DIY Battery Spot Welder from Microwave Transformer Issues

Thread Starter

MadPatch

Joined Apr 3, 2017
15
Hi guys,

I built a battery spot welder out of a microwave transformer by just replacing the secondary coil with 6 AWG wire wrapped 3.5 times. I am measuring 5 volts on my multimeter (not under load, not during a weld). But when I try to weld some thin nickel strip it burns a hole right through and throws red sparks everywhere. I have tried adjusting the number of wraps on the new secondardy coil but still not getting a good weld. Is this issue too much voltage or too much current? And do I need to adjust the gage of the wire?

Extra note. my electrodes are 12awg solid copper, should this be adjusted?

Any insight would be great,
thanks.
 

shortbus

Joined Sep 30, 2009
8,429
Any insight would be great,
May or may not be your problem, but how much pressure are you putting on the strips when you try to weld? Spot welding is a pressure type weld not like a Mig or other type weld where there is a gap between the electrode and work. There needs to be pressure from the electrode on the strips before you turn on the power.
You do have some type of power switch on this don't you? To only power up during the actual weld.

Then there is the strips them selves, they need to be both flat and very clean, to get a good weld.
 

Thread Starter

MadPatch

Joined Apr 3, 2017
15
May or may not be your problem, but how much pressure are you putting on the strips when you try to weld? Spot welding is a pressure type weld not like a Mig or other type weld where there is a gap between the electrode and work. There needs to be pressure from the electrode on the strips before you turn on the power.
You do have some type of power switch on this don't you? To only power up during the actual weld.

Then there is the strips them selves, they need to be both flat and very clean, to get a good weld.
Yeah I programed and arduino to switch a relay when I enter '0' in the serial monitor and have it only power for 30 milliseconds. The strips are nice and flat and there is sufficient pressure. It could be possible the pure copper wire is not a suitable electrode. Has anyone experienced that?
 

kennybobby

Joined Mar 22, 2019
75
30msec is too long, pulse needs to be down around 3-5msec, especially with such thin nickel strips. Copper is the best electrode, high conductivity, high thermal capacity, high melting point.
 

shortbus

Joined Sep 30, 2009
8,429
One other thought, make sure your electrode is flat on the end, not the normal wedge shape from cutting with wire cutters. File or sand it flat.
 

Thread Starter

MadPatch

Joined Apr 3, 2017
15
One other thought, make sure your electrode is flat on the end, not the normal wedge shape from cutting with wire cutters. File or sand it flat.
I tired to make sure it was flat but I think the angle wasn't perfectly flat do it was only catching on an edge
 

MrAl

Joined Jun 17, 2014
8,061
Hello,

I took apart a nice big microwave transformer a while back too taking the secondary off. What a job that was. Didnt get to use it for anything yet i was going to make a very high current 12 to 16v DC power supply out of it.

As far as welding though it sounds like you have too much power going to the joint or not enough surface area, or the time of application of current is too long. The temperature rise has to be just enough to get the joint hot but if it is too hot things will start to vaporize. Something like this happens with soldering too except the joint just oxidizes too much and so it dirties up the metals so the solder does not stick to them anymore. This is especially noticeable when the tip itself is too hot i will not tin properly because the new solder just seems to oxidize too quickly.

But anyway, the temperature rise expression looks something like this:
T=K*(P/S)^a

where K is a constant, P is power , S is surface area, a is an exponent less than 1.
The key point here is that inner part:
(P/S)

which is power over surface area. This means that the more power you apply the higher the temperature rise, and the smaller the surface area the higher the temperature rise.

The required current to melt the metals at a given ambient temperature is proportional to the square of the 'circle' diameter of contact divided by the square root of the time:
I=Kb*d^2/sqrt(t)

and combining the two we end up with:
T=Kc*(D^2*V)/(sqrt(t)*S))^a

So the temperature rise over time t depends on the square of the 'circle' of contact, the voltage drop, and the surface area, and of course the time itself.

What this really means though is very simple. If the metals melt too much then either:
1. reduce the time of current application
2. reduce the current applied
3. increase the area of contact

Either one of those or any combination of those three.
Pretty simple huh? :)

BTW i used the 'circle' of contact because that's the usual shape of at least one of the electrodes but this could vary. The conclusions and thus the remedies will be the same though.
 
Last edited:

drc_567

Joined Dec 29, 2008
1,088
... A possible adjustment to the input transformer power might be accomplished by forming a coil in series with one of the input leads to the transformer primary winding. This coil would constitute an inductive reactance, which would develop a voltage drop according to 2πfL, with a consequent decrease in the transforner primary voltage. The coil inductance would be adjustable by changing the number of coils, the coil diameter, or possibly by just sliding a ferrous bar in and out of the winding assembly. ... Some trial and error may yield a suitable combination for your spot welding project. The principle of this approach is that power in equals power out.
... This concept is not an original idea, just something that I happened to read in a 1950's era book on basic electricity. The idea there was to prevent an electric clothes iron from becoming too hot.
 

Thread Starter

MadPatch

Joined Apr 3, 2017
15
Hello,

I took apart a nice big microwave transformer a while back too taking the secondary off. What a job that was. Didnt get to use it for anything yet i was going to make a very high current 12 to 16v DC power supply out of it.

As far as welding though it sounds like you have too much power going to the joint or not enough surface area, or the time of application of current is too long. The temperature rise has to be just enough to get the joint hot but if it is too hot things will start to vaporize. Something like this happens with soldering too except the joint just oxidizes too much and so it dirties up the metals so the solder does not stick to them anymore. This is especially noticeable when the tip itself is too hot i will not tin properly because the new solder just seems to oxidize too quickly.

But anyway, the temperature rise expression looks something like this:
T=K*(P/S)^a

where K is a constant, P is power , S is surface area, a is an exponent less than 1.
The key point here is that inner part:
(P/S)

which is power over surface area. This means that the more power you apply the higher the temperature rise, and the smaller the surface area the higher the temperature rise.

The required current to melt the metals at a given ambient temperature is proportional to the square of the 'circle' diameter of contact divided by the square root of the time:
I=Kb*d^2/sqrt(t)

and combining the two we end up with:
T=Kc*(D^2*V)/(sqrt(t)*S))^a

So the temperature rise over time t depends on the square of the 'circle' of contact, the voltage drop, and the surface area, and of course the time itself.

What this really means though is very simple. If the metals melt too much then either:
1. reduce the time of current application
2. reduce the current applied
3. increase the area of contact

Either one of those or any combination of those three.
Pretty simple huh? :)

BTW i used the 'circle' of contact because that's the usual shape of at least one of the electrodes but this could vary. The conclusions and thus the remedies will be the same though.
So glad you explained the math too, I love it. Will try playing around and see what I can do
 

shortbus

Joined Sep 30, 2009
8,429
To my knowledge on these battery tab welders the best approach is a capacitance discharge, not a direct from transformer like your using. Doing that lets the size of the caps decide the current applied not a "time" constraint. Now with a regular spot weld it is the time but they are usually welding much thicker parts.

Some good Youtube videos on people doing it - https://www.google.com/search?clien...YAQCgAQGqAQdnd3Mtd2l6yAEIwAEB&sclient=gws-wiz
 

Thread Starter

MadPatch

Joined Apr 3, 2017
15
To my knowledge on these battery tab welders the best approach is a capacitance discharge, not a direct from transformer like your using. Doing that lets the size of the caps decide the current applied not a "time" constraint. Now with a regular spot weld it is the time but they are usually welding much thicker parts.

Some good Youtube videos on people doing it - https://www.google.com/search?client=firefox-b-1-d&sxsrf=ALeKk01vEe6zsVNdFu7VuphapyU3W5LK9A:1614378181294&ei=xXQ5YIW4EZu4tQbOsYyICw&q=diy+capacitive+discharge+spot+welder+youtube&oq=diy+capacitive+discharge+spot+welder+you&gs_lcp=Cgdnd3Mtd2l6EAEYADIICCEQFhAdEB46BwgAEEcQsAM6BggAEBYQHlDg2QJYmOICYK_8AmgBcAJ4AIABtQmIAZkRkgELMC4xLjIuMS43LTGYAQCgAQGqAQdnd3Mtd2l6yAEIwAEB&sclient=gws-wiz
Yeah I am aware this might be a better solution but I had an old microwave and don't have any of those caps. I know this solution will work with some fine tuning.
 
Top