kWeld - "Next level" DIY battery spot welder

tomjasz said:
Great work,

I'm looking forward to the next release.

I really have no interest in owning Lipo again. What are the options?

Also, since I have 2 other welders and three sets of weld electrodes and wiring, can the next kit be purchased without selecting "cable / electrode holder / electrodes completion grade"?


Thanks!

Thanks!
Any power source with low impedance (less than 10 milliOhm), a suitable voltage (12 - 30V DC) and at least 500J of stored energy will do (only 10-20% of that energy will arrive at the weld spot). Other than Lipos, your immediate options are car batteries and supercapacitors. The latter are very expensive though.

I made the article structure in the shop flexible enough that you can either purchase a full kit, or you can select just the components that you really need. The assembly manual gives an overview of all the required parts.
 
tatus1969 said:
I made the article structure in the shop flexible enough that you can either purchase a full kit, or you can select just the components that you really need. The assembly manual gives an overview of all the required parts.
Thanks I was trying to add the complete without the electrodes, as I have several sets already.
 

Attachments

  • Screen Shot 2017-08-15 at 4.31.26 PM.png
    Screen Shot 2017-08-15 at 4.31.26 PM.png
    29.8 KB · Views: 5,213
tomjasz said:
tatus1969 said:
I made the article structure in the shop flexible enough that you can either purchase a full kit, or you can select just the components that you really need. The assembly manual gives an overview of all the required parts.
Thanks I was trying to add the complete without the electrodes, as I have several sets already.
In your case you do not go for the complete kit, but order an electronics module, some fuses, a set of bus bars, maybe the mechanical stuff like screws, and maybe a foot switch. Maybe also the cable/electrode holder set, but I doubt that the electrode holder will match your existing electrodes.

The "complete kit" always includes everything and only lets you choose to get the stuff pre-manufactured or loose, this is neither what you want. Have a look in the assembly manual, the first section describes the different items one by one.

Anyway, let's better discuss this via PM or email, as it may not be of public interest.
 
Fantastic! Really enjoying your videos, please keep it up!

Glad to see that there's no issues with the threaded connection, results in a very clean electrode!

I would recommend you also compare the Hobbyking graphene battery, https://hobbyking.com/en_us/graphene-5000mah-3s-65c-w-xt90.html

These have proven to be even more power dense than majority of the older nanotech packs and longevity is dramatically improved. Given that you've got an easy method for comparing batteries it could be very interesting indeed.
 
Ohbse said:
Fantastic! Really enjoying your videos, please keep it up!

Glad to see that there's no issues with the threaded connection, results in a very clean electrode!

I would recommend you also compare the Hobbyking graphene battery, https://hobbyking.com/en_us/graphene-5000mah-3s-65c-w-xt90.html

These have proven to be even more power dense than majority of the older nanotech packs and longevity is dramatically improved. Given that you've got an easy method for comparing batteries it could be very interesting indeed.

Do you have proof that the graphene packs are any good ?

From what I have read they are the same as the others....
 
[edit: new information was posted below by Ohbse, the graphene LiPo appear to be just as good in this application]

They are roughly the same price, but...the 3S ones Frank/tatus1969 posted are a very high C-rate (130C vs 65C)...If I already had a 65C packs, I would use them, but...if I am buying them , why not get the best?
 
cheapcookie said:
Do you have proof that the graphene packs are any good ?

From what I have read they are the same as the others....
Any real data on Graphene?
I could suck real world 1200A from 2P7S10Ah Nanos for repeated helicopter start with warming up. Going 1200-600A in 10-30 sec (manual limits start intent to 30 sec) with 15 min rest, depending on temp and other conditions, they got warm at most. :D
P.S. Russian-Polish Mi2
 
spinningmagnets said:
They are roughly the same price, but...the 3S ones Frank/tatus1969 posted are a very high C-rate (130C vs 65C)...If I already had a 65C packs, I would use them, but...if I am buying them , why not get the best?

As you're no doubt aware, the hobby C rating is essentially meaningless outside of establishing approximate categories of energy VS power density for differing applications. If you read the detail in the listing for the graphene packs you'll see they also include the '130C burst' in the specifications, just not in the headline.

In the real world they appear to hold up to high discharge cycles better than nanotechs and appear to have a lower IR. For this type of pulse my guess is they will prove superior by a small amount in power output, but substantially better in longer term usage. There's an enormous wealth of information and comparisons on hobby flight forums.

Of course an alternative to improved performance and longevity is to use a larger capacity pack or double up in parallel.
 
Ah I looked for the disassembly video I watched a while ago, a rant against those batteries.

The videos contained no testing what so ever. (only disassembly)

I stand corrected, didn't mean to start that kind of stupid war on a nice thread :(

Peace and love !
 
In the real world they appear to hold up to high discharge cycles better than nanotechs and appear to have a lower IR

Thank you for posting that. I have been away from watching LiPo for a few years, and didn't know that. It is very good information, especially since there are frequently times when the HK warehouse has one product in stock, but not the other. It is very useful to hear that both would work well.
 
Ohbse said:
In the real world they appear to hold up to high discharge cycles better than nanotechs and appear to have a lower IR. For this type of pulse my guess is they will prove superior by a small amount in power output, but substantially better in longer term usage. There's an enormous wealth of information and comparisons on hobby flight forums.
I will concentrate on shipping the kits for now, and start evaluating the use of supercapacitors thereafter (I already have six Maxwell BCAP0310 lying on my desk). But I will be happy to collect your experience with other power sources, and add them to my list of recommended ones after your thorough testing.

No matter which power source you choose for the welder, you also need to make sure that the current stays below 2000A. To limit it to 1700A when using a 12V power source, the total system resistance must therefore stay above 7 milliOhms (=12 / 1700). Cabling and power switch account for approx 2.5 milliOhms, the weld spot resistance should be assumed to be zero to be on the safe side. This results in a minimum internal resistance of the power source of 4.5 milliOhms. Large supercapacitors, massively paralleled Lipos, or large lead acid batteries may fall below that limit. In this case, further resistance needs to be added, for example by extending the cables.
 
It was quite some work, but I have managed to ship all orders in time (week 34) as promised! Thank you for your patient waiting, and I am curious about your feedback.
 
Impressive!! I wish I saw this sooner and was getting one too!! If anyone changes their mind I'll pay extra for one. I've got a sunkko that I prefer not to ever use but I don't want to wait months longer for my build.
 
I am busy preparing the user manual. When I need a break from that, I'll be playing with two different options of supercapacitors that I have here now:

guesses-on-what-i-am-attempting-here


On the right, 6x Maxwell BCAP0310 P270 T10 (2.2 milloOhms per cell). On the left 30x 100F unbranded 100F capacitors from far east. I did a quick measurement to see if they are overrated or even fake: they are not, measured 113F capacitance and 13mOhm ESR per cell. Not too bad!
 
I've got some first results from using Maxwell capacitors. Four of them are enough for 90 joules into a 1 milliohm weld spot. My requirement was 100 joules, and that is not reached yet.

But their ESR is surprisingly low, I measured only 1.5 milliohms per cell, in contrast to the 2.2 milliohm in the datasheet.

guesses-on-what-i-am-attempting-here
 

Attachments

  • test maxwell.zip
    245.8 KB · Views: 73
I have test my spotwelder.
Info
Cable 16mm2
Car battery 50AH

Cal R 1.77 mR

Spotwelder

E 28J
I 1443A
T 14,56 ms
R 0,95 mR

I have a few photo's of the spotwelder
 

Attachments

  • 20170908_142827_resized-1.jpg
    20170908_142827_resized-1.jpg
    88 KB · Views: 4,660
  • 20170908_142848_resized.jpg
    20170908_142848_resized.jpg
    102.8 KB · Views: 4,660
Package arrived today, and the parts were well-packed. Now I need to find a way to get some days off work!

The parts were not difficult to figure out what parts went where. Start with the most obvious parts and it becomes easy very fast. Install the four longer standoffs under the board first, because the digital display covers one of the screws. I had to pull the display off after I installed it, install that screw and standoff, and then re-install the display. Not hard , but...perhaps this note here will save you a minute of frustration. I will post pics here soon.
 
A customer had reported a kWeld that had its main voltage regulator broken (see picture). He is using a 6S Lipo instead of the recommended 3S. After some research, I realized that I paid much attention to the robustness of the power switch while overlooking what is happening at the battery terminals. Of course that also sees high voltage spikes from inductive kickback, which is exceeding the main regulator's input voltage ratings.

Although it never happened to me during all my experiments that the regulator broke down, I highly recommend applying the modification that I summarized in the attached document. If you do not have the resources to do this by yourself, then of course I offer to have the unit sent back to me, modified, and delivered back to you, all free of charge.

kweld-broken-reg.jpg
 

Attachments

  • kWeld r2 - input protection.pdf
    218.2 KB · Views: 428
I am working on a stationery power supply for kWeld. It will consist of
- a 1000W server power supply, HP DPS800GB. These are available in used condition via ebay for 20€.
- a capacitor charging module for the above. I have a first working prototype on my bench (see pictures), and I plan to sell them for 70€. The board basically converts the server PSU into a high power laboratory supply with adjustable voltage and current 0-12V / 0-80A. The 80A is the design target, I'll have to see how close it will get to delivering that continuously. The design is temperature controlled and will reduce current when reaching its limit.
- a capacitor bank with 6x BCAP0310 P270 T10, passive balancing, and fan cooling.

If anyone is interested, please post here or send me a PM.

guesses-on-what-i-am-attempting-here

guesses-on-what-i-am-attempting-here

guesses-on-what-i-am-attempting-here

guesses-on-what-i-am-attempting-here
 
The capacitor charger test load has arrived, and I couldn't resist to make a first test with a higher load. Results: 35A at 8.5V -> 297W. Hottest spot of the charger are the large heat sink and the inductors: 50°C. It amazes me how small power circuits can be with modern components 8)

Sadly, I had to stop here because the server PSU seems to have problems, it reaches 105°C at one spot of its PCB. I'll have to solve that before continuing. I have another ten bulbs waiting (these are 50W).

guesses-on-what-i-am-attempting-here


guesses-on-what-i-am-attempting-here
 
tatus1969 said:
It was quite some work, but I have managed to ship all orders in time (week 34) as promised! Thank you for your patient waiting, and I am curious about your feedback.

Thanks! Are you going to do another run of these? I'd love one!
 
Ingineer said:
Thanks! Are you going to do another run of these? I'd love one!
I'm busy with that. I have sourced almost all loose components, many of them are already here in large quantities. However, I have not finished the board redesign yet, I expect that I can complete that this weekend. I think that I can open my shop for pre-orders by next weekend.
 
I've been thinking about weather an AVR would be capable of something like this. The ADC is nowhere near as fast (15kHz vs. 1MHz) and slightly lower resolution (10-bit vs. 12-bit) and there's only 1 instead of 2, but I'm wondering if all that horsepower is really needed. What are your thoughts on this?
 
Back
Top