Tag Archives: welder

Inappropriate Uses for a Welder

Last week at work I was given the challenge of designing a custom heat sealing tool for one of our automated manufacturing systems. Since a lot of impulse heat sealers use strips of Nichrome foil as heating elements, we thought we’d start there. (As a side note, unless we were totally ripped off, Nichrome foil is apparently rather expensive; a 34″ x 6″ x 0.005″ sheet cost about $600.)

A few “back of the envelope” calculations indicated that we’d need to pump about 180 amps through the Nichrome; the resulting voltage would be about 12V. That’s quite a large power requirement – 2.16kW to be exact – enough to run a couple hair dryers on full blast. So the question was – how are we going to deliver this enormous amount of current? A 10:1 step-down transformer would do the trick, but finding one rated for more than 2kVA (at a reasonable price) was a bit of a challenge.

Lincoln Electric K2535Enter the Lincoln Electric 225A AC/DC (K2535) arc welder. Our machine shop picked one up about a year ago with year-end funds – hooray accounting! It’s not often used, so I decided to borrow it for a quick test. It’s really an impressive device – all you have to do is set AC/DC, then dial in your desired current (settable from 5 to 230A at 1A increments) and hit the foot pedal. There are a number of other settings available which I’m sure any legitimate operator would appreciate.

Anyway, my intention was to use this lovely device as a glorified power supply running in constant-current mode. To do so, we purchased a new plug and wire for the working electrode that would allow us to attach to the Nichrome via wire lugs. The ground electrode simply clamped onto the other end of our test apparatus. After frustratedly poring through a ton of Amphenol datasheets, I finally found the right connector for the foot pedal (parts MS3106A18-12P and MS3057-10A). With this connector I could wire up my own electronic control (a double-pole relay plus driver circuitry) to precisely set how long the welder should operate.

So down to testing! We cut ourselves a strip of Nichrome foil approximately 5″ x 0.5″ and strapped it down to a block of aluminum covered with an electrically insulating layer of Kapton. A couple thermocouples were stuck between the Kapton as well. With everything set, we clicked on the welder and hit the pedal. As it turns out, nothing exploded. In fact, set to DC, the welder made a very nice high-current PSU. Set to about 100A, we had our Nichrome glowing red-hot within a second or two (we calculated a strip resistance of about 0.1 ohms, so this means we were heating at 1kW).

A thermal camera indicated a maximum temperature of about 600C. The assembly, after quickly cooling, looked like this in the infrared:

Nichrome Strip Test ApparatusThe working (positive) electrode connects on the left and is held in place with a couple chunks of plastic (which are starting to get a little warm, as indicated by the orange/yellow false color) . The ground electrode is clamped to the right side of the bar.

It’s interesting to note that this welder begins each operation with about half a second of high-frequency (~2kHz) AC. My best guess is that this is designed to start an arc when one is actually welding:

Heater Strip VoltageThe scope screen-shot above shows an initial spike in current when the welder first kicks on, followed by a period of high-frequency pulsing. While this is happening you can actually hear a sort of “hiss” from the welder, which I imagine is some type of internal spark gap.

As far as the Nichrome is concerned, this is no problem. The strip heats just as well with AC as it does with DC. However, this 2kHz pulse can radiate and wreak havoc on nearby electronics. During testing I noticed that every time the welder started, an AMREL programmable load across the room would beep. It never actually produced an error message or malfunctioned, it just beeped (and beeped consistently). In addition, the first USB DAQ device I tried for controlling the welder would occasionally reset once the welder turned on. This was rather alarming the first time it happened, because the device failed in the “ON” position, leaving me scrambling to unplug the relay before the Nichrome melted. The machinist I work with also told me that “back in the day” he had a welder that would make his watch run faster…

In summary, we wound up with some interesting data on the heating rates of Nichrome. We may actually require a bit more than the 180A originally specified, but this may still be in range for our welder. It’ll be interesting to see how this plays with the rest of our equipment.