of 3.3mOhms, plus wiring and PCB tracks.

650mV over the 600A/60mV shunt resistor so that makes it 6.5kA! 

Great job :-)

Measured voltage over the mosfet RDS-on during switching = 3 volts, 8 mOhm typical rds-on -> 0.44mOhm typical and 6.8kA measured that way. Scary amount of avalanche energy but the mosfet pack seem to handle it fine. (final use is going to be at less current than 6.5kA)

If you control welding energy through capacitor voltage, discharge them fully in one pass, then this could work out. Interrupting this amount of current will very likely destroy the MOSFETs from the stored magnetic energy in the conductor loop.

We've started shipping kCap modules and enclosures today, here's a first impression of how that assembly looks:

[kulla]

Any news on stock update, is it going to happen on the end of June as planned or there might be delays?

I ordered mine complete kWeld in the beginning of May, can't wait to get my hands on it