Gregory Hildstrom Projects Publications Resume Links Contact About Google+ Facebook Youtube Donate




Lincoln Electric AC-225 (AC225) AC/DC stick/TIG welder conversion

Introduction
120V Operation
Initial Untangling
Extension Cord
AC/DC-225 Conversion
- Inductor
- DIY Lugs
- Rectifier
- Connectors and Assembly
- DC Results
- Voltages
- Currents
Lincoln Electric Hi-Freq
Foot Pedal
TIG Torch
Overall Schematic
TIG Results

Introduction

I bought this American-made
Lincoln Electric AC-225 (AC225) arc/stick welder from Home Depot in July 2009, here is the manual. It has been produced basically unchanged for many decades, so it has reliability and reputation in its corner. It is also known as a buzzbox or tombstone welder. I decided to get this after an imported inverter welder of mine bit the dust and its warranty replacement went up in smoke. The simple copper windings inside the AC-225 should be extremely reliable for the power output. This page details all of the modifications I made to my AC-225 over a few years. Here is my YouTube playlist about the welder conversion. Here are a couple shots of the finished product:



120V Operation

I discovered that you can actually run the AC-225 on 120V AC power in addition to the normal 240V AC power. The only problem is that the open-circuit voltage is much lower and it is more difficult to maintain an arc, but I did manage to lay down several inches of weld during testing. The bottom line is that it can work at limited welding currents if you are ever in a pinch. The internal cooling fan did not seem to run at 120V and the idle supply current was 1A. The cooling fan ran at 240V and the idle supply current was 4.4A. Here are some power measurements I made while testing 120V operation.

AC-225 Currents With 120V Supply Instead of 240V
Current Setting (A) Measured Welding Current (A) Measured Supply Current (A)
105 60 20
225 120 29


Initial Untangling

The AC-225 is a heavy unwieldy beast. Moving it in stock form is quite a pain. After I installed the simple wheel kit, I cut the front welding leads and rear power cord short. I was tired of trying to manage the cables every time I needed to move the thing. So, I put a NEMA 6-50P connector on the short power cable on the rear of the welder. Then I installed a set of Lenco LDC-50 Dinse-style welding cable quick disconnects rated for 250A on the welding leads. I highly recommend these modifications. Now, nothing gets caught, stepped on, or dragged when I need to roll, lift, or carry the welder. A side benefit is that I can use the same connectors, when I build a rectifier box, to switch between AC and DC operation. It would have been a bit cleaner to get panel-mount connectors, but this was much easier.

Extension Cord

I made a 50' 240V extension cord using three 12 AWG 50' 120V extension cords. I braided them together, cut the connectors off, twisted the conductors together, and attached a NEMA 6-50 connector to each end. Each extension cord functions as one conductor equivalent to 7-8 AWG. There may be a better way to make an extension cord like this, but this one is cost effective and very flexible. I also made a short NEMA 10-30 (dryer) to NEMA 6-50 (welder) adapter using a section of the original power cord.

AC/DC-225 Conversion

Direct current (DC) is necessary for many types of electrodes. DC produces a more stable welding arc, allows greater heat control, and allows me to use my TIG torch with scratch-start. The Lincoln AC/DC 225/125 is interesting, but it is a lot more expensive than my AC-225 and it only produces a maximum of 125A in DC mode. I decided to convert my AC-225 to DC with the goal of supporting all 225A in DC. The following schematic illustrates my plan. It's SVG, so it should zoom and scale in most web browsers.

Inductor

I decided to wind my own arc current smoothing reactor/choke/inductor on a used microwave oven transformer core. I also decided to omit any type of arc voltage smoothing capacitor because I figured the inductor would be sufficient on its own and much cheaper. I wound it with 2 AWG stranded cable. Luckily, the transformer core was about the perfect size to fit in the exhaust channel.

DIY Lugs

I made my own lugs by hammering and drilling copper plumbing couplers. They made very strong connections for about $0.50 each.

Rectifier

I got a great deal on some 1000V 50A bridge rectifiers, so I used 6 of them in parallel. I bent some flat aluminum bar into a bracket and mounted them as close together as possible. Close physical proximity insures close operating temperatures, which helps insure even current distribution. Equal length 12 AWG wiring helps insure some lead-wire resistance and even current distribution. I mounted the bracket where it should get decent cooling air flow and not interfere with transformer cooling.

Connectors and Assembly

I decided to get four Lenco LDPM-50 panel connectors to make the install cleaner, interface with the connectors already on my cables, and make switching between AC and DC very easy. When the cables are connected to the top two panel connectors, output is AC. When the cables are connected to the bottom two panel connectors, output is DC with positive on the right.

DC Results

It works great. The arc is full power, stable, and reasonably smooth. I hooked up my clamp-on current meter to double check the current. I burned rods at 40A, 75A, 120A, and 225A and the DC current was right on the money at each setting. This was quite a bit more work than simply buying a DC welder, but it was interesting and very cheap for something capable of 225A DC. Here is one of my test beads on hot 3/8" plate with 1/8" 6013 @ 90A DCEP.

The DC output was not smooth enough for
stick welding aluminum. The ripple voltage dips too low to maintain an arc with aluminum electrodes. This is a result of too little inductance, but I coudn't fit much more inside the case. Greater inductance could be achieved with more inductors in series, a larger inductor core, or more turns of wire. Another person doing a DC conversion verified that more series inductance, I think he put it in series with the external electrode lead, makes stick welding aluminum possible.

Voltages

I recorded a bunch of voltage data after doing the conversion. I used my multimeter to measure open-circuit AC voltages and, with the help of a capacitor, to measure peak open-circuit DC voltage. I used those measurements to design a voltage divider so I could record data with my Arduino, which requires between 0V and 5V for analog input without clipping. After performing the multimeter measurements and getting the
ArDAQ (Arduino Data Acquisition) sketch working, I recorded the open-circuit voltage as I ran through each of the current settings. Then I welded a couple beads at 120A and 225A, finishing with ridiculously long arcs. As you can see in the welding voltage plots, the DC output is not a smooth line because of the significant ripple from the small inductance. Increased inductance would smooth the output current more if required. Notice the little inductive voltage spikes in the zoomed-in welding voltage plots. During those times, the input current to the transformer crosses through zero, but the inductor prevents the output current from going to zero.
Open-Circuit Voltages
Current Setting (A) Measured VACrms Measured VDCpeak
135 83 118
225 58 82



Welding Voltages
Current Setting (A) Measured VDCavg
120 21
225 27



Long Arc Voltages
Current Setting (A) Measured VDCavg
120 33
225 36



Next, I compiled a spreadsheet of various welding machine output ratings. I included the process-specific voltage and amperage output ratings for stick, TIG, and MIG welding machines. I plotted my two test weld average voltages and two long arc average voltages for comparison.
welding voltage comparison, TIG welding voltage, stick welding voltage, arc welding voltage, MIG welding voltage

Currents



Lincoln Electric Hi-Freq

I learned to do some scratch-start DC TIG welding with a little 120V inverter stick welder. Then I upgraded to a much more powerful 240V inverter machine, which had high-frequency arc start, a foot pedal, and AC/DC operation. It proved less than reliable, but I was able to practice a little TIG on aluminum, steel, and stainless. To keep with the reliable old school theme, I bought a Lincoln High-Freq, which is a "High Frequency Generator For TIG Welding Applications". Here is
the manual. It provides three basic features built into most modern AC/DC TIG welders: gas flow control, high-frequency arc start, and high-frequency arc stabilization. It provides fixed gas pre flow and adjustable post flow. The high-frequency arc start allows you to start TIG welding without scratching the tungsten to the work. The high-frequency arc stabilization (constant on) allows AC TIG welding. Without this, the AC TIG arc would extinguish during the zero voltage portion of the transformer's sinusoidal output. Square-wave AC TIG machines do not have sinusoidal output, so they spend very little time near zero voltage and may not require high-frequency arc stabilization. Just like with my DC conversion, the idea here is to add capability that operates independently of the building blocks beneath it. If this fails, I can still DC stick weld. If that fails I can still AC stick weld. Also, having wiring diagrams of every component in the system should ease maintenance and troubleshooting.



I followed advice from someone who was having Hi-Freq problems and replaced the 7 electrolytic capacitors on the control board with brand new ones. My Hi-Freq was made in 1981, so I figure it couldn't hurt. I chose 105C capacitors, but I chose higher voltage ratings just in case (16->25, 25->35, 150->200). Here is my Digikey part list. All of the original capacitors and values are shown here:


Page 7 of the manual boldly states: "CAUTION: A by-pass capacitor Kit (T-12246) shipped with each Hi-Freq must be installed in the welding power source per instructions included with the Kit. This capacitor protects the welder used with the Hi-Freq from any high frequency voltage feedback through the welding cables. Failure to install this Kit may damage the welder."

Thanks for the warning. After opening up the Hi-Freq and seeing how it works, this makes perfect sense. The unit induces high voltage high frequency AC in the electrode cable. In order to complete the circuit, this HV/HF AC must flow through the electrode cable, through the arc, through the work, through the work cable, and back to the electrode cable through the welder. I'd rather not expose my transformer windings and bridge rectifier diodes to that kind of abuse if I can avoid it. I ordered the T-12246 kit, which consisted of several leads of wire and a dual 0.05uF 1000V capacitor. The capacitor is supposed to be installed just inside the welder in parallel with the welding leads and 'com' grounded to the welder chassis. The only problem is I have two sets of outputs on my welder instead of one selectable set, which means I needed another RF bypass capacitor kit. I managed to find the same Aerovox T11577-12 dual "bathtub" capacitor much cheaper through another source. The capacitor allows the HV/HF AC to bypass the welder's internal circuitry including the transformer secondary windings and my bridge rectifier diodes.



The arc start switch connector says Amphenol 14S-5S. After finding a great PDF on "Amphenol 97 Series Standard Cylindrical Connector" on Digikey, I realize those numbers specify the insert size, number of conductors, and socket. The panel-mount receptacle is an Amphenol MS3102A-14S-5S. So, for the corresponding cable plug, I needed an MS3106A-14S-5P, which I got from Mouser. The remote switch cable leads should be attached to pins D and E on the connector. Pin E is very close to the chassis ground and pin D is +24VDC.

Foot Pedal

Here are some really cool threads with schematics and discussions on adding foot pedal control to transformer-based welders like the AC-225. These threads are definintely inispiring and they all use a similar approach. They basically create a high-current solid-state dimmer circuit in series with one of the 240V input leads to the welder transformer.
Thread: Here is the schematic for SCR control of my arc welder
Thread: SCR Welder amperage controller circuit
Thread: Lincoln tombstone tig
Dan's Workshop - Homebuilt arc welder

First, I decided to try J0K3R-X's dual-SCR module and lamp dimmer approach mentioned in the second thread above, which is very similar to the current control implemented in the first thread and on Dan's welder in the fourth link. Many thanks to the authors for the schematics and discussions. It seemed easier because there were no circuits to build from scratch. I found the same IRKT91 dual Thyristor/SCR module (specifically the IRKT91/12A) on eBay. Many other components like the MCC56 or individual SCRs should also work. For the "lamp dimmer", I thought about buying a Leviton Trimatron 6602-220 600W 220VAC rotary dimmer, but I opted for a 4000W open-frame 220-240VAC dimmer I found on eBay instead. It was cheaper, higher power, and I figured it would be easier to work on if I wanted to separate the potentiometer from it. The TRIAC approach discussed toward the end of the first thread is interesting, but I could not find any TRIACs that handle enough current to support the welder's full power, let alone with much headroom. The original thread and schematic mentioned a 600pF 2kV ceramic capacitor, but I tried 0.5nF and 1nF 500V in my test circuit. The highest voltage I measured at the capacitor was equal to the peak input voltage of 340V, so 500V was more than adequate with my non-inductive test load of two 120V 75W light bulbs in series. I prototyped the IRKT91-based schematic and it chopped up the 240VAC input in exactly the same way as a purpose-built lamp dimmer circuit, which is excellent.


The IRKT91 is by far the biggest semiconductor device I've ever held; it's huge. It shipped from China, but it was made in Italy. Interesting.


My relatively inexpensive battery-powered oscilloscope is great for working with mains/line/power company voltage, but it's still good to be cautious around exposed conductors. The voltage across the non-inductive test load looked great.


The voltage across the actual transformer primary looked quite a bit different. I tested it with an open secondary and a shorted secondary (4 1/8 electrodes welded together). Low-end control was very good, but under load it would latch high once the current reached half of the front panel setting. With the front panel set to 120A, current could be increased smoothly from 0A to 60A, but then it would latch high at 120A and it could not be turned back down. That's unacceptable. I tried a few different simple snubber RC circuits across the SCRs and across the transformer primary, but nothing seemed to help.


Next, I tried Mike W's trigger circuit from the first thread above. I ordered all of the parts from Digikey and it really didn't take long to build. His circuit worked much better for me. It wouldn't go all the way down to zero voltage, but it controlled all 225A smoothly without latching. With the shorted secondary, I had about 3:1 max:min current ratio. For example, with the front panel set to 75A, the minimum output was about 25A and the maximum was the full 75A. I swapped a 470K potentiometer for the 250K potentiometer the schematic originally called for, which allowed zero voltage and much better low-end current control all the way down to 5A.



Here are the parts I ended up using for Mike W's foot pedal control circuit.
QuantityPartDescription
1IRKT91/12Adual SCR/thyristor module
21N4007FSCT-NDdiode 1KV 1A
2478-4045-NDceramic capacitor 0.1uF 1KV
2DB3CT-NDdiac 28-36V 2A
1RV4N254C-NDpotentiometer 250K ohm 2W
1salvagedpotentiometer 470K ohm 2W
2100KZCT-NDresistor 100K ohm 2W
11KZCT-NDresistor 1K ohm 2W
1V2018-NDpc board 0.1"sp 3x3.5"
2486-1258-NDfuse block cartridge 500V 10A
2486-1848-NDfuse 1.6A 125V 5x20 glass

I picked up an S18815 switch, the same as the power switch, to use as a local/remote power control switch. When it is in the local/on position it will short power across the dimmer leads effectifely removing it from the power circuit. In the remote/off position, power will have to flow through the dimmer to get to the transformer. After cramming all of this extra stuff in there, here's what it looks like inside:


I bought an Excel foot pedal control on eBay with the intent of changing the internal potentiometer and the cable connectors. I bought an Amphenol MS3102A-14S-1S 3-conductor panel receptacle socket and a matching MS3106A-14S-1P plug for the foot pedal dimmer/potentiometer connection to the welder. I couldn't go with the 14S-5S/5P combination used on the Hi-Freq because the voltage rating of that shell/insert size is only 200VACrms. The 14S-1S/1P rating is 500VACrms, which is sufficient for the 240VACrms needed for the potentiometer. I fused both hot potentiometer leads inside the welder and grounded the foot pedal chassis to make it reasonably safe. I split the cable near the welder to attach the switch leads to the separate MS3106A-14S-5P connector for the Hi-Freq. I got a couple 97-3057-1007 cable clamps that match the back of the plugs. I also got a couple 9760-14 receptacle caps and 9760-14P plug caps. The Excel foot pedal had a slide style potentiometer with a little bit of backlast in the actuator. Since nobody had the replacement slide potentiometer I wanted in stock, I modified the foot pedal to use my 470K rotary potentiometer. I bent up a simple mounting bracket, created an idler pulley using a skate bearing and two bottle caps, and mounted the rotary potentiometer. I glued a piece of very thin wall 0.8"-1" steel tubing to the knob to create a simple string pulley. It seems to work well so far, but if I have any problems, I'll probably create a nicer string pulley with flanges and a bearing for the unsupported end. As I discuss in the video, the end result is very adjustable.



TIG Torch

I wanted a torch that could handle the full 225A output of the welder, but I didn't want to go the water-cooled route because of the added complexity and expense. I decided on a CK Worldwide CK210 RG air-cooled TIG torch. It's rated for 200A 100% duty cycle on AC or DCSP/DCEN. It should be able to handle 225A, but at a slightly lower duty cycle. That shouldn't be a problem since the AC-225 has only 20% duty cycle at full power. The one I ordered was configured with a 25' two-piece cable with part number
CK-2125NSF RG. I'll probably want a smaller more flexible torch in the future, to compliment this one, but this should do fine for now. The power cable was terminated with a big copper lug that easily attaches to the Hi-Freq. The gas hose was already terminated with a connector that matches the Hi-Freq's gas connections, so easy. I picked up a dual flowmeter with two gas hoses, a stubby gas lens kit, and some other accessories from HTP. I picked up a torch-mounted switch from eBay and a cable sheath from Arc Zone. I used the same connector as the foot pedal for the Hi-Freq connection, so I can choose either the torch mounted switch or the foot pedal switch for arc start.

Overall Schematic

I created
this schematic, and a couple extra parts in this library, using KiCad. Here is the SVG version, which can be zoomed/scaled in most web browsers:


TIG Results

After I got everything hooked up, I started testing. I had to read the Hi-Freq manual a second time because of some strange behavior I saw. It turns out the Hi-Freq was not designed to be used with a foot pedal, so stopping the Hi-Freq on AC is not as simple as completely letting off the foot pedal. However, the great news is that it all works! I was able to run some test puddles and test beads on steel and aluminum. I was able to keep a reasonably sharp point on an 1/8" tungsten at 75A despite the 50/50 AC balance, which surprised me. I expected it to ball up immediately, but I guess that is a pretty big electrode for that current. Following the advice of Jody from Welding Tips and Tricks, I did a bunch of ambidextrous TIG drills on some plate and switched hands every few beads.




Now for the quirks. I can't really use the arc start switch in the foot pedal. The Hi-Freq uses the arc start switch only for starting; it should not be held in continuously and letting go of the switch doesn't extinguish the arc. I've been using the torch-mounted arc start switch and I only use the foot pedal when I want amperage control. The arc must be extinguished by snapping out / pulling away or by letting the welder's output go to zero. However, the Hi-Freq controls certain things like the high frequency arc and the gas solenoid based on sensing the welder's output voltage. So, if you are in AC/continuous Hi-Freq mode and you let the foot pedal bring the welder's output to zero, the arc stabilizing continuous high frequency continues to run and must be manually switched off on the Hi-Freq panel. In either continuous or start only mode, letting the welder's output go to zero may interrupt the Hi-Freq's post flow timer. I'm still experimenting and figuring out how to best use everything together, but it's very capable and I'm pretty happy with it overall. I've been able to use the full DC output range without any issues. On AC TIG with continuous high frequency, I can't weld aluminum over 90A for more than a minute or two without tripping my 50A breaker. I haven't had any problems at 90A and below, but at 105A and above, it trips after a bit of use. I guess that's my biggest dissapointment at the moment; I can't use the full output range on AC TIG with continuous high frequency on my 50A breaker. A 60A breaker might do better.