# Some question about spot welder&welding process

#### HotFurnace

Joined Mar 31, 2018
29
Hi everybody,
- I seen these spot welder are quite common on the internet and market:

And a DIY one:

These spot welders main part is a high current transformer which will transform main voltage to lower AC voltage, then the output is fed directly to the weld parts. Because of leakage inductance, these cannot deliver very high peak current, compare to capacitors.
But in the industry these are rarely used, what are used are spot welder based on capacitors. A capacitor (or more common a bank of capacitors) is charged to high voltage then dumped in very short time to a transformer, which step down the voltage. These capacitors are pulsed rated and can discharge its energy in very short time, resulting in extreme current and fast fusion of welding material.
How much superior is capacitors based spot welders are compare to main transformer based spot welders?? I can see that the "DIY" spot welder make welds look so awful, the weld material had its mechanics properties degraded.
- Why do DIY community& market built such a low quality devices? Capacitors banks aren't very expansive?? Why can't we spent a little more buck to improve the quality of the weld? Can I use such DIY devices to weld parts exposed to high tension?

#### Tonyr1084

Joined Sep 24, 2015
6,611
Not sure about spot welders, but as I understand it - you pass a high current through the parts being pinched together by the electrical probes. The resistance of the metal being welded causes sufficient heat to melt and bond the two points together. Timing is important. Not enough dwell time and the joint is weak. Too long and you can overheat the part and warp it. But that's all I know on the subject.

As for capacitors driving a transformer? ? ? I'm not certain you can. Perhaps you get a high energy pulse from the discharging capacitors which results in a high magnetic field in the transformer. As the caps drain and the voltage drops you MIGHT get a high spike of energy but I'm not convinced of this. Maybe what you're referring to is capacitors that store up DC voltage which has been rectified from a transformer. The capacitor can then deliver a high current in a very short time. Depending on the rating of the capacitor and its charge you can control the welding process. But I guess at this stuff with only a limited bit of experience.

#### HotFurnace

Joined Mar 31, 2018
29
Thanks for the input, from what I seen DIY spot welders takes too long time to melt the metal hence overheat the parts, the heats spreads far away from the joint, causing more troubles. I didn't know that, thanks again!
A capacitor in theory cannot drive a transformer as the current it produce is DC. But note that the transformer works based on Faraday laws stated that V=Ldi/dt you just need to make changes in current, for example rapidly increasing the current by discharge the capacitor through the transformer primary, then at the secondary it will induce a voltage. If the primary is not current limited (for example by spray inductance) then the secondary current can also be very high. Discharge capacitor without transformer is okay but I think it is only valid for low voltage/high capacitance capacitors, not for high voltage pulse capacitors.
I think to get correct timing we must:
- Use low inductance and low equivalent resistance capacitors.
- Adjust capacitor voltage based on energy needed.
Adjust capacitor voltage for correct timing is impossible because our capacitor can only store very specific amount of energy, and voltage affect energy storage in laws of square. If we adjust the voltage then we may run into situation where a high voltage is need for correct energy delivery and part fusion, while the timing suggest a lower voltage. In such situation adjust the inductance is the only way to go.

#### jpanhalt

Joined Jan 18, 2008
11,088
1) Are you trying to make a spot welder for sheet metal or for battery tabs? The capacitor welders I have seen are mostly for battery tabs, not for thicker sheet metal.
2) DIY welders often use just a single pulse. That includes both battery tab welders and sheet metal welders. More advanced welders often include a "cleaning pulse" followed by a welding pulse. A welder with a cleaning pulse needs a way to limit that pulse's duration. Those that are capacitor based and without a cleaning pulse will often just switch an SCR on and duration is controlled by the capacitance and current (i.e., current continues until the capacitors are discharged). Transformer-based welders (DIY) that use a modified microwave oven transformer ("MOT") for low voltage will often time the primary current as that is technically easier to do.

1) If you are making a welder for sheet metal, I suggest checking into the many DIY designs based on MOT's.
2) If you are making a battery tab welder, do you want to use a cleaning pulse?

Joined Jul 18, 2013
24,218
But in the industry these are rarely used, what are used are spot welder based on capacitors.
Not from my experience, I have maintained resistance welders for many decades, including Seam Welders where the electrodes are copper wheels, that weld a seam around items such as automotive gas tanks etc.
The control method for this is a rapid cycle on/off of the welding pulse as the wheels traverse the weld.
You would have difficulty using Capacitor discharge method for this.
Also the cap/discharge method is often used more for stud welding of bolts to sheet metal etc.
Max.

#### shortbus

Joined Sep 30, 2009
9,059
Not real clear what the TS is wanting. He shows spot welders to do two pieces of sheet metal then starts talking like he wants to weld battery tabs. The two are totally different types of welders.

#### strantor

Joined Oct 3, 2010
5,752
Not real clear what the TS is wanting. He shows spot welders to do two pieces of sheet metal then starts talking like he wants to weld battery tabs. The two are totally different types of welders.
I think he just wanted to take cheap shots at DIY spot welders. He hasn't seen mine or he would be singing a different tune. It does not suffer from any "leakage inductance" issues that I am aware of. And if it did, I'm sure the inductance leaked, is less of a problem than the inductance issues introduced by placing a transformer into a DC capacitive discharge circuit.

#### HotFurnace

Joined Mar 31, 2018
29
MOT-----> Microwave oven transformers
1) Are you trying to make a spot welder for sheet metal or for battery tabs? The capacitor welders I have seen are mostly for battery tabs, not for thicker sheet metal.
2) DIY welders often use just a single pulse. That includes both battery tab welders and sheet metal welders. More advanced welders often include a "cleaning pulse" followed by a welding pulse. A welder with a cleaning pulse needs a way to limit that pulse's duration. Those that are capacitor based and without a cleaning pulse will often just switch an SCR on and duration is controlled by the capacitance and current (i.e., current continues until the capacitors are discharged). Transformer-based welders (DIY) that use a modified microwave oven transformer ("MOT") for low voltage will often time the primary current as that is technically easier to do.

1) If you are making a welder for sheet metal, I suggest checking into the many DIY designs based on MOT's.
2) If you are making a battery tab welder, do you want to use a cleaning pulse?
I'm trying to make all-purpose welder, but the first thing I want to weld is 20 18650 battery tabs. More about that later (read below).
Not from my experience, I have maintained resistance welders for many decades, including Seam Welders where the electrodes are copper wheels, that weld a seam around items such as automotive gas tanks etc.
The control method for this is a rapid cycle on/off of the welding pulse as the wheels traverse the weld.
You would have difficulty using Capacitor discharge method for this.
Also the cap/discharge method is often used more for stud welding of bolts to sheet metal etc.
Max.
Yes I agree, I didn't want to say that transformer based is bad, but I tried 1 MOT welder and then got unsatisfactory result. I think they would need more welding current and less welding time. I don't know much (just get into) about spot welding so any help would be highly appreciated.
Not real clear what the TS is wanting. He shows spot welders to do two pieces of sheet metal then starts talking like he wants to weld battery tabs. The two are totally different types of welders.
I stated it above, I want all-purpose welder.
I think he just wanted to take cheap shots at DIY spot welders. He hasn't seen mine or he would be singing a different tune. It does not suffer from any "leakage inductance" issues that I am aware of. And if it did, I'm sure the inductance leaked, is less of a problem than the inductance issues introduced by placing a transformer into a DC capacitive discharge circuit.
Yes I haven't seen your welder, but I'm still not fully convicted about how well it works. Did you remove the magnetic shunt? Could you upload a picture of it welding metal sheet?
Anyway, have anyone really think how bad those transformer are? The manufacturer use their iron and copper very well: The iron cores are very small compare to their rated power. The secondary and primary are small and use as least copper as possible. I don't know what material do North American's and European's MOT windings are made of, but Vietnamese's MOT are usually made of aluminium. I asked the repairmen if he see any copper MOT and he said he seen it only once.
Moreover, do you notice that the manufacturer WELD the core? Normally transformer cores are made of insulated iron sheet to minimize eddy current loss. But in the case of MOT they weld the Es to the Is. One possible reason for this is that if they use regular stacking in normal transformers then the manufacturing time would be much longer than just compress the E and the I then weld them together.
Did you weld the cores again when you put together your MOT? Or you just tape it?
If you use those MOT, you will surely have hard time extracting power from them. Forced air cooling is a must, and if you forget to turn off the power then they are likely to burn.
And if you still question my reasoning, please search google for "microwave oven transformer core loss" or visit this link, this guy analyze the MOT very thoroughly. Other than residing in a Microwave oven, I see no uses for those MOT. Maybe for recycling.
Okay, so I ought to introduce my welder, although not completed.

This welder is capacitor based: 105 capacitors, all metallized polypropylene capacitor with 4.7uF, 630V rating in parallel, resulting in ~500uF and energy storage of 100J. The switch is a KP100 thyristor, the gate triggering circuit ensure highest di/dt possible.
Now if I discharge this into a metal sheet without anything to shorten the pulse, I will surely destroy the thyristor. But making a air core inductor of suitable size would be expansive. I'm thinking about a workaround for this problem.

Last edited:

Joined Jul 18, 2013
24,218
. But in the case of MOT they weld the Es to the Is. One possible reason for this is that if they use regular stacking in normal transformers then the manufacturing time would be much longer than just compress the E and the I then weld them together.
.
This is done in many power transformers of different types, the principle reason is to prevent or lessen 50/60Hz buzzing due to vibration of the plates..
If using a MOT the magnetic shunts should be driven out.
Max.

#### HotFurnace

Joined Mar 31, 2018
29
This is done in many power transformers of different types, the principle reason is to prevent or lessen 50/60Hz buzzing due to vibration of the plates..
If using a MOT the magnetic shunts should be driven out.
Max.
No I think that was not the case, usually transformer core laminations are clamped together using bolts, that should eliminate most of the noise. Why did they choose to hugely increasing loses just to minimize the noise when there's a more power friendly alternative? Bolts are cheap aren't they?
Hello,

Here you can find some PDF's on welding (including spot welding):
https://the-eye.eu/public/Books/cnqzu.com/Anarchy Folder/Workshop/Welding/

Bertus
Thanks for the infomation.

#### shortbus

Joined Sep 30, 2009
9,059
I stated it above, I want all-purpose welder.
And as I stated they are totally different machines. In my 70 years on this Earth, any machine that claims to do more than one thing, doesn't do anything as good as one dedicated to the job.

#### strantor

Joined Oct 3, 2010
5,752
I tried 1 MOT welder and then got unsatisfactory result. I think they would need more welding current and less welding time. I don't know much (just get into) about spot welding so any help would be highly appreciated.

[...]

Yes I haven't seen your welder, but I'm still not fully convicted about how well it works. Did you remove the magnetic shunt? Could you upload a picture of it welding metal sheet?
Anyway, have anyone really think how bad those transformer are? The manufacturer use their iron and copper very well: The iron cores are very small compare to their rated power. The secondary and primary are small and use as least copper as possible. I don't know what material do North American's and European's MOT windings are made of, but Vietnamese's MOT are usually made of aluminium. [...]
Did you weld the cores again when you put together your MOT? Or you just tape it?
If you use those MOT, you will surely have hard time extracting power from them. Forced air cooling is a must, and if you forget to turn off the power then they are likely to burn.
And if you still question my reasoning, please search google for "microwave oven transformer core loss"
I did include a link to my welder that would answer a lot of your questions. Your points on MOTs are all valid. They are cheaply made, and over-rated. If you look at the size of a "1,000VA" MOT, it is a fraction the size of an industrial unit of the same rating. They are horribly inefficient and They saturate well below their rated voltage. They "get away" with this, because microwaves are not continuous duty machines, and nobody is out there measuring temperature rise in their leftover lasagna, comparing to watts consumed, and blowing a horn on the evening news about it. Similar to the "3HP" shop vac or the "5HP" air compressor. Anyone who knows a little about electricity can figure it out. But ya know what? I can buy a "1000W" microwave for less than a 500VA industrial transformer, so MOTs are still my go-to source for transformers for projects.

My spot welder uses 3 MOTs. They are "120V" transformers with the shunts removed, and I found that they start to saturate around 75-90VAC, so I placed the 3 primaries in series across 240V. They are very happy in that configuration.

I did not cut my transformers open. I cut the HV secondaries out, leaving the cores intact.

I suggest to reference the thread I linked for more info. If you want to make a spot welder that actually welds real sheet metal (or thicker, even), a single MOT will not do what you want. There is not enough iron to dissipate the heat, or to magnetically couple enough power to the weld fast enough. I found that out by many experiments. All these YouTube videos of single-MOT spot welders are misleading. People who just want a pat on the back. Anyone serious about spot welding would throw them in the trash. Or re-salvage the MOT and throw a few more MOTs at the problem. Or just go buy a commercial unit.

Last edited:

#### HotFurnace

Joined Mar 31, 2018
29
Sorry for the long delays, I got sick.
Yeah strantor I agree with your points, if you quote how you built it earlier then I will happily agree.
I usually wound the transformer myself, since I had the time to do so. I pick a bigger core and using big aluminium wire for the winding really save up building cost, so it's not expansive to make 2000W transformer. But yes getting that same power from series MOTs would be significant cheaper, I can get MOTs for 1$each. The results are great, can they be improved? The oxidized "aura" surrounding the weld seem big. I will try with 9 MOTs spot welder and see what I could get. #### MaxHeadRoom Joined Jul 18, 2013 24,218 The results are great, can they be improved? The oxidized "aura" surrounding the weld seem big. I will try with 9 MOTs spot welder and see what I could get. Knock the magnetic shunts out. Max. #### strantor Joined Oct 3, 2010 5,752 Sorry for the long delays, I got sick. Yeah strantor I agree with your points, if you quote how you built it earlier then I will happily agree. I usually wound the transformer myself, since I had the time to do so. I pick a bigger core and using big aluminium wire for the winding really save up building cost, so it's not expansive to make 2000W transformer. But yes getting that same power from series MOTs would be significant cheaper, I can get MOTs for 1$ each.
The results are great, can they be improved? The oxidized "aura" surrounding the weld seem big. I will try with 9 MOTs spot welder and see what I could get.
I don't know what pictures you're referring to when you speak of the heat affected zone ("aura") in my spot welds. I posted several pictures along the way as I was designing/modifying/improving the welder. In the beginning the welds were crap. By the time I reached the final configuration they were looking way better. Not perfect, still room for improvement, but very acceptable.

I would advise against using aluminum conductor for the secondary; the resistance is much higher than copper, and reducing voltage drop is the key. Maybe it will work fine if you have 9 MOTs in series, but you have to consider heating in the welding leads and duty cycle. Ex: How many welds will you do in an hour? Enough to make the high resistance aluminum leads scalding hot? Go with copper. Just one or two in a day? Maybe aluminum is fine.

Note that I crammed so much copper through the transformer cores that there is only room for a single pass of the secondary lead through each side of each transformer. So the secondary lead goes through the right side of transformer A, then the right side of transformer B, then C, then comes back through the left side of C, then B, then A. So each transformer has a single "turn" (or half a turn, not sure trchnically) and each transformer adds like 1V or 2V to the output. The output is very low voltage, very high amps. If you intend to do yours the same way, you will need to ensure that all 9 transformers are phased correctly so that the voltage is added to the output, not subtracted.