Battery Spot Welder (CD Welder) || DIY or Buy - YouTube

Channel: unknown

[1]
In a previous project video I showed you how to create a battery pack out of
[6]
18650 lithium ion batteries.
[9]
Back, then I utilized a rather riskier and not recommended methods of connecting the battery cells together.
[16]
By soldering the nickel strips directly to them.
[19]
This can be dangerous because you heat up the battery quite a bit which according to its data sheets can lead to problems.
[27]
So even back then I tried an immediately failed at doing some DIY spot welding.
[33]
With the two salvaged microwave transformers which I had lying around from another failed wedding video.
[39]
But since I want to create a bigger battery pack in the near future
[43]
I'm in dire need of a battery spot welder and truth be told I almost bought a cheap one for around 200 euros.
[51]
Which seems to be capable of doing the job without a problem.
[56]
But before buying I wanted to try the DIY route and see whether I could create my own battery spot welder.
[63]
Which is the subject of this DIY or by episode.
[67]
And spoilers I will fail terribly.
[70]
Let's get started!
[72]
[music]
[83]
When it comes to spot welders
[85]
Then we can either use a big transformer in combination with some power electronics
[90]
To create small energy pulses with the length of a couple of milliseconds
[94]
Or we can create a CD welder aka a capacitive discharge welder
[100]
This method still uses power electronics to create millisecond pulses of energy
[106]
But the main power source are this time capacitors.
[110]
As you might know if capacitors get connected to a voltage source then they get charged up
[116]
and can also hold this charge for a while
[118]
and if you short the capacitor leads with a nickel strip
[122]
then we get an energy burst which in this case is not enough for welding though
[127]
Since the 220 microfarad capacitor only held an energy of zero point zero two four seven five joules.
[135]
But if we instead utilized super capacitors with a capacity of whopping 100 farads
[141]
connect six of them in series
[144]
Utilize Kapton tape to create a poor man's capacitor bank and charge the pack up to 15 volts
[151]
Then we successfully stored around
[153]
1875 joules of energy
[156]
Which should be plenty for spot-welding.
[159]
That means it was time to create the control circuitry around the capacitor bank
[164]
As electrical switches I went with IRL 2505 N channel MOSFETs,
[170]
which can handle up to 104 amps
[173]
Thus I used one for the charging process of the capacitors and three in parallel to discharge the capacitors during welding.
[182]
Additionally I utilized four TCO four 4420 MOSFET driver ICs
[188]
Which already ends the power electronics part for this project.
[192]
For he control electronics I went with an Arduino Nano as the brains of the organization,
[198]
a 128 by 64 OLED display so that we can see what pulse time is currently used and whether the capacitors get charged up
[208]
a rotary encoder to adjust the port's time and initialize the charging of the capacitors
[214]
and finally a pedal for sewing, to which I soldered one channel and the ground cable of an audio cable
[220]
So that I can connect it to the circuits through an audio jack in order to activate the discharge pulse.
[227]
To those main components I then added some complimentary ones
[232]
and started turning this theoretical composition of components into a proper schematic
[237]
through the free on-line circuit design software called Easy EDA
[241]
Not only did I make sure to connect all the components to the correct Arduino pin and counterpart
[247]
But also make sure to utilize the correct package with included PCB footprint
[252]
Because once the schematic was complete. I did not feel like connecting this big number of components to one another on a piece of perfboard
[261]
Instead I click the convert to PCB button and started arranging all the PCB footprints of the components
[268]
Once I was happy with the layouts I then started tracking the copper traces on the top and bottom layer
[276]
at the end I placed thicker copper traces for the power electronics and
[281]
created a solid copper region for the ground potential
[285]
and with that being done, I exported the Gerber files and ordered five of the PCBs through JLC PCB.
[293]
Which was not only affordable, but also only took one week for them to arrive.
[298]
Now even though the PCBs look like they would work without a problem, there still was a problem.
[305]
Which was that I did not use to correct footprint for the DC jack.
[309]
So I had to enlarge the holes, which did work out eventually.
[314]
That means it was time to first [solder] all the SMD capacitors to the PCB
[319]
and afterwards all the through-hole components which all and all took roughly around 30 minutes.
[326]
And if you're wondering for what the big 10 ohm power resistors is used
[330]
Then let me tell you that it is used to slowly charge the capacitors while providing feedback voltage to the Arduino.
[337]
So that the microcontroller knows when to stop charging.
[341]
Only problem was that the feedback voltage can easily exceed 5 volts,
[346]
which the 5 volt Arduino certainly does not like.
[350]
So as an afterthought I interrupted the feedback copper trace
[355]
and added a 2 kilo ohm/1 kilo ohm voltage divider to lower the voltage.
[361]
And with that being done the hardware was complete, and it was time to program the microcontroller
[368]
Now I will not go into too much detail on how I created the [code].
[372]
Since I already talked about interfacing each one of the utilized components in previous videos.
[378]
Let me just say that I used an external interrupt for the rotary encoder,
[383]
pin change interrupts for the push button and the paddle button and a super simple delay function to create the welding pulse.
[391]
So after uploading the code, connecting the capacitor bank and a DC voltage of 15 volts to the circuits
[398]
It was time for testing and as you can see the rotary encoder control works.
[404]
Charging the capacitors is possible
[406]
and even the pedal switch created a precise pulse according to what we have set with the rotary encoder
[413]
Beautiful, right?
[415]
Well here begins the depressing part, since the capacitor bank was always connected to the welding terminals
[422]
no matter whether there was an impulse or not
[426]
The problem was, which I found out after too many hours of troubleshooting
[430]
the missing reference potential from the discharge MOSFETs
[434]
Which I could not add as an afterthought because that would have shortened out the main power supply.
[440]
So all in all this project was a big bummer for me.
[444]
But it shows beautifully that buying a product can often saves you a lot of nerves.
[450]
Now for me going back to the schematic and fixing this problem is not worth it.
[455]
Since welding projects seemed to be my kryptonite.
[459]
But if you want my project information to fix it or see a functioning CD welder
[464]
then simply have a look in the video description.
[467]
As always, thanks for watching.
[469]
If you enjoyed watching my failed attempt, then don't forget to like, share and subscribe
[475]
Stay creative, and I will see you next time! With a functioning project...hopefully.