I bought my Synchrowave (SN preface code LG) new in June of 2007. It probably has less than 25 hours (no more than 50 for sure) total power-on time on it in my home hobby shop. Most of my (occasional) shop time is spent machining; welding is not done very frequently and is mostly small items. Grinding/filing/machining doesn't take place next ti the welder.
I'd had an old Miller 330 that I replaced with this new Synchrowave when the 330 died in the middle of a project. I'd been hoping that spending the money on a brand new machine would get me trouble-free operation for a very long time considering how infrequently I use it. The 250DX worked fine until it didn't.
A couple weeks ago I was doing some aluminum welding and the machine started acting up. I tried to troubleshoot it (no dangling wires/blown capacitors, good power to it, no grinding particle/dust build up inside) and after a talk with Miller tech support I arranged for a service call, as Miller figured it was probably the main circuit board (what I'd consider a motherboard since it has the microprocessor and control circuits on it) that was at fault.
Yesterday, after 3.5 hours of tech time and a $900 replacement board (a newer version that also included a replacement of different specification for one of the giant resistors in the machine as well as an extra bridge rectifier) the 250DX was fixed.
The original motherboard looks like it gets a 3 year warranty, so that was up June 2010.
Is this a common failure? It appears Miller had reason to make the replacement a new board (different part number, blue phenolic for the board instead of green, extra parts included with the replacement board) so there must have been some shortcoming in the original part
The original board in the welder was #209877F and I was told the current part uses #237587. There are also different software/revision codes showing up when the machine is turned on.
It would be nice if Miller could add a runtime clock to their control boards and factor the actual running hours into their warranty. It seems like that should be pretty cheap to implement/program at a time when they do a board revision. I'd think they might even find the clock data useful for determining mean time between failures (MTBF) for their manufacturing and design process.
Sure, probably most people who buy a welder at this price level are going to use it a fair amount because they are buying it for a commercial shop and the machine needs to pay for itself. But I suspect that I'm not the only hobbyist who thinks "I'll go with a nice industry standard machine because I can squeeze it into the hobby budget and it will save me money and bother in the long run and bring me peace of mind."
Unfortunately, Mr. Murphy seemed to be watching when I made that decision and between the parts and labor I've just spent an additional 50% of the purchase price of the welder.
Intellectually I can appreciate that sometimes things just break, but "just breaks" combined with "oh, we've got a new and improved replacement part for that with additional replacement parts and new software as a part of the package" makes me wonder about the quality of the original part and whether there should have been some sort of recall and/or warranty extension on it.