I worked the first 18 years of my career at a small company that makes automated testing equipment, primarily for the fluid power industry. An example of their application are the wheel drive motors for Bobcat brand skid-steer loaders. Each motor has to pass one of these test stands before it can be put on a loader.
In the early '90s, we built a hydraulic pump/motor test stand for the China Railway Ministry in Beijing. It was going be installed at a maintenance depot. The prime mover was a large 3-phase motor that operated in both motoring mode to drive pumps, regenerative mode to load hydraulic motors, and was controlled by a variable speed drive. Industrial 3-phase mains in China are 380V, and since the drive and some other components are designed to accept 480V, we supplied the machine with a large step-up transformer assembly. To make sure the supply compatibility worked out right at our factory in the US, all of our testing was done with 380V, 50 Hz input being supplied by a 500 kW Cummins diesel genset.
Commissioning these machines at a customer site is usually a multi-day affair for a small, mixed-discipline team. In this case, it was just two of us. We arrived on a weekday and took most of two days to remove shipping mounts, install packed-away components, and ensure that the utilities were properly connected. In performing the light-up procedure, everything went fine until we tried to start the main drive and got nothing. A peek in the drive cabinet revealed that it had faulted for undervoltage.
I checked the drive manual and found that it tolerated only about ±5 percent deviation on the input. With the mains sitting at about 350V, we were nearly 10 percent low at the feed to the drive. After some head scratching and chatting with the customer, we decided that due to issues outside our control, we’d just have to live with low mains voltage. Not a problem, though, because the step-up transformer had several taps on the secondary. I found a set of taps that got me pretty close to 480, moved the output leads to them, and that did the trick -- until the weekend.
On Saturday, the drive faulted for overvoltage. With the industrial load on the power grid reduced on the weekend, the mains voltage rose to around 410V. The good news is that the customer didn’t expect to use the machine much on weekends, so I left the transformer tapped for low mains and the customer with explicit instructions on how to change it if necessary. It all worked out fine in the end, as the customer beefed up the supply a few months later.
This entry was submitted by R. Andrew Thomas and edited by Rob Spiegel.
Reagan Thomas has done electronic and software design for embedded applications, industrial controls, and automation. He is currently a software engineer in the defense industry working on sensor data distribution on computer networks.
Tell us your experience in solving a knotty engineering problem. Send to Rob Spiegel for Sherlock Ohms.
Click here to access the Sherlock archives.