Wow that is a lot different than my settings. I think mine is at around 1100a and 27J gives a good weld single layer and i started tacking down the second layer at 50j, seems to be ok.
4 Likes
Around the same here in regards to the first layer. No idea how 8 could be enough lol
Starts ripping holes in the Nickle when pulled of with pliers at 6j but only on 1 terminal constantly, 8j rips a hole in the Nickle on both contacts for me
I found my biggest bottle neck point of my set up was the XT90 splitter so I changed it to XT150 this is still the part theat generates the most heat for me but it barly gets warm with normal welding now if I take my time.
I think amp delivery has a large diffrance in the 1200-1800amps from what I have seen I would be interested to see if this has any thing to do with voltsage as it’s been mentioned 4s has better results for other people.
So it’s constant with almost doubling the jules for 2nd layer. Good to know otheres come found the same thing and I’m not doing some thing strange.
Just use some fine grit one, the nickel gets off easy
I use 60J on nickel to nickel, the power source wont make a huge difference because kWeld makes the same weld on 2 different batteries more or less, i mean there will be a small difference but nothing to be worried of
2 Likes
This is your opinion not the opinion of me or the creator of the system.
Pleas back this comment up with numbers, facts and pictures
1200a vs 1600 is 30% extra power
27j vs 8j is 35% ish
Would be interesting to see the time it took to deliver the energy
It is stated that some people have had results with power supply that can deliver as low as 800a but a miniumum of 1000 amps is recommended. Higher is better up to the limit of 2000amps the welder can handle why because.
A higher supply deliveres the energy in a shorter amount of time resulting in more consultant welds and less heat spreading out over a wider Area. The question is when dose diminishing gains happen.
Using a supply with 3/4 of the ampage delivery requires 3 times the energy needed to do the weld. I don’t see using a better power supply as diminishing returns.
I don’t think you need 1800 amps for 0.15 as the work around is to apply more energy but no harm in useing the advantage if you have it as it’s better.
It’s a much larger difference than I would have expected but is this to do with the curve or is there other factors to consider more pressure cleaner tips lower resistance of build.
1 Like
Im still working on it, will let you know the time
Edit: dunno what one is the time 
which reading is it?
I have the best solution you can thank me later.
After trying different sandpapers I found out that pure alcohol (methylated spirit) cleans like a champ without the need of sandpaper. When the electrode(s) start to stick on the nickel during welding put some spirit on a kitchen roll and clean/rub the tip and you are ready to weld again.
This 0.2mm nickel was welded with 30J
Lipo is nano-tech 3s 5Ah 65-130C
Here some pics from tests with 0.2mm nickel on a blade
bottom side of the blade
after removing the nickel with pliers
The wood underneath the blades got burned spots from welding
0.3mm nickel is also no problem above 50J but everything gets hot quicker.
Trying to remove 0.3mm nickel (80-100J) from a blade with pliers, impossible task.
One important thing I noticed is the angle of the electrodes during welding. It’s best to hold them with about 30° angle. Once I did some welds and they were poor with 30J. Even with 40J it didn’t look good and I thought there’s something damaged. Then I realized I held them with too less angle (maybe 15° only). After correcting the angle and went down to 30J it was perfect again, the welding spots are much bigger.
10 Likes
If you hold down the pedal, it’s the one that ends with ms.
So god damn stoked to get mines up and running. Finally got the battery from Hobbyking. Building stuff bitches!!! Well fixing stuff first …thanks Unik POS battery.
Thanks for all the info gents
1 Like
T = 50.62 for 0.15 to .15 50j and 1150a
from my experience of 5ah green and red lipos the difference is around 400A, green was at around 1800 while the red lipos were around 1400A, both batteries got the same welds more or less
I was saying that above 1300A or so the results will be always good, up to I would say around 70J welds
Not that a 1000A and 1800A power source work the same
2 Likes
Why don’t you just solder the two male XT90’s right to the input of the welder? You could eliminate the XT150 and additional wiring entirely, and still have a system with detachable batteries. What is that black thing left to the upper XT150 in the pictures? Just a sleeve, right?
Yes, based on my tests here. I had been using solid strips of pure 0.3mm nickel, and it doesn’t weld that reliably here. It depends on what you actually use - Hilumin (nickel plated steel) welds significantly easier, and slotted strips help guiding the current and also improve the situation.
It’s really just Ohm’s law. Desired current is 1500A. The resistance of the kWeld system is roughly 3.2 milliOhms (stock cables). The weld spot itself contributes another 1 milliOhm typical. This gives 4.2 milliOhms total. To push 1500A through that, you need U = R*I = 6.3V from an ideal battery. For every milliOhm of battery’s internal resistance, you need another 1.5V accordingly. Or in other words, for every volt above 6.3V that the battery has, it needs to have 1 milliOhm of internal resistance to stay in the sweet spot for the current. A 3S battery therefore needs to have 12-6.3 = 5.7 mOhms. Then calculate the power loss in the battery: P = 0.5 * R * I^2 = 6412 W. That explains why they heat up quickly, even though the pulses are short. Now consider a 6S battery. That must have an internal resistance of 24-6.3 = 17.7 mOhms. Then P = 19912W. You see that the power loss in that battery is huge. This is why I recommend 2S to 3S batteries for the welder.
Having the sides touch instead of the tips gives you an incorrect calibration result. It is very importan to have the tips touch correctly (pushing the conical areas is okay). Make sure you get no sparks, otherwise repeat. This is important for the energy metering to work properly.
That should definitely be possible without problems, especially because you are running the welder at higher amperage than I do here, and it works well here. The result also depends on the type and thickness of the battery can itself, is it possible that your batteries have aluminium cans? kWeld can definitely not weld nickel strips to that material.
Another important advice (sorry the user manual lacks this info): kWeld does not like voltage applied to its electrodes; this can happen when accidentally touching different poles of your (large) battery under construction. Please be careful to avoid that, otherwise you may kill the unit.
That’s public on the website:
keenlab UG
Langestrasse 9
26388 Wilhelmshaven
Germany
Much appreciated from my side!
If you have, please make tests with slotted strips. You should get better results with that. Nevertheless, I share your opinion that welding overly thick strips to your battery requires a lot of heat that can potentially damage the cell internals. What happens to the back of your blade also happens to the battery. When a customer asks me, then my recommendtion always is to rethink the solution and check if thinner material might be the better option.
Exactly. These are officially called “resistance spot welders”, as the operating principle is to push current through the electrical resistance of the material to weld.
It’s voltage that causes current, and it’s current that causes heat. So the right term would be ‘current’ punch. I hope that I could clear up that relationship in my above calculations 
That’s a clear sign of bad calibration, please repeat that and check. 8J barely be enough to weld 0.1mm nickel.
The joule metering algorithm automatically adapts to varying situations (battery voltage, current, corrosion,…) A higher current allows to weld thicker material, and it keeps the pulse duration shorter. The latter is desirable as it avoids heat spreading out during the pulse. That causes some variation, and is generally good to experiment to find the correct setting before making large batteries. Do this for each type of connection: strip to negative, strip to positive, strip to strip. Different types of batteries again can require varying this, because they may have cans of different thickness for example.
It’s important to get as little arcing as possible, hence push hard and work with clean surfaces where possible.
4 Likes
Dident want to have a floating live plug when using the kcaps (when cutting Nickle bits go every we’re and I’m scared some day it will short a battery) did suggest it earlya in this thread for any one thinking of doing more than 0.2mm
also with the warmest part of the unit after use was the XT90 (I presume this is because this is the part of the system has the hiest resistance) so I thought using a XT150 with the kcaps in the futere was a good idea ones you sorted it out.
the XT150 neg conection?
its a dule XT90 to XT150. 2 AWG 10 cables striped is a super tight stuff in to a XT150 and there is no way the insulation is going in the XT150 case so you have to strip extra so multiple layers of shrink tubing around the double AWG10 silicon cables next to the XT150
I’m consistently getting 2.15 with the XT150 use to get 2.35 with the XT90 before all with stock cables. is this some thing to be concerned with how low is some thing off and why im welding with such low Jules setting? I think my battery cells resistance from memory is 1.1-1.4 milliohm each resistance and I have 3 in series 2 sets of them in parallel. so 1.2*3=3.6 derided by the 2 packs 1.8 mill ohm thats leaves 0.35 mill ohms of resistance for the Kweld system and probes seems to low even for a bad calibration unless I’m doing maths wrong some were.
I take it then the dule 3s 6ah Turnigy green Graphene delivering 1600-1800amps im using is pretty much the best power source I could be using no point in changing as a few other options was discussed earlya
the 0.15 nickel welds grate to it so definitely not Ali can and im struggling with craft blades below 100j starting to think I need to try 0.3mm suppler that’s not nikon
found it after posting sent it beginning of last week hopefully its arrived
Got it I was afraid that this would be some kind of fuse or whatever.
The number should be independent from the amount of current that flows. Thinking of this, it could also be a flaw in the algorithm. The welder waits for the current to ramp up before starting to make measurements. The time that this needs depends on the amount of inductance in the system, and the waiting time may be too short. This would explain why the measured resistance varies in your case. Do you have the firmware updater so that we can experiment with this?
I’m afraid yes The displayed resistance is that of the output of the kWeld unit only (output cables, electrode holders, electrode tips). It is very important to press the conical ares of the tips together, as the wanted result is all resistance that the system later has when welding. Also make sure that the set screws are tight (don’t overtorque, as it is copper), maybe also check all the screws. In rare cases a customer even had to disassemble the bus bars to clean the surfaces. They need to make good contact over the entire surface, or the accuracy of the current measurement is reduced.
There must be something wrong with your unit or setup, otherwise you wouldn’t need to dial in such low Joule values to get the wanted result.
Arrived this week, I’ll inspect it early next week.
2 Likes
Yes some were but I only have my MacBook with me for the next few weeks windows PC at home if I need to run software.
So not like a Zs then calculating the supply in as well. As I thought it did.
In my head the lower the resistance of the calibration the better a bad assembly would add resistance not take it away.
I did check all the connections and tweak all the bolts when I opened it up when adding the XT150 connectors would explain the .2 drop (may or may not have soldered the red and black connectors the wrong way round and it’s easya to unbolt than re solder didn’t want to forget one day and power it up with the - and + reversed)
I think this is the correct one
I did 10 calibrations pushing the tips together and I’m getting 2.16 mR pretty consistently unless I don’t push them together so hard then iv getting 2.20ish
Doesn’t seem to matter what battery I use if it’s a 6Ah or 12Ah only affects the amps delivered as expected.
3 different Nickle pieces non of them rusted after 5 days in salt water 1 and 2 are 0.15 3rd pice is 0.2
Perfect! That’s what I had expected. In that case, no need for experiments. And yes, that’s the firmware updater.
So iv not Heard any thing back from this since you revived the return back over a month ago unless iv missed some thing. Any update?