kWeld - "Next level" DIY battery spot welder

[Fraser]

Frank,

I will always be using my kWeld on 14V or less so is there any reason why I should not fit a lower voltage TVS better matched to my maximum intended supply ? I was thinking of one that engaged at say 20V and fully clamped at say around 22V. It's working voltage of 18V would prevent any unwelcome issues with a fully charged car battery p.d.

Best Wishes

Fraser

[tatus1969]

Frank,

I will always be using my kWeld on 14V or less so is there any reason why I should not fit a lower voltage TVS better matched to my maximum intended supply ? I was thinking of one that engaged at say 16V and fully clamped at say around 20V.

Best Wishes

Fraser

Yes, that would leave much more safety margin. And it would also allow you using a PTC / resettable fuse of lets say 0.5~2 ohm instead of a resistor. The 10 ohm resistor adds ~0.5V of voltage drop from the battery (operating current is 40-50mA, with pulses of ~300mA during holdup capacitor recharge). I'll revise the entire supply circuit in the next revision, and tend to use a linear regulator with enough input voltage rating instead. Power dissipation will be 1W at 30V, but who cares at welding power in the kW range...

[Fraser]

Many thanks Frank. That is what I will do on my kWeld them.

Fraser

[anishkgt]

silly question here. Forgive me for that.

what does the ohm mean here ? resistance of the nickel between the two points at which the elctrodes touch ? or the wire during weld ?

[tatus1969]

silly question here. Forgive me for that.

what does the ohm mean here ? resistance of the nickel between the two points at which the elctrodes touch ? or the wire during weld ?

It displays the measured resistance between the electrode tips. EDIT: this is true when welding. During calibration, it displays the combined wiring/electrode holder resistance at the output. So somehow both of your answers are correct. And it shows that your question was not silly 

[tatus1969]

Many thanks Frank. That is what I will do on my kWeld them.

One more thought: a 0.5R polyfuse is maybe too little resistance. You have to assume that the inductively driven voltage at the input is ~50V. When your TVS clamps at 20V, then the current would be (50-20)/0.5=60A. If the TVS can sufficiently clamp this current, then you're fine. Otherwise it may be wise to increase the resistance. At 2 Ohms, the current is reduced to 15A which I like much more. Power dissipation is not an issue, because the voltage spikes last less than 20 microseconds.

[Fraser]

Hi Frank,

Thanks for the advice. I will add a little resistance as you suggest.

I have elected to use a P6KE18A TVS rated at 100A so it is quite robust. I already have them in my component stock so may as well use them 🙂

Best Wishes

Fraser

[anishkgt]

Maybe a bit off topic.

Just need to clear my doubts on getting amps and resistance put into the nickel strips by this formula
I = U * diameter [mm2] / (0.0175 * length [m])

with that i get this result
I = 0.29 * 25 / (0.0175 * 0.13)
  = 3,186A for 20ms

How true is this ? i did start a thread on this but am just curious and lost at the same time as to how to calculate it without any additional components or costly gadgets.

[tatus1969]

I = 0.29 * 25 / (0.0175 * 0.13)

Please elaborate more on the numbers, except for the specific resistance of copper (not nickel) I cannot see how you came to them.

[anishkgt]

Oh i am sorry that i missed that out

I = 0.29 * 25 / (0.0175 * 0.13)
V = 0.29 (on a cable with two tapping 13cm apart)
size of cable in mm² = 25
0.0175 is a constant (not sure what it is but i got the formula from here http://avdweb.nl/contact/12-contacts/1-avandalen.html)
distance between two taps on the cable = 13cm (here 0.13m)

[SirJMD]

Just finished testing copper strips and copper wire. No success at all. I think the copper conducts way too good.

[tatus1969]

Just finished testing copper strips and copper wire. No success at all. I think the copper conducts way too good.

I did some quick research on this. To weld copper strips, one supposedly needs 2500A+, tungsten electrodes, an inert gas. There is also an interesting thread here: https://endless-sphere.com/forums/viewtopic.php?t=84680

[tatus1969]

Oh i am sorry that i missed that out

I = 0.29 * 25 / (0.0175 * 0.13)
V = 0.29 (on a cable with two tapping 13cm apart)
size of cable in mm² = 25
0.0175 is a constant (not sure what it is but i got the formula from here http://avdweb.nl/contact/12-contacts/1-avandalen.html)
distance between two taps on the cable = 13cm (here 0.13m)

That is Ohm's law combined with https://en.wikipedia.org/wiki/Electrical_resistivity_and_conductivity

[anishkgt]

Yes but how reliable would it be ?

[tatus1969]

Yes but how reliable would it be ?

What do you mean by that?

[anishkgt]

I mean can we determine the amp draw using this method, that is with that formula ?

[tatus1969]

I mean can we determine the amp draw using this method, that is with that formula ?

Did you take the time to understand the wiki article and relate it to the formula? That should answer your question.

[anishkgt]

well i did read it but with lack of knowledge on them i was lost if this road would be correct.

[tatus1969]

well i did read it but with lack of knowledge on them i was lost if this road would be correct.

It is only Ohms law, R=V/I, and the specific resistance of copper, R= 0.0175 * length / crosssection. 

[tatus1969]

well i did read it but with lack of knowledge on them i was lost if this road would be correct.

It is only Ohms law, R=V/I, and the specific resistance of copper, R= 0.0175 * length / crosssection. 

By the way. It seems that you want to copy my invention of using an energy metering approach into your design. If you do this for yourself, I am fine with that, but if you want to implement and sell it and by this damage my business, I would find this very immoral. I am not a large company that has enough money and doesn't care about this, but a husband and father trying to make a living from this and other projects in a self employed business.

[anishkgt]

Oh Tatus i had begun this project of mine a year back and they are not based on the capacitve discharge welding, rather they work on ac and determining the amp drawn was harder where the ac had to be rectified and then fed into the MCU and again there would be voltage drop down the way.

I brought this question up because it was done here http://avdweb.nl/contact/12-contacts/1-avandalen.html and wish to know how true it was by asking for others opinion. It had nothing to with your project and in fact am not going to add it into my either though the approach to it is totally different. Was only shedding some light to it so maybe yourself or others could make use of it.

Secondly i know how hard to sell something to some pennies out of it. I had sold quite a few of earlier spot welders. I have the second version pcbs with me that i had posted here in the forum https://www.eevblog.com/forum/manufacture/some-china-pcb-manufacturer-recommend/50/.

I too am a dad and a husband but this is not my primary job i work in the IT industry and these are just my hobby which am interested in with that said i do not copy and paste ideas, i try to do it my way with my own ideas. So Tatus need not worry about copy your ideas or designs into mine.


Hope that clears your mind.

[anishkgt]

Impressive project by the way Tatus . Seems one of a kind and i hope this would bring you fortune. As far as i have read, growth and fame has always begun in a closed small room, a grage or in the basement.

[tatus1969]

Impressive project by the way Tatus . Seems one of a kind and i hope this would bring you fortune. As far as i have read, growth and fame has always begun in a closed small room, a grage or in the basement.

Thanks, in this case it would an attic 

[anishkgt]

Impressive project by the way Tatus . Seems one of a kind and i hope this would bring you fortune. As far as i have read, growth and fame has always begun in a closed small room, a grage or in the basement.

Thanks, in this case it would an attic 

ha ha ha  that is a close match to 'closed small room'

[mazzic518]

I definitely want one looking forward to the second batch.