We've had annoying failures of electronic ballasts in the fluorescent ceiling fixtures in the children's museum where I volunteer a couple of days a week. And, since the staff is aware of my power quality testing experience at Fluke, they asked me to look into the matter.
I've long been a proponent of starting with simple checks first when troubleshooting an electrical problem. Most of them are the result of loose connections or poor grounds. So, it follows that I would start my testing with a DMM before resorting to a more complex and expensive monitor or analyzer.
I'd done that in this case, but the simple tests haven't uncovered anything surprising, so now I've decided to take DMM measurements to the next level. I'm going to call in "Old Reliable," the 189 Logging DMM, and apply one of its more subtle (undocumented until now) features.
The museum facility is wired with 3 Phase 120/208, so the lighting fixtures are operating at 120 V, phase to neutral. The wiring is in flexible conduit, which doubles as the safety ground to the fixtures via snap-in connectors - a fairly common technique which meets code and is adequate when classic magnetic ballasts are used.
But, I also know that some contractors who are experienced with electronic ballasts already recommend installing a green-wire safety ground through the conduit and bonding it directly to the fixture to avoid problems such as ours. And the fact that at least one electronic ballast manufacturer cautions that neutral to ground voltage should not exceed two volts leads me to the suspicion that the failures might be due to excessive neutral to ground voltage. I've measured between 1 and 1.5 V rms, but it's possible that there might be fast transients of several volts during switching that could be the cause of our problems.
So, how can the 189 help?
First, I'll use the logging feature to record neutral to ground voltage over an extended period of time, sampling the voltage over 50 ms intervals (3 cycles). Then, to look for faster transients, I'll activate the 189's Fast Min Max at the same time. That way, I'll get samples of the rms level in the normal logging mode, but I will also record one high and one low extreme of any excursions as short as 2 ms over during the logging session. And, I'll have a date-time stamp for any such occurrence.
The method was successfully applied by a dairy farmer and his wife several years ago as they searched for electrical causes of major difficulties they were having with their animals. During one of their many sessions logging stray voltage using multiple meters, a thunderstorm occurred nearby. The simultaneous Fast Min Max recording showed transients much higher than the logged rms data in that case.
You can learn more about their experience by going to:
Electrical - Application Notes and selecting: Is that a tingle she feels? Minnesota dairy farmer pioneers stray voltage measurement.
[An update on the dairy farm - Chuck and Wanda Untiedt, who used the 189 precursor 89-IV meter in their tests, say things are going much better today. Milk production is at an all-time high, and the cows now appear to be well and happy. They credit Fluke meters, and the measurement methods they've applied for helping to bring about the change.]
I'll warn you now, the procedure to be followed is intricate, and must be performed in the indicated order, but it is achievable. I'll go through it, step-by-step in the sidebar below.
So, since the 189 can now be used to log data for a week or more using a relatively new accessory, the BP189 High Capacity Battery Pack, I hope I'll be collecting results during the time that another failure occurs, and can verify my theory.
Now it's off to the museum to try the technique on the ballast failure problem.
If I find anything earthshaking, I'll let you know in a future column.