kWeld - "Next level" DIY battery spot welder

[Upnorth]

Any more information on using the Ksupply as a bench top power supply?
The manual I found online has a section for details on this but it is left blank.

I would like to hook up a couple pots and an lcd screen would be great. Some sort of indicator to show if it is in CC or CV mode....

 

[Upnorth]

I'm staving off buyer remorse....
Back me up here.

I just put an order for a Kweld, KCap, Ksupply and a HSTNS-PR01 - HP 850 to 1000 -Watt supply.

Its a BIG chunk of change for a guy who has NEVER spot welded a battery.....

"Buy once cry once" right?

In the past I have spent far too much time trying to reinvent a cheaper solution to the problem. This thread has convinced me that there is simply no better way to build a pack. my reasons are:
-supply and caps will get rid of delays
-no high amperage Lipos, I hate them because I wont use them for anything else and would likely neglect the care they need, and they would become a fire hazard
-probably the biggest reason is that this is "straight from the source", I want the designer to make a bit of coin, selfishly because if he does then there may be more good products from him in the future (Tantus if you are listening a nice LCD and controls on the supply.........)
-NOT supporting knock off companies that try to copy and pass off goods they don't truly understand

This set up will set me back about $1000 cdn all in. A huge sum of cash for me.
I have a LUNA Wolf battery that died after 185 cycles, it set me back $800 cdn, I actually have two, so I am in for $1600. My intent is to commute and that $/km ratio sucks, $0.15/km for the Wolf that got me 4500km in two years.

Soooo. I ordered the cells from Battery Hookup, 120 cells for $300cdn, add in the $1000, and another $100 for a bms, and I can do two batteries for $1400cdn....... $1500 including incidentals. For a couple batteries that are rated for 2000cycles.

I hope its a wise investment......

If I can use the powersupply and the Ksupply module to anodize an aluminum case I'll really be able to wring the value out of this.

 

[john61ct]

Battery packs based on the 3.6-3.7V chemistries are always going to be a consumable

for regular commuting use likely 3 years is optimistic for even the best build quality based on brand new cells.

Whereas your infrastructure gear should last much longer, delivering that quality and value over and over again.

Might even be a good investment for those looking for side-gig income if you can avoid the liability downsides.

 

[spinningmagnets]

The few times I have seen a used kWeld for sale, it had no problems finding a buyer, at 60%-80% of the new price. The sale typically was not due to unhappiness with performance, but once a pack or two was built...a few months later the owner felt that they were unlikely to build another pack and wished to sell while they are in demand.

 

[goatman]

I'm staving off buyer remorse....
Back me up here.

I just put an order for a Kweld, KCap, Ksupply and a HSTNS-PR01 - HP 850 to 1000 -Watt supply.

Its a BIG chunk of change for a guy who has NEVER spot welded a battery.....

"Buy once cry once" right?

In the past I have spent far too much time trying to reinvent a cheaper solution to the problem. This thread has convinced me that there is simply no better way to build a pack. my reasons are:
-supply and caps will get rid of delays
-no high amperage Lipos, I hate them because I wont use them for anything else and would likely neglect the care they need, and they would become a fire hazard
-probably the biggest reason is that this is "straight from the source", I want the designer to make a bit of coin, selfishly because if he does then there may be more good products from him in the future (Tantus if you are listening a nice LCD and controls on the supply.........)
-NOT supporting knock off companies that try to copy and pass off goods they don't truly understand

This set up will set me back about $1000 cdn all in. A huge sum of cash for me.
I have a LUNA Wolf battery that died after 185 cycles, it set me back $800 cdn, I actually have two, so I am in for $1600. My intent is to commute and that $/km ratio sucks, $0.15/km for the Wolf that got me 4500km in two years.

Soooo. I ordered the cells from Battery Hookup, 120 cells for $300cdn, add in the $1000, and another $100 for a bms, and I can do two batteries for $1400cdn....... $1500 including incidentals. For a couple batteries that are rated for 2000cycles.

I hope its a wise investment......

If I can use the powersupply and the Ksupply module to anodize an aluminum case I'll really be able to wring the value out of this.

you want to sell that dead luna wolf?

 

[tatus1969]

Any more information on using the Ksupply as a bench top power supply?
The manual I found online has a section for details on this but it is left blank.

I would like to hook up a couple pots and an lcd screen would be great. Some sort of indicator to show if it is in CC or CV mode....

Just seeing this now, somehow the ES email notifications don't arrive anymore since a while...

Bench supply conversion requires removal of the two trimmers and soldering two 10k linear pots. Current limiting is a bit voltage dependent because the circuit uses PWM peak current limiting, and adding a second control loop around that (and the extra cost and size) wasn't the main focus. I also recommend adding a bunch of low ESR capacitors close to input and output, a) to maintain stability and b) to reduce output ripple (if desired).

CC/CV steering is not done in the classic way and does not allow direction insertion of LEDs. Dual diode D1 pulls down the control signal when current or thermal limiting takes over, but it needs to be low voltage drop Schottky. R10 works against them, controlling voltage. The opamp outputs should be able to drive an LED however: IC2 pin 7 goes low in thermal limit, IC2 pin 1 goes low in current limit. The system is in voltage control when none of these are low.

The more fancy method would be to use kSupply as a slave module only, and use a discrete of microcontroller based controller with separate voltage and current measurements, LCD, ... In this case, after removing R10 the control signal (representing current) could be fed in directly.

 

[tatus1969]

Tantus if you are listening a nice LCD and controls on the supply.........)

I am Meanwhile I have graphical LCD code in my toolbox, made for another project (using lvgl). The feature list for a kWeld2 is long, all is missing is time. At the moment I am fully loaded with customer projects till beyond end 2021...

 

[Rafalkut]

Hi can some help.me to indentyficate thia element on Kweld it pop out last time , and i cant find in web parameters of this is next to cappactors

 

[tatus1969]

Hi can some help.me to indentyficate thia element on Kweld it pop out last time , and i cant find in web parameters of this is next to cappactors

That is a Littelfuse SMDJ28A.

 

[Upnorth]

I feel like a total noob…..
I just got my power supply and hooked up my KWeld.

I am using an HP power supply power, KSupply and Kcap. I have no switch on the power supply, so I just pull the plug on it.

The component are all hardwired (bolted terminals).

How do I turn this thing off?

I welded on scrap with the supply un plugged until the “low current” error came on, but it’s been over an hour and the power light on the cap is still on and the KWeld is still on, finally now showing “low volts 4.0”

 

[Merlin]

it takes some to bleed down the energy.
That "endless" Squeeking" is also annoying for me.

But you can discharge it with resistors(big ones lol) or simply with a light bulb.

 

[Upnorth]

Where should I connect the load?

 

[Merlin]

...output of your kcap

 

[Silvaticus]

tatus1969, could you tell us how the calculation done in kWeld of the energy applied to the place of welding occurs. At least in general terms.
If you refuse to answer this question, then perhaps other forum members will help figure it out.

Here's what my hobby colleague wrote:

«Resistance of an open welding circuit (from electrode to electrode) with open mosfets (Ri + Rc1 + Rc2 +Rs +Rc3 + Rc4) is 1.35mΩ.
Resistance of the circuit section (which we use as a current shunt) 0.32mΩ
The duration of the welding pulses is 6.1ms.
Voltage drop at the power source at the time of welding from 5.6 to 5.4 volts.
The voltage drop in the measured section of the circuit (shunt) 0.8 volts.
Know the current is 2500А stable, about 6000А at the beginning of the pulse.
I didn't know what this splash was. And I can't find myself the reason why it appeared on the oscillogram.
If we take the current equal to 2500А, then the resistance of the closed circuit (Ri + Rc1 + Rc2 +Rs +Rc3 + Rc4 + Rw) at the time of welding is 2.16mΩ
We subtract the welding resistance, we get (Rw) 0.81mΩ. This is the resistance of the weld spot.»

Is it also correct to assume that approximately 37.5% of all energy expended was spent on creating welded joints and the remaining 62.5% became losses turning into heat?

 

[tatus1969]

tatus1969, could you tell us how the calculation done in kWeld of the energy applied to the place of welding occurs. At least in general terms.

In general terms, the unit measures the power that is put into the weld spot. In detail, that is what differentiates kWeld from other solutions

 

[Silvaticus]

In general terms, the unit measures the power that is put into the weld spot. In detail, that is what differentiates kWeld from other solutions

I beg your pardon, but I sincerely believe that it does not matter whether the total energy is displayed on the screen or only spent on work to create welded joints. So understanding the principles and using the values from the screen allows you to achieve guaranteed results in both cases.

But the differences between kWeld and all the others are best to be discussed on photographic evidence of welding results.

https://ibb.co/ZfFshFf
https://ibb.co/LRyXqGv
https://ibb.co/MpHx2hv
https://ibb.co/w0G9jHf

 

[tatus1969]

it does not matter what and how accurately you measure the energy applied to the weld

I fully disagree - doing that is exactly what creates the repeatability that kWeld is known for.

 

[Silvaticus]

I fully disagree - doing that is exactly what creates the repeatability that kWeld is known for.

“Fides sine operibus mortua est” is therefore best discussed using examples of welding results. Could you please show the welding result of pure nickel or Hilumin 200 μm. I would like to see this compared to our results.

 

[tatus1969]

The calculation algorithm and its relative accuracy within reasonable limits do not affect the result in any way. We will not take into account Chinese garage crafts. It makes no difference Sunstone Welders display directly on the screen the energy consumed by the entire system or that which went directly into the creation of the welded joints.

What is your professional background to make that statement?

 

[Silvaticus]

Unfortunately, I have not yet received an answer to my original question.

Here's what my hobby colleague wrote:

«Resistance of an open welding circuit (from electrode to electrode) with open mosfets (Ri + Rc1 + Rc2 +Rs +Rc3 + Rc4) is 1.35mΩ.
Resistance of the circuit section (which we use as a current shunt) 0.32mΩ
The duration of the welding pulses is 6.1ms.
Voltage drop at the power source at the time of welding from 5.6 to 5.4 volts.
The voltage drop in the measured section of the circuit (shunt) 0.8 volts.
Know the current is 2500А stable, about 6000А at the beginning of the pulse.
I didn't know what this splash was. And I can't find myself the reason why it appeared on the oscillogram.
If we take the current equal to 2500А, then the resistance of the closed circuit (Ri + Rc1 + Rc2 +Rs +Rc3 + Rc4 + Rw) at the time of welding is 2.16mΩ
We subtract the welding resistance, we get (Rw) 0.81mΩ. This is the resistance of the weld spot.»

It would be great if someone who is competent in this matter could confirm or deny the correctness of the assumption.

 

[john61ct]

What does this have to do with the fact that you could not or did not want to answer a question not related to kWeld? Just for fear of revealing the secret of the unique kWeld algorithm?

...

P.S. I fully realized the complete futility of trying to get answers in this thread. I will try to find a topic where I can ask again and maybe even get an answer.

Familiar writing style here.

Why do you feel so entitled?

 

[Silvaticus]

Familiar writing style here.
Why do you feel so entitled?

Why do you feel so entitled? Why did you decide that? It's a mystery to me why such a question caused so much resistance. We use a different type of spot welding and would like to better understand exactly how much energy is spent on welding. The results have been achieved in one way or another and they are stable, despite certain difficulties caused by the design. The purpose of the question was simple, there was a desire to be able to better understand how much energy was spent on welding.

Possibly you can / would like to confirm or deny my colleague's calculations? I believed that this could be useful not only for me and him. Otherwise, I would not write in this topic, but went to visit a familiar electronics engineer. I just really didn't want to distract him.
I would greatly appreciate it.

 

[john61ct]

I do not have answers, was just pointing out one reason you may not be getting what you want.

http://www.catb.org/~esr/faqs/smart-questions.html

 

[Silvaticus]

Yes, you're right, my answers sound a little arrogant. My apologies. In fact, I just wanted to figure out if we think correctly.

 

[john61ct]

Overly dramatic and counter productive.

Someone may well be able to answer your questions, which would be A Good Thing for the community.

Just ask a bit more nicely that's all, recognizing that the experts here are volunteering / donating very valuable time and energy helping us noobs