PAGE3: The insides of an ABI pulse arc welder

These pictures of an ABI welder were sent to me by a good friend, thank you Borowski!!!

This welder looks very primitive to me and basically uses a high voltage low current pulse to start the plasma flame and then a capacitor is discharged a couple on microseconds later to produce a very hot plasma flame. The problem with these types of high voltage starters is that when welding inside a tube or a channel where the sides or a spot on the work piece is closer than the spot that needs to be welded, the high voltage arc will jump to the closest spot and miss the actual point that needed to be welded.

 WELDER PROGRESS 11/17/2012

The start circuit works very good but I'm not happy and are trying to get it to work better.  What i see is that the capacitor discharges to around 13 to 14 volt and then the plasma flame goes out. I read up and found on a couple of sites where they say that the lowest plasma voltage is about 14.6V. So this is normal...please help me out if this is wrong.
After some head scratching i realized what the problem was. The start of the plasma arc is using up too much energy and basically discharges the capacitor too far before a plasma flame gets lit.
I am working on a secondary start circuit and will hopefully test it tomorrow 11/18/2012 and will post the results if everything works out as planned.
After a successful test I will be able to move on to design the rest of this welder.
There are still lots to do, designing the power supplies for this welder.....so far three....I'm not sure if I'm going to make it switching supplies?? and then i have to design an agitating circuit which will help to weld metals that are difficult to weld with frequency adjust and pulse width modulation. I will also have to think how I'm going to control all of this, touch screen? or lots of buttons?
My main objective here is to try and make it as easy as possible to build and keep the cost low.


MORE TO FOLLOW.....

PAGE 2: First weld

I put all the boards and circuits together this weekend and made a couple of test welds. The capacitor I used  is the biggest of 7 capacitors  and the capacitance is 35 % lower than it should be.  I welded a couple of small washers together and there were problems from the start. At first I could not get a plasma flame going and after adjusting the argon gas pressure lower I started seeing results. Another problem was that the capacitor was not discharging completely and it means that the plasma flame was dieing half way through the weld. The total energy roughly worked out was about 50Ws and with only half the capacitor discharging was more in the 25Ws range. This welder is going to be 250Ws and after I get all the little growing pains shorted out, It just might live up to its full potential.
 The following two photos was taken with my microscope of both sides of the washers :

 Three welds on this side

Two welds on this side

It doesn't look like much but it is a start.

SUCCESS

The welder needed to be tweaked and after receiving some parts from Digikey today I managed to adjust the precise pulse timing for this welder.I welded two screws together with the weld energy set to 50Ws. The welds are very shiny, clean and strong.

Two 9mm casings (Brass) welded together

PAGE 1: PULSE ARC WELDER

I designed a battery spot welder HERE and had many requests from people to design a pulse arc welder to weld and repair jewelry. These welders are like small TIG welders that will produce a very hot plasma flame and basically melt all type of metals like gold, silver, aluminum etc together. This welder will be adjustable from 0 to 250WattSecond and will have certain adjustments that will  make it possible to also weld metals that are very hard to weld.

In a previous attempt I tried to use my capacitor discharge welder to achieve the above and believe it or not I did manage to make a couple of successful welds.

 

Two metal washers

Copper wire to a metal washer

I machined a stylus and mounted a solenoid with a collet and a tungsten electrode inside the stylus. The stylus also has ports for argon gas to shield the weld from the air around it producing a very clean and shiny weld.

 

I also designed an electronic board with two Pulse width modulated  relay driver circuits to operate the two solenoids. This board has a microprocessor that will determine when the electrode touches the work piece to energize the argon solenoid and pull the electrode back.

I can hear everyone asking, why a relay driver circuit and why do you need a processor to energize these solenoids? Why can't you just do it with a switch?

The pulse width relay driver circuit is just to limit the current and also save some money on big heat sinks and components getting hot and burning out etc. This circuit works great and will definitely be used in the final design. A better explanation about how the circuit works can be found in my other blog HERE and HERE

A solenoid however is very slow. We are talking "PULSE ARC" and with the pulse I mean, energizing an electromagnet, waiting for the buildup of a magnetic force, retract an electrode away from the work piece,  firing a pulse, creating an arc and getting a plasma flame lit, all within 0.1 thousands of a second(MIN) to 40 thousands of a second(MAX). 
Clearly, a switch is not going to work here, we need a fast processor that will calculate the resistance between the electrode and the work piece as it moves away and then fires the pulse at the right moment or the electrode will get burned to the work piece, contaminating and ruining the weld. 

Here is the prototype microprocessor relay driver board(Schematic above) that I will be using to energize the argon gas and electrode solenoid. The processor will calculate the exact moment to fire the mosfets which will ignite a very controllable plasma flame. I included a flashing led(LED 3)  to indicate if the processor is running for trouble shooting purposes.

I spent the last two weeks building a small mosfet board to fire the biggest of 7 capacitors. I also changed some of the components on the board above for the new voltages this welder will be using and ordered a capacitor from China, which at NO SURPRISE IS 35% smaller than advertised. 

I noticed that one of the readers of this blog already said that he hopes this project does not takes as long as my Capacitor Discharge project. I guess he meant it good and was trying to say that he is really interested in this project and in an hurry to start to use it. 
Designing and building something like this takes a lot of time and money. I am working on a budget and trying to make time for this project between work and Family so please be patient. 

I worked on the software this weekend and

the oscilloscope graph above is from the relay board firing a pulse while the electrode is moving away from the work piece. The red graph(5Volts/ block vertical) is the electrode touching the work piece +- 18 Volt and then moving away, going down to 0 volt In about 25 micro seconds(25 millions of a second.....Measured between the blue vertical lines). 
The processor is doing calculations while the electrode is moving away and fires a very repeatable and precise 5V pulse (Yellow Graph 1Volt /block vertical) at about 14uS before the electrode completely disconnects from the work piece. This pulse will ignite the plasma flame and is software adjustable. 



Any donation will speed up this project and will be very appreciated, thank you.
 

MORE TO FOLLOW.............