Had a customer in Adelanto CA, in the middle of nowhere between LA and Vegas, with a soft starter on a large blower at a melt furnace giving him trouble. The soft starter kept tripping on shorted SCR, but when his shop electrician tested it, they were fine and the soft starter worked fine again after testing it. As a general rule when SCRs fail, they fail for good, there is no "maybe" about it because they become full time conductors and stay that way. But it kept happening over and over, so I assumed it must be a problem with the detection circuit and sent them a new board. It still happened, so I took a new soft starter out there and put it in. Worked fine, and as I was driving home, I got a call that the same fricken thing was happening with the new one. Tripping on shorted SCR, but the SCRs tested out fine, then the starter worked again. I spent the rest of that day and most of the next out there, watching it, and sure enough, every now and then it would trip, only when not running the motor (that's the only time you can detect them anyway). I did figure out that testing had nothing to do with it, powering it down to do the test was resetting the fault and it would work for a while, then trip.
After watching it do this over and over all day long the second day, I finally noticed a sound in the shop changing at the exact same time as the trip. I followed the sound and found a 25HP air compressor running outside that had just turned on. So I experimented and sure enough, every time the compressor turned OFF when the soft starter was not running its motor, the soft starter tripped. What was happening was that the coil on the compressor starter was line voltage, they ran the 480V line right through the pressure switch to the starter coil. When the coil de-energized, it gave an inductive kickback spike of voltage on the 480V line that caused the SCRs on that phase to turn themselves on from a brief fraction of a second, but it was long enough for the Shorted SCR detection circuit to think that because the voltage drop across the SCR went away, the SCR was shorted. It's something I had forgotten could happen because I had not seen a line voltage coil on anything in decades! It was common at the still mill where I started as an electrician in the 70s, but people stopped doing that.
So I knew I needed a snubber on the coil, but there is nothing in that town but a Radio Shack. They didn't know what a snubber was, nor did they have an MOV, so I had to buy resistors and a capacitor to make my own. But because it was radio shack, selection was limited and nothing was rated high enough for 480V. So I had to string together a bunch of smaller resistors and parallel some caps to make it work and connect them with solder, making a string of parts about 6" long. I put them into a piece of heat shrink tube with just the leads peaking out of each end to attach to the coil terminals. It was butt ugly, but it worked! I told the plant electrician to order a real coil snubber for that starter from Sq. D, because I didn't like the chances of my string of Radio Shack parts lasting.
4 years later I was driving through there coming back from Vegas and I stopped in to see them, the $12 string of parts was still there, still working.