Opening the electrodes will cause a splash usually, it is best to keep pressure on, most welder use either a timer/contactor on the primary or a couple of SCR's back to back.
I would have tended to use multi-strand welding cable for the secondary.
Max.

Yes I leave the workpiece clamped for a bit after the weld. the large copper plates sink the heat out of the electrodes and the weldspot and strengthen the weld. After a few seconds of post-weld clamping, the workpiece is safe to touch.

I would have used stranded welding cable too, if I had any to spare. But this is a scrap project. The idea was to improve on the low-cost spot welders like the one available from harbor freight. The HF spotwelder has 6" tongs and that's it. If you need to weld deeper than 6" you're S.O.L. There are more expensive units by Miller et. al. that come with 6, 8, and 10" tongs, but they are still the "tongs attached to a heavy box" spot welder. I wanted something with flexible leads and long reach that I could manipulate without holding the bulky beast in my hands. I achieved that, but at the cost of a lot of copper, much stiffer-than-desirable cables, and less-than-optimal welding current. Production spot welders put out 20k-50k amps. I think 4.5k amps is reasonable for my machine; that would be the max I could get out of it on a 240V 50A circuit at 90% efficiency.

 

How about you add up all the circular mils of your existing cable, calculate the apparent ohms, pick some number of inches, and place voltage taps that far apart?

But it will change with heat!
Blow a fan on those few inches of cable.

 

1/4 inch copper? You're thinking to small.
I'm thinking, 35 millivolts for the last analog meter movement I worked with. @1000 amps, that's 35 watts lost in the shunt.
1 gauge copper @ 8072 ft/ohm
35 micro-ohms needed
math...math...math
3.39 inches

hmmm.... and I can get 1ga copper on ebay for $6/ft.

 

How about you add up all the circular mils of your existing cable, calculate the apparent ohms, pick some number of inches, and place voltage taps that far apart?

But it will change with heat!
Blow a fan on those few inches of cable.

I got .03486mOhm/ft
@4000A, that's 139mV/ft, or 3" of cable for a 35mV meter.

Yep, I think I know what to do now.
Thanks for the help!

 

FEAST THINE EYES!!!!!
I can honestly say I've never seen anything like this before!!
Are the primaries really in series, or does the same windings go through all the cores?
What size circuit are you powering this from?
What is the OCV on the secondary?

 

FEAST THINE EYES!!!!!
I can honestly say I've never seen anything like this before!!
Are the primaries really in series, or does the same windings go through all the cores?
What size circuit are you powering this from?
What is the OCV on the secondary?

They are 120V primary transformers. I have 3 of them, primaries in series, across a 240V 50A circuit. Each transformers sees 80V (actually, 80V, 81V, and 82V respectively, with 244V actual input). This is beneficial because these cheapass MOTs saturate when applied their "rated" input of 120V. Each of them, when powered from 120v, draws 9.5+A with no load - that's true whether powered with their original HV secondaries, no secondaries, or LV scondaries - AND that's true whether the magnetic shunts are in place or not. With the primaries 3 in series across 240 (each seeing 80V) the buzzbox only draws 0.9A no load.

The secondaries are in a sort-of-I-guess "series" configuration. As you can see in the pics, the shared seconday "winding" (hodgepodge of scrap wire) passes through one side of each transformer, and then comes back and goes through the other side of each transformer. The resultant OCV (2.2V) is exactly the same as if a wire is passed in and out of one transformer and then in and out of the next, and the next, in the traditional series configuration. It's functionally one big transformer, except with the iron split up into chunks for more efficient heat dissipation. Putting the secondaries in this quasi-series configuration allows me to addresse the problem of unequal output voltages from the transformers. Transformer A puts out slightly higher voltage than the other two; when I was experimenting with them in parallel, I was drawing excessive no-load amps by transformer A fighting with B and C. Also this configuration allows me to pass retarded amounts of copper through the transformer windows; even just making one more pass through the window would cut my copper cross-section in half.

At direct short I pull 35A on the primary side. That seems to coincide with max current permissible by the resistance of my leads. They're about 18 ft long end-to-end (from top electrode, through transformers, to bottom electrode); at .00003486 Ohms/ft, that's 627mOhm for the entire length. 2.2V/.000627ohms = 3500A. I originally went with the long leads because I wanted a flexible cable that I could get into tight spaces. What I ended up with is about as flexible as a cane pole, so there's really no point in leaving them long. I'm hoping that by cutting a few feet out, I can boost my output amps a great deal. My math says that at 90% efficiency, if I cut 8ft out of the loop, I can get over 6300A out of it before tripping the 50A service breaker. that will make it a 13KVA machine! Me so horny.

 

For simple spot welding, the diameter of your electrode tips seem a tad large when compared to industrial strength machines.
I have a couple of MIG transformers I am thinking of setting one up as a spot welder.
Max.

 

I'm not really willing to drop any voltage. Voltage dropped across a shunt is voltage not dropped across my workpiece, which is less welding power. As I said above, I'm trying to get even more out of it by shortening the leads.

If you want to measure the current, you're going to have to drop something somewhere. When I was at NIST we had a 4000 A supply that could put out at most about 1V. So our busbars were four ganged 6" x 1/4" OFHC copper bar stock (which got noticeably, but not excessively, warm at 4000A (this was not pulsed, this was DC and often maintained for hours at a time). Our current sensor was a 0.01 mΩ serpentine resistor with forced air cooling that was designed for 50mV output at a full-scale current of 5000 A.

 

For simple spot welding, the diameter of your electrode tips seem a tad large when compared to industrial strength machines.
I have a couple of MIG transformers I am thinking of setting one up as a spot welder.
Max.

Thanks. I can reshape the electrode tip. Currently it is blunt-faced 4AWG solid copper. I am new to resistive spot welding; learning as I go

 

If you want to measure the current, you're going to have to drop something somewhere.

Yes my solution plan as of now is to measure the existing drop of the welding leads - that drop is already there and I can't get around it so it might as well serve as my shunt. With cooling as suggested to keep the resistance more-or-less consistent.

 

You have defeated several gremlins. I thank you for the education.

 

Alright I think I found my panel meter:

@12$ shipped, it's the cheapest thing I could find that measures mV and has a hold feature. It's not a PEAK HOLD, but I know that the peak happens at the instant power is applied, so I can hack the HOLD button and wire it to a relay on the output of my main relay - the time it takes the relay to close should be pretty consistent, so I should get a HOLD value from the same point in time each time. Not ideal, but still workable. Also I could use a multipoint selector to have it monitor other parameters, like internal temp and electrode voltage

 

We had some spot welders where I used to work, I'm no authority , but the on time was critical for a quality weld. Another factor is the squeeze, the spot welders would pinch the metal together (air cylinder) and you could actually see the tips move together while the weld was happening. There may be enough spring in the deep throat vise grips to accommodate that. Your Frankenwelder looks like a fun project, keep us appraised of your progress.

 

To switch the primary the welders had something like this..http://pdf1.alldatasheet.com/datasheet-pdf/view/21245/POWEREX/FT500AH.html
In inverse parallel.
The gate signal had several settings, single pulse, several short pulses, etc.. We used the several short pulse method, all of our metal was galvanized and the short pulses were supposed to burn through the coating before doing the actual weld.

 

We had some spot welders where I used to work, I'm no authority , but the on time was critical for a quality weld. Another factor is the squeeze, the spot welders would pinch the metal together (air cylinder) and you could actually see the tips move together while the weld was happening. There may be enough spring in the deep throat vise grips to accommodate that. Your Frankenwelder looks like a fun project, keep us appraised of your progress.

Yes the squeeze force is definitely a factor. Contrary to what I expected, the less the squeeze force, the better the weld....to a point; too little force and the weld becomes an example of violence. The greater the squeeze force, the more weld time it takes. It seems like something you just gave to get a "feel" for, although I am aware there are tomes of scholarly information regarding all the parameters of spot welding. But what I boiled out of what I have read of that scholarly information is that you want EXTREMELY high current for an extremely brief time (like 50,000A for 2 AC cycles) to get beautiful permanent spot welds. I can't achieve that, but I'm shooting for the highest current and the lowest time that I can achieve.

Here is an example of (low current, high weld time) Vs. (high current, low weld time):




The spot welds in the foreground I did with a single MOT putting out about 480A @2.7V. With that single MOT I had to weld for 7-10 seconds before it would stick, and it would ONLY stick if I applied sanded the surfaces, cleaned them off with brake parts cleaner, and then applied soldering paste flux between the 20ga sheets before welding. The spot weld in the background was performed with the 3-series-transformer configuration at about .5sec weld time (or however fast my foot is). Quite a difference. Also the example in the background was two dirty oily sheets with absolutely no prep and no flux, and the bond is unbreakable - the surrounding metal will bend/break/tear before the weld.

Here's my torture test. I welded two galvanized 1/8" thick washers together (and yes, I breathed deeply while doing it, to cure my zinc deficiency). I welded for 20sec with the 3-series-transformer setup and both washers were glowing. The weld was stronger than I could break by hand, but putting it in the vice and pulling with pliers broke the weld. Maybe better luck if it wasn't galvanized.

 

To switch the primary the welders had something like this..http://pdf1.alldatasheet.com/datasheet-pdf/view/21245/POWEREX/FT500AH.html
In inverse parallel.
The gate signal had several settings, single pulse, several short pulses, etc.. We used the several short pulse method, all of our metal was galvanized and the short pulses were supposed to burn through the coating before doing the actual weld.

That sounds cool. Maybe a good opportunity to throw a uC into the mix. But for now, using what I already have, I was planning on throwing this bad boy onto the primary. Its just an ordinary triac SSR, with a big heat sink on the back and a phase angle controller board on the front. I'm going to swap out the 30A 480V SSR that's in it, for a 90A SSR that I have. Coupled with the timer relay, should give me control over power output and weld time; maybe not down to-the-cycle, but "good enough" I hope.

 

To switch the primary the welders had something like this..http://pdf1.alldatasheet.com/datasheet-pdf/view/21245/POWEREX/FT500AH.html
In inverse parallel.

Those hockey pucks can use up 1.5 volts at 1000 amps.
Probably better to control the primary.

 

If you're interested in what I made with this, check out this thread. It's a baby-sized version of an operator's console that will be used as a desktop simulator/training aid so I can teach people how to run a subsea cutting tool.

 

I have a question. This appears to be an AC spot welder, would that be correct? Using some microwave oven transformers (have to love those things) correct? The weld time duration is said to be a few milliseconds derived using a foot switch. So, what I don't get is how we know where we are on the AC sine wave with respect to time? I mean here in the US a full repetition of an AC 60 Hz sine wave id 16.666 mSec and across the pond where 50 Hz is the go to mains frequency that would be 20 mSec. So if the cycle time of the weld is measured in milliseconds how do we know where we are at on the sine wave?

My experiences years ago with large spot welding machines and magnaflux machines were all with DC type machines. Large banks of capacitors were charged from a large DC power supply and a Triac or SCR about the size of my fist was fired. A sort of 'thud" sound was heard and the parts were forever bonded in a matrimonial type state.

While I can appreciate this is a scrap parts inexpensive project it would be sweet if a shunt could be fabricated or even bought where the peak current could be captured and held on a display.

Anyway, really nice and from the looks of the welds gets the job done quite well, real nice job!

Ron

 

My experiences years ago with large spot welding machines and magnaflux machines were all with DC type machines. Large banks of capacitors were charged from a large DC power supply and a Triac or SCR about the size of my fist was fired. A sort of 'thud" sound was heard and the parts were forever bonded in a matrimonial type state.
Ron

The DC capacitor bank was a stud welder and the caps are charged up to a predetermined level and the whole charge released in one shot, this is a high DC voltage compared to spot/projection/seam welding which is very low voltage high current AC, Usually one turn of a secondary.
The main reason for AC is the energy and time can be accurately controlled, this is either done with Ignitron tubes for very large currents these are SCR equivalents, the amount of current can either be for so many cycles, or a percentage of a cycle depending on the sophistication of the controller, in some simple types, it is done with a contactor and a timer.
The electrode pressure is usually very high via a pneumatic cylinder.
The lower the current capacity of the machine, the smaller the electrode tip, otherwise overheating and deformation of the surrounding area occurs, due to excess time required.
Max.