kWeld - "Next level" DIY battery spot welder

I'm up and running with my kWeld using the recommended Turnigy Graphene 5.0Ah 75C https://hobbyking.com/en_us/graphene-5000mah-3s-75c.html . Results thus far have been absolutely amazing. Now I get it. Wow!! :)

I've noticed that after a row of 12 welds @50J on .2mm x 8mm nickel strip using the super convenient automatic mode (another wow!) that the probes and leads get pretty warm and so does the battery. I don't have a good feel yet for how warm is too warm and thus time for a cool-down.

Similarly, I see there's an adjustable low voltage alarm for the input battery but I don't know what I should set it to to protect the battery for a long service life.

How do others monitor and care for their LiPo input batteries?
 
tatus1969 said:
SubnetMask said:
@tatus1969, what are your thoughts on Headway LiFePo4 cells and the kWeld?
Click to expand...
Yes these are much safer and can be used. You'll need to run the calculations with regards to voltage (should be ~ 10 to 15V) and the pack's internal resistance (~ 3 to 5 milliohms) however, and that's where it becomes tricky. The cell ESR numbers can be unknown, measured poorly, or tweaked for marketing. If you can afford experimenting (starting with low cell count to avoid overcurrent) then there's nothing against that approach. I don't know however how well they can withstand the repeated abuse from being short circuited by the welder, as that could drastically reduce their lifetime.
Click to expand...

Just so I'm clear - even though the kWeld can go down to 4v, suggesting that these in a 2s arrangement should work, they would really need to be in a 4s arrangement because the kWeld would actually want the voltage higher than 6v on batteries?


Unrelated to that, just a thought - if you ever do another PCB redesign or adjustment, perhaps move the connector for the foot switch closer to the edge and use a socket like this with this plug to allow easy disconnection of the foot switch while the kWeld is enclosed in the case. With the current setup, if you wanted to remove the foot switch, or if you had the connector jumped for 'permanent' automatic mode, you'd have to take it out of the case to disconnect the foot switch or jumper.
 
Why would you want to disconnect it anyways? I always roll up mine when I'm done with it. You could also solder a barrel plug inline if you really needed to remove it every time you use the thing.
 
I quess it would be neat for storage to disconnect it but since it leaves electrodes and battery leads still hanging there its not that big deal. And making plugs for high power wires is more expensive and reduces performance :roll:

Or you can just make nice box/case for it that has place for welder, pedal etc and has nice foam inside like if you would buy expensive tools 8)
 
JoelR said:
I'm up and running with my kWeld using the recommended Turnigy Graphene 5.0Ah 75C https://hobbyking.com/en_us/graphene-5000mah-3s-75c.html . Results thus far have been absolutely amazing. Now I get it. Wow!! :)

I've noticed that after a row of 12 welds @50J on .2mm x 8mm nickel strip using the super convenient automatic mode (another wow!) that the probes and leads get pretty warm and so does the battery. I don't have a good feel yet for how warm is too warm and thus time for a cool-down.

Similarly, I see there's an adjustable low voltage alarm for the input battery but I don't know what I should set it to to protect the battery for a long service life.

How do others monitor and care for their LiPo input batteries?
Click to expand...

Exact same questions for me. : - )
 
To reduce the heating up of the battery, parallel a second, identical battery when welding. You can also aim a small fan at the pack when working. It won't help cable heating though. When a Lipo is at 3.7 Volts per cell - resting, not under load, it is down to about 18-20% capacity remaining. Best not to drop below 20%. Depending on how fine you can adjust the alarm, (I have not used it) I'd set it to between 3.7 and 3.73 Volts Per cell. Also, don't leave a Lipo in a fully charged state for any length of time - even over night. It kills them. Best to leave it at "Storage Charge" which is about 3.75 Volts Per Cell when you're not using it.

Hopefully, you're using a balance charger so that all cells stay balanced.
 
JoelR said:
leads get pretty warm and so does the battery. I don't have a good feel yet for how warm is too warm and thus time for a cool-down.
Click to expand...
Please make sure the battery doesn't exceed 50degC in order not to further reduce its lifetime. In most cases, the necessary cooling periods for the electrodes are enough to ensure that, but it is worth double checking.

JoelR said:
Similarly, I see there's an adjustable low voltage alarm for the input battery but I don't know what I should set it to to protect the battery for a long service life.
Click to expand...
@BVH, fully agree. But there is a second option - you can set the undercurrent cutoff (menu "MinCurr") such that, as soon as the current drops to lets say 80% nominal, the welder will abort.

SubnetMask said:
perhaps move the connector for the foot switch closer to the edge and use a socket
Click to expand...
Thanks, and that's on the list among other things for a redesign. At the moment I don't have much time for this, in favor of other projects. I'm working on something big actually, a novelty. :)

SubnetMask said:
Just so I'm clear - even though the kWeld can go down to 4v, suggesting that these in a 2s arrangement should work, they would really need to be in a 4s arrangement because the kWeld would actually want the voltage higher than 6v on batteries?
Click to expand...
The battery needs to push enough current through the loop, it's literally Ohms law. More voltage means more current for a given resistance. On the other hand, too much voltage requires additional resistance and therefore waste of power. With the stock cabling, a welding pulse typically needs about 5 volts at the kWeld's input terminals, the rest is 'lost' in the battery. That's where the 12V "sweet spot" comes from.
 
Just received my new Kweld unit and I’m almost ready to start. Has anyone used these Tungsten electrodes from Sunstone? This was the closest size I could get to the 5 mm electrodes that came with the kit. Any feedback is welcome. Thanks

2C0131BA-0A85-4D57-984D-DA263668C61C.jpeg
 
Nothing wrong with tungsten electrodes. Be aware they have a much higher resistance than the copper electrodes, so they will get very hot. The benefit of tungsten is that it has a very high melting point, so if the copper electrodes are sticking to the work, the tungsten will not.

A third option is to buy electrodes that are half-copper and half tungsten. Slightly lower price, too.
 
spinningmagnets said:
A third option is to buy electrodes that are half-copper and half tungsten. Slightly lower price, too.
Click to expand...
Did not think of that idea. Makes perfect sense. Do you have a recommended source? I will do some research as well. As always thanks for sharing your knowledge. Gary.
 
Got my kWeld yesterday and got it all together, just waiting on the power source now.

For the power source, I ordered two Turnigy Nano-tech 5000mAh batteries, with the intent to run them in parallel to reduce the 'abuse' on each battery, but I've read what seem to be conflicting information. I've read posts saying that you can't calibrate it with both attached, but it'll work fine. I've read that a single pack should be around 1300-1500A, so if two only increases by 50%, that's still bringing it real close to the 2000A limit, if not over. And I've seen pictures in a thread that Frank has posted in of the same setup I have/had planned - a Y cable with a XT90 on one end of each for the batteries and the other end of both soldered to a XT150 to power the kWeld. Is running the two packs in parallel going to be OK for the kWeld?
 
SubnetMask said:
Got my kWeld yesterday and got it all together, just waiting on the power source now.

For the power source, I ordered two Turnigy Nano-tech 5000mAh batteries, with the intent to run them in parallel to reduce the 'abuse' on each battery, but I've read what seem to be conflicting information. I've read posts saying that you can't calibrate it with both attached, but it'll work fine. I've read that a single pack should be around 1300-1500A, so if two only increases by 50%, that's still bringing it real close to the 2000A limit, if not over. And I've seen pictures in a thread that Frank has posted in of the same setup I have/had planned - a Y cable with a XT90 on one end of each for the batteries and the other end of both soldered to a XT150 to power the kWeld. Is running the two packs in parallel going to be OK for the kWeld?
Click to expand...
do not charge the battery beyond 4V, that should keep the amps under 2k.
 
SubnetMask said:
For the power source, I ordered two Turnigy Nano-tech 5000mAh batteries, with the intent to run them in parallel to reduce the 'abuse' on each battery, but I've read what seem to be conflicting information. I've read posts saying that you can't calibrate it with both attached, but it'll work fine. I've read that a single pack should be around 1300-1500A, so if two only increases by 50%, that's still bringing it real close to the 2000A limit, if not over. And I've seen pictures in a thread that Frank has posted in of the same setup I have/had planned - a Y cable with a XT90 on one end of each for the batteries and the other end of both soldered to a XT150 to power the kWeld. Is running the two packs in parallel going to be OK for the kWeld?
Click to expand...
kWeld has overcurrent protection (2000A), but if the battery is way too strong then it can't react quickly enough anymore. Two of them in parallel will be borderline but may work, but I suggest to start with 50% charge state (reduced voltage). When running at high amperage, make absolutely sure not to extend the cables. The single-battery calibration trick actually works, because it requires directly shorting the probes whereas regular welding has additional resistive material in the path.
 
Got new Kweld. Awesome machine. Followed setup instructions for calibration but I do not get a resistance reading after pressing the electrodes firmly together and pressing the knob when the “Short” message is displayed. I can feel the pulse but instead of an IR rating it just displays “10J”. Any ideas? Looking forward to using this awesome machine. Thanks. Gary
 
garolittle said:
Got new Kweld. Awesome machine. Followed setup instructions for calibration but I do not get a resistance reading after pressing the electrodes firmly together and pressing the knob when the “Short” message is displayed. I can feel the pulse but instead of an IR rating it just displays “10J”. Any ideas? Looking forward to using this awesome machine. Thanks. Gary
Click to expand...
Can be either an incorrectly wired foot switch (check that you get an 'open' when it is not pressed), or insufficient power from the battery (it aborts calibration with undercurrent when below 800A). In any case, keep the trigger pressed during the steps as described in the manual to get diagnostic data.
 
tatus1969 said:
SubnetMask said:
For the power source, I ordered two Turnigy Nano-tech 5000mAh batteries, with the intent to run them in parallel to reduce the 'abuse' on each battery, but I've read what seem to be conflicting information. I've read posts saying that you can't calibrate it with both attached, but it'll work fine. I've read that a single pack should be around 1300-1500A, so if two only increases by 50%, that's still bringing it real close to the 2000A limit, if not over. And I've seen pictures in a thread that Frank has posted in of the same setup I have/had planned - a Y cable with a XT90 on one end of each for the batteries and the other end of both soldered to a XT150 to power the kWeld. Is running the two packs in parallel going to be OK for the kWeld?
Click to expand...
kWeld has overcurrent protection (2000A), but if the battery is way too strong then it can't react quickly enough anymore. Two of them in parallel will be borderline but may work, but I suggest to start with 50% charge state (reduced voltage). When running at high amperage, make absolutely sure not to extend the cables. The single-battery calibration trick actually works, because it requires directly shorting the probes whereas regular welding has additional resistive material in the path.
Click to expand...

Thanks Frank - when you say 'make absolutely sure not to extend the cables', I assume you mean ANY cable, including the inputs? If so, I'm guessing that rather than make a 'Y' (Or really more of a 'V' cable) with two XT90's on one end for the packs and a XT150 pair on the other to join up with the stock cables with the XT150's, I'd be better off creating this cable as short as possible with XT90's on one end for the packs to ring terminals on the other where they join to replace the stock 8AWG cables? In doing so, I assume I should also adjust the cable length in the settings to match whatever the round trip end sup being from the packs to the electrodes and back?
 
tatus1969 said:
garolittle said:
Got new Kweld. Awesome machine. Followed setup instructions for calibration but I do not get a resistance reading after pressing the electrodes firmly together and pressing the knob when the “Short” message is displayed. I can feel the pulse but instead of an IR rating it just displays “10J”. Any ideas? Looking forward to using this awesome machine. Thanks. Gary
Click to expand...
Can be either an incorrectly wired foot switch (check that you get an 'open' when it is not pressed), or insufficient power from the battery (it aborts calibration with undercurrent when below 800A). In any case, keep the trigger pressed during the steps as described in the manual to get diagnostic data.
Click to expand...

Thank you sir. It worked. It was the foot switch. Measured 2.29 milliohms during calibration. Thanks for the response. Gary
 
My question is very similar to subnetmask .... Can I use the XT90 “parallel” connector as shown in the pictures below? I have not plugged in the paralleled battery packs to the Kweld unit yet. Wanted to ask the question before I made the move. :)

64643DB8-78E0-442B-A027-362B1B159CDD.jpege
096DE436-1242-4044-B459-21F822C71CBF.jpegCDB8EACE-32A5-4549-BC80-375D99FC4F9D.jpegF93DC799-500F-42EF-8612-BC2FF0DBCA53.jpeg021364EB-FA36-4516-BB1B-2557CE25D69F.jpeg
 

Attachments

  • BBC84AA9-50BD-4840-8E3A-7C1E0E261D63.jpeg
    BBC84AA9-50BD-4840-8E3A-7C1E0E261D63.jpeg
    43.2 KB · Views: 1,427
My guess is that the XT90 on the leads into the kWeld (as well as the parallel adapter) would have trouble with the current. What gauge wire is used from that XT90 to the kWeld? It's not going to blow up or melt down just plugging it in, but doing welding could result in 'some heat'. My original plan was to use 10ga wire from each XT90 and join the two at a XT150 connector pair that would connect to the stock provided 8ga wires (the battery packs have 10ga wires, so a heavier gauge from each battery to where they join will have minimal benefit if kept short); possible alternate plan would be to join the two at ring terminals right on the kWeld.
 
SubnetMask said:
What gauge wire is used from that XT90 to the kWeld? ... possible alternate plan would be to join the two at ring terminals right on the kWeld.
Click to expand...

8AWG Wire used from the XT90 connector leading into the Kweld. Your alternative plan sounds interesting. Really appreciate the comments and ideas.
 
XT90 is already too weak to carry the current from a single battery (~1400A). In fact, given the ratings of these batteries, both AWG10 and XT90 are not appropriate. Running up to 2000 amps through them will probably cause them to melt -keep in mind: 41% more current results in doubled conduction loss. If you don't want to swap the battery connectors, the best compromise is to run separate cables with ring lugs to the welder.

The 1.0m limitation is important because otherwise you'd need to dial in the extra length into the welder, and it would then reduce its current limit - not good if you intend to run it close to 2000A. All pieces of wire count, although my initial calculations never included the length of the battery cables themselves, nor the bus bars, nor the electrode system. There is some safety margin in the design, and I have made destructive tests at 2800A (this again corresponds to doubled stress). Still, the less stress the better, so every piece of wire that you can shorten is worth it.
 
"If you don't want to swap the battery connectors, the best compromise is to run separate cables with ring lugs to the welder"

Now why did I not think of that? Excellent. Thanks Sir. I really appreciate your great product and customer service!
 
I use the 120 Amp rated Anderson connectors and feed both Negs and Pos into one connector for each pair. Matching connector is on KWeld.

 
You must log in or register to reply here.

Similar threads

H
spot welder recommendation: Glitter vs Sunkko for .3mm pure
Replies
0
Views
492
Hempstaed
H
P
BifRC PCB spot welder - gate driver modification help & advice needed.
Replies
8
Views
536
rafalg
R
Lucifer
Want a battery spot welder that works for £20?
Replies
23
Views
2,279
Lucifer
Lucifer
F
Sunkko S709a battery spot welder and DIY battery parts
Replies
1
Views
554
jvm37
J
Z
20S DIY triangle battery configuration
Replies
4
Views
776
zorbas
Z
Back
Top