3
\$\begingroup\$

I'm working on a friends inoperative Miller 300 Amp DC inverter Tungsten Inert Gas (TIG) machine. These machines convert 240v single phase A/C to DC, then use IGBT chips to generate controlled DC output. Rated output for this machine is 250A @ 30 VDC. The actually welding amperage range is 5-300 amps, wit maximum open circuit voltage at 95 volts. Typically you would set a maximum output current on a panel, then further control that output with a footpedal (full pedal = maximum set output current).

The control boards seem to communicate well with the front panel. There are no faults set. There are no signs of any component overheating, nor any sign of blown capacitors. Note: Factory repair is priced in the $1000 of dollars. My suspicion is some key high power components are not functioning correctly. I'd like to verify this, and I was hoping to verify this in circuit. I'm reasonably active on the welding forums and component failure is a known cause for non-function.

Tools I have..

  • Old analog two channel oscilloscope, with new aftermarket probes (x10, x1), a device I have little experience with.
  • Typical $100 multimeter, nothing fancy.
  • Fluke 36 Clamp Meter (rated at 600v, 600A AC, 1000A DC)

I'd like to check the diode bridge rectifier, the IGBT chips and the output diodes in circuit. My initial thought was to just use the oscilloscope and look at the interim signals.

And then I pause, to think. Is this even safe? I don't want to hurt myself, or any tools or any devices being tested. Can I safely probe up to a 240 VAC signal with the oscilloscope? Is there an isolation device available? If not, is there an alternative way to read a full rectified 240v output, or troubleshoot a switching power supply board with the above rated power?

\$\endgroup\$
  • 1
    \$\begingroup\$ It may be possible to probe the high voltages safely with your scope and probes. Both should have markings like "CAT II 300V" or "CAT III 1000V" on them. On the scope it would be near the inputs. If so, post what it says. \$\endgroup\$ – Jack B May 9 '17 at 20:55
  • \$\begingroup\$ Also note that it is possible to get x100 and even x1000 probes for a 'scope for testing even higher voltages. Ironically, the "multiplier" is not multiplying sensitivity, but dividing it. Think of the rating as "x to 1", i.e. x10:1. A 100v signal using a x1 probe would output 100v into the scope (possibly damaging it.) But a x10 probe would output 10v, a x100 probe 1v, etc. Be very careful probing around any electrical device operating above 50v or so, especially high-power devices. \$\endgroup\$ – rdtsc May 10 '17 at 2:04
4
\$\begingroup\$

I used to repair welding machines for a living. First thing I would do was to disconnect the things from the power completely. Then I'd disconnect the big diodes and SCRs, etc. and just test them with a meter. 95 times out of 100 one or more of the big semiconductors failed open (you'll know if they fail shorted).

You might have to disconnect one side of them and it could get tedious with 3 phase welders but with such a ratio I considered it worth it instead of troubleshooting with wave-forms right off.

\$\endgroup\$
  • 1
    \$\begingroup\$ How do I accept both the mark b and the Joren Vaes responses? I'm going to follow the mark b recommendations for the safest route to success... \$\endgroup\$ – zipzit May 10 '17 at 0:46
4
\$\begingroup\$

You always need to watch out with probing mains-powerd stuff with an oscilloscope. Dave Jones, from the EEVBlog, explains:

How not to blow up your oscilloscope!

To safely probe this, you need to either verify it's isolated, or use an isolation transformer to isolate the device under test.

\$\endgroup\$
  • 5
    \$\begingroup\$ Using an isolation transformer on the oscilloscope can be dangerous, it would usually be a better idea to use the isolation transformer on the device under test (if the isolation transformer can take it) \$\endgroup\$ – Jack B May 9 '17 at 20:53
  • \$\begingroup\$ I just went thru the entire video.. and yes, I was just about to do that way wrong. But here is my question. Can I just use my probe in single wire mode (tape back the ground jumper?), knowing full well we've got connected ground between the 240 main on the welder and the 120 main on the Oscilloscope? How ugly would that look? \$\endgroup\$ – zipzit May 9 '17 at 21:07
  • 1
    \$\begingroup\$ This advice to use an isolation transformer on the scope is bad. Please see this answer: electronics.stackexchange.com/questions/73991/… \$\endgroup\$ – replete May 10 '17 at 1:31
  • \$\begingroup\$ It was a mistake on my part, I edited it to "float the device under test". \$\endgroup\$ – Joren Vaes May 10 '17 at 6:44
2
\$\begingroup\$

You know about the kilovolts on the tungsten tip, right?

I haven't repaired TIG machines, but I know that they contain a "high frequency unit" supplying ~3KV AC, which is more than enough to fry all your test equipment. Identify these sections and connections so you can avoid them (or perhaps shut that section down entirely.) Some connections will destroy your test equipment if probed.

\$\endgroup\$
  • \$\begingroup\$ I'm well aware of the high frequency unit. In this case its a old school point-gap-point system. Because the TIG welder has a stick weld option (which totally bypasses the pedal controller and the hi-frequency start system) that really is not an issue. \$\endgroup\$ – zipzit May 10 '17 at 3:28

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.