kWeld - "Next level" DIY battery spot welder
- Thread starter tatus1969
- Start date
ElectricGod
10 MW
john61ct said:
Sure...if you don't mind wearing bulky gloves while doing fine motor skill work.
The welding probes get bloody hot after a steady bit of spot welding. The higher the joule setting the hotter they get quicker. Give it a bit and you can't hold the probes anymore, have to stop and let them cool down. Insulating the probes so you can keep welding and still hold something slender and easy to manipulate is key here. Heavy gloves won't cut it since they are so bulky. I've already tried 5 layers of heat shrink and I have to stop after a while to let things cool down. That's still better than the single layer of heat shrink they come with.
AKA...needing better insulation on the welding probes. Aerogel infused cloth ought to do the trick to keep the heat away from my fingers.
The new goal will be can the probes get hot enough to melt solder at the 8 awg wires and I'm still welding.
ElectricGod
10 MW
Both my super cap modules have conformal on the small electronics now and I added balance leads to them. This is the second module. Sure enough they are all pretty close to the same voltage. So far after quite a lot of spot welding, the caps have worked fine at 9.44ish volts on both boards.
I was part way into adding conformal to the super cap module that I destructively tested previously when I noticed a missing resistor. While I was taking it apart to add a balance cable and conformal, I accidentally fumbled the screwdriver and it struck the board. At the time I didn't notice anything. It was later when I was applying conformal that I noticed this. Clearly this is where the screw driver hit and it tore off the resistor at R31. Looking at the other board it is labeled 657 or 650,000,000 ohms or 650 meg ohms, but that seems unlikely. That's almost an air gap! 657 ohms seems more likely.The resistor is very tiny. I've looked all over my bench and don't see it. It's so tiny, spotting it is virtually impossible. I'll try to scrounge one off some scrap board I have. I've torn a trace off the board too. It's totally repairable, but still...grrrrr!
Completed precharger and discharger:
I wanted to follow the KISS principal and use the same resistor bank for graceful charging and discharging. The red wire is screwed down to PSU+. The black wire gets plugged into the negative side of the super cap modules only when I want to discharge. THe 8mm bullet connects to the positive terminal on the super cap board. The black wire is also a direct path back to PSU+. Since the negative wire from the PSU is disconnected to connect this wire, there is not a complete circuit path back to the PSU if it is on or off. If the super caps are charged, they can't back feed into the PSU via the positive wire only. The rocker switch when on bypasses the resistor bank for precharging. It has 3 30 amp parallel diodes in line with it. The diodes would be reverse biased when discharging. This prevents me from mistakenly leaving the rocker switch on when discharging and create a dead short across the super caps. With the switch on or off, all current flow to discharge has to pass through the resistor bank. When charging the super caps via the PSU, the discharge wire is unplugged and the PSU- wire is plugged in again. IF by mistake I did leave the discharge wire plugged into CAP-, nothing would happen. There wouldn't be a complete current path between the PSU and the super cap board. It's simple and pretty fool proof. It's far simpler and uses 1/2 as many parts as my other idea for graceful charge and discharge. The physical disconnect of the precharge wire could have been done via a second rocker switch, but I had only 1 rocker switch with contacts that could carry the current so the physical disconnect was a simple solution.
Here it is plugged in between the PSU and super cap module. It's hard to see, but the PSU- wire is plugged into the super cap module. The discharge wire is hanging next to it. There's no exposed current path from the discharge wire to anything else so letting it hang there is pretty safe. To gracefully precharge, I turn off the switch. Once close to fully charged, I close the switch. One detail I need to check is if I'm limiting current flow through those 3 diodes. It doesn't seem like it as the super caps charge pretty rapidly once I close the switch with or without the diodes in there. I don't think they make much off a difference. The super caps still charge up to 9.44 volts.
I was part way into adding conformal to the super cap module that I destructively tested previously when I noticed a missing resistor. While I was taking it apart to add a balance cable and conformal, I accidentally fumbled the screwdriver and it struck the board. At the time I didn't notice anything. It was later when I was applying conformal that I noticed this. Clearly this is where the screw driver hit and it tore off the resistor at R31. Looking at the other board it is labeled 657 or 650,000,000 ohms or 650 meg ohms, but that seems unlikely. That's almost an air gap! 657 ohms seems more likely.The resistor is very tiny. I've looked all over my bench and don't see it. It's so tiny, spotting it is virtually impossible. I'll try to scrounge one off some scrap board I have. I've torn a trace off the board too. It's totally repairable, but still...grrrrr!
Completed precharger and discharger:
I wanted to follow the KISS principal and use the same resistor bank for graceful charging and discharging. The red wire is screwed down to PSU+. The black wire gets plugged into the negative side of the super cap modules only when I want to discharge. THe 8mm bullet connects to the positive terminal on the super cap board. The black wire is also a direct path back to PSU+. Since the negative wire from the PSU is disconnected to connect this wire, there is not a complete circuit path back to the PSU if it is on or off. If the super caps are charged, they can't back feed into the PSU via the positive wire only. The rocker switch when on bypasses the resistor bank for precharging. It has 3 30 amp parallel diodes in line with it. The diodes would be reverse biased when discharging. This prevents me from mistakenly leaving the rocker switch on when discharging and create a dead short across the super caps. With the switch on or off, all current flow to discharge has to pass through the resistor bank. When charging the super caps via the PSU, the discharge wire is unplugged and the PSU- wire is plugged in again. IF by mistake I did leave the discharge wire plugged into CAP-, nothing would happen. There wouldn't be a complete current path between the PSU and the super cap board. It's simple and pretty fool proof. It's far simpler and uses 1/2 as many parts as my other idea for graceful charge and discharge. The physical disconnect of the precharge wire could have been done via a second rocker switch, but I had only 1 rocker switch with contacts that could carry the current so the physical disconnect was a simple solution.
Here it is plugged in between the PSU and super cap module. It's hard to see, but the PSU- wire is plugged into the super cap module. The discharge wire is hanging next to it. There's no exposed current path from the discharge wire to anything else so letting it hang there is pretty safe. To gracefully precharge, I turn off the switch. Once close to fully charged, I close the switch. One detail I need to check is if I'm limiting current flow through those 3 diodes. It doesn't seem like it as the super caps charge pretty rapidly once I close the switch with or without the diodes in there. I don't think they make much off a difference. The super caps still charge up to 9.44 volts.
Murphy got you, because that's a 6k57 0.1% precision resistor, which you most likely don't have in your drawerElectricGod said:
That's part of the internal voltage reference generation and must be that exact value and precision. The resistor connects to that third IC pin from the left, and to R32 right to it (directions according to your picture).
ElectricGod
10 MW
tatus1969 said:Murphy got you, because that's a 6k57 0.1% precision resistor, which you most likely don't have in your drawerElectricGod said:
lol...figures...the one resistor of an impossible to scrounge value...
Thanks for the info...ordering some 6.57K .1% resistors from Mouser.
Welp! Back to using a single super cap board again. Not that for what I'm doing right now do I need two in parallel. That was more about testing them both at the somewhat higher voltage I run at. The fans run at low RPM's for a few seconds after lots of continuous welds. By then the welding probes are too hot to hold anyway so I have to stop and let them cool down.
I'm absolutely pleased with the power and consistency of the KWeld.
The only things I'd change or add from the existing kit is:
Adding a few bullets at some strategic locations so it comes apart easily
Insulated welding probes
This is more a safety thing, but a warning sticker on the super caps so noobs don't do what I did deliberately.
All my mods are about me getting it set up exactly how I want it to work and having easy access to check stuff.
My own deliberate screwing around and breaking the one super cap board twice, has nothing to do with the KWeld goodness. That's ALL on me!
ElectricGod
10 MW
The aerogel fabric arrived Saturday, but so far I've had no time to do anything with it.
It's 1/4" thick, but then it's also rather compactable so I think it might compress to more like 1/8" thick.
It's 1/4" thick, but then it's also rather compactable so I think it might compress to more like 1/8" thick.
ElectricGod
10 MW
I bought some .3mm x 8mm nickel since the KWeld can handle it. The guy I got it from told me about a new welding strip he will be carrying soon.
It's nickel, stainless steel and copper. I asked if he could send me a sample of it so he sent me 12" of it. He has no supply yet so this with a lot of his very limited stock.
I wish I had widths and thicknesses all the same to make things a bit more obvious.
Top to bottom:
SS/Ni/Cu .1mm x 9mm 11.1 mOhm
Ni .15mm x 7mm 20.9 mOhm
Ni .3mm x 8mm 9.85 mOhm
As expected the .3mm nickel has the lowest resistance. But check out that .1mm thick alloy! Not much more resistance despite being 1/3rd as thick. This alloy is supposed to have significantly less resistance thanks to the copper content than pure nickel and not oxidize or corrode like copper thanks to the SS and nickel.
There is a very slight color difference between pure Ni and the alloy. The alloy is a bit stiffer than the .15mm nickel. Looking at them side by side, if I didn't know which one was the alloy, I would not be able to tell them apart.
I want to see about salt water resistivity next. I've left pure nickel all scratched up in salt water for a week and it was unharmed at all. I think since I have 3 thicknesses of metal, I'll scratch sections of each one and drop them all in the same salt water and see what happens. I might even have some nickel plated steel.
To keep glare from the camera flash off the metal, perspective makes these 3 strips look like they are different lengths, but they are all 12" long.
Close-up on the values:
It's nickel, stainless steel and copper. I asked if he could send me a sample of it so he sent me 12" of it. He has no supply yet so this with a lot of his very limited stock.
I wish I had widths and thicknesses all the same to make things a bit more obvious.
Top to bottom:
SS/Ni/Cu .1mm x 9mm 11.1 mOhm
Ni .15mm x 7mm 20.9 mOhm
Ni .3mm x 8mm 9.85 mOhm
As expected the .3mm nickel has the lowest resistance. But check out that .1mm thick alloy! Not much more resistance despite being 1/3rd as thick. This alloy is supposed to have significantly less resistance thanks to the copper content than pure nickel and not oxidize or corrode like copper thanks to the SS and nickel.
There is a very slight color difference between pure Ni and the alloy. The alloy is a bit stiffer than the .15mm nickel. Looking at them side by side, if I didn't know which one was the alloy, I would not be able to tell them apart.
I want to see about salt water resistivity next. I've left pure nickel all scratched up in salt water for a week and it was unharmed at all. I think since I have 3 thicknesses of metal, I'll scratch sections of each one and drop them all in the same salt water and see what happens. I might even have some nickel plated steel.
To keep glare from the camera flash off the metal, perspective makes these 3 strips look like they are different lengths, but they are all 12" long.
Close-up on the values:
ElectricGod
10 MW
tatus1969 said:
I'm not welding the alloy yet, but I'll get to trying it soon. Since the metal is so much thinner, I'm betting weld power needs to be reduced.
This is my test based results for good welds and not blowing through the bottom of 18650 cells.
.15mm = 50 joules
.3mm = 100 joules
So I'm guessing I need 30-40 joules for the .1mm alloy. But then since it is part copper and SS...who knows!
I've never welded .3mm before. It takes a lot of energy to get decent welds and even then they are not as strong as the .15mm welds. 50 joules and .15mm dies not come apart without pliers. .3mm can be broken off the cell with some semi aggressive wiggling. I tried all the way up to 300 joules last night and that didn't seem to improve the strength over 100 joules. Maybe a single super cap module is unable to provide the current needed?
On a side note, the 50 amp PSU keeps up nicely with 50 joule welds, but starts lagging behind on 100 joule welds like is needed for the .3mm nickel. My weld cadence at 100 joules has to be slower...or at least it is while I'm running a single super cap board. I haven't checked what the amp draw is through the 3 diodes. I don't care too much, but ideally I want to maintain that 50 amps I was seeing without them. The diodes were old and scrounged from who knows what or when. I have lots of much newer half wave rectifiers. I think I'll try a few of those.
PaulD
1 kW
ElectricGod said:
Sounds like EMS SigmaClad : https://www.emsclad.com/solutions-by-sector/energy-storage.html
I'm assuming it's clad, not an alloy. If so, it's been around for a while, but a bit expensive. It's used in less cost sensitive high power stuff. It welds nicely considering it's conductivity due to the stainless layers.
ElectricGod said:
I think that when you get to the 0.3mm nickel or to pure copper it is important to have lots of current. Even when energy is constant it makes some difference if the 50joules is delivered in 50 or 100ms pulse. I have also noticed that after certain point more joules don't chance results. I assume that is because weldtime gets so long that it just starts to spread the heat to larger area..
ElectricGod
10 MW
tatus1969 said:
I have current set to 900 amps. So... go higher?
ElectricGod
10 MW
PaulD said:
I specifically said alloy, not laminate or plated or clad.
There's no layers that I can see and it was not described as anything other than as an alloy of metals.
If you feel like you need to. 2000 amps is the max allowed but some naughty people have gone over it and for some reason I think you will try it too :lol:. Calibration will not be possible at over 2000 amps but I hear it still welds..ElectricGod said:
ElectricGod
10 MW
ossivirt said:
Whaaaat?! ME naughty? I resemble that remark! hahaha
900 amps is more than enough for .15mm nickel. It was also enough to weld .5mm soft steel to .5mm soft steel. I assumed it would also do the job for .3mm nickel. I'll set current to 1500 amps and then mess with the joule setting. The current setting is a maximum, not what you are actually using...which may be far below the maximum.
Infernal Bill
10 µW
- Joined
- Oct 17, 2019
- Messages
- 5
Sorry for a basic and specific question, but I'm new to this forum and my research over the past week lead me to the KWeld (which I purchased yesterday). I've read through this thread and the KWeld manual as well, but I don't have a formal education in electronics. I'm a hobbyist. I'm going to be using the KWeld to make battery packs for my long range RC planes.
I'm hoping to get some input on what thickness I should use for the nickel tabs. I found the chart on the first page of this thread: https://endless-sphere.com/forums/viewtopic.php?f=14&t=68005&hilit=spot+welder
My power use (long range RC Planes) typically falls along the following lines: 1) launch for 10 seconds at full throttle and pull around 60 amps, 2) cruise for 1 hour to 1.5 hours drawing about 10 to 15 amps on average, 3) occasionally draw around 20 to 25 amps when climbing. I've been using pre-made Li-Io packs for years in the above application so I know they work (and work far better for me than Li-Pos for my purposes).
Based on the chart, am I looking at .3mm Pure Nickel only? Some posts in this thread note that while KWeld can manage that thickness, there are some issues (e.g., heat, possibly killing my Li-Po power source (HK 3S 5A Graphene)). Seems easier on KWeld to do .2mm tabs but not quite sure.
Any thoughts on this are most appreciated before I get too far along one path. These planes can sometimes fly way out over the ocean, so a battery fail in flight would obviously be an unfortunate end to quite a bit of equipment. Thanks in advance.
I'm hoping to get some input on what thickness I should use for the nickel tabs. I found the chart on the first page of this thread: https://endless-sphere.com/forums/viewtopic.php?f=14&t=68005&hilit=spot+welder
My power use (long range RC Planes) typically falls along the following lines: 1) launch for 10 seconds at full throttle and pull around 60 amps, 2) cruise for 1 hour to 1.5 hours drawing about 10 to 15 amps on average, 3) occasionally draw around 20 to 25 amps when climbing. I've been using pre-made Li-Io packs for years in the above application so I know they work (and work far better for me than Li-Pos for my purposes).
Based on the chart, am I looking at .3mm Pure Nickel only? Some posts in this thread note that while KWeld can manage that thickness, there are some issues (e.g., heat, possibly killing my Li-Po power source (HK 3S 5A Graphene)). Seems easier on KWeld to do .2mm tabs but not quite sure.
Any thoughts on this are most appreciated before I get too far along one path. These planes can sometimes fly way out over the ocean, so a battery fail in flight would obviously be an unfortunate end to quite a bit of equipment. Thanks in advance.
