An HVAC student advancing in the trades recently asked me a question. I was visiting the shop down at the local community college hoping to borrow a little time on the sheet metal brake to make a chassis for an underground cable/pipe locator (more on that later). The student asked about zero sequence currents and how they add up in the neutral conductor for various loads. He asked the question in easy terms.
“I work at a shop with a lot of old lighting, pre PFC (power factor correction). We have strict rules on cutting and splicing the neutral conductor anywhere along the bus, yet the loads are perfectly balanced. How could there be current in the neutral wire with balanced loads?” he asked.
That’s a wonderful question, well worthy of a wonderful answer. But could we do it with simple first-order meters and stuff around the HVAC lab? I opted to just simulate it on BIAS (build it and see). The concept is nothing new, though the proof was a fun experiment that ended with a journeyman electrician looking me in the eye and saying, “I understand that now. I get it!”
We headed off to the parts room to see what it had for goodies. I found some 1200-V bridge rectifiers rated for 50 A with standard recovery diodes. I then found some large electrolytic capacitors from an old motor drive. They were rated at 3900 µF at 450 V and improperly stored without bleeder resistors across them.
I then found some light bulbs (yes, people still use them), Edison sockets, fuseholders, and switches. The supply of THWN wire was nearly endless. Figure 1 shows what we built. The three-phase service in the shop was 480 L-L.
Before moving on, we need to note that this is live circuitry, with 480-V L-L, 277-V L-N voltages. These potentials can be lethal. Working with these voltage levels requires proper training, personal protective equipment, lock-out tag-out (LOTO) hardware, and deeply rooted technical common sense and respect. Do not attempt this without all of the above.
I made sure the student had taken the prerequisite basic electricity course and was at a journeyman level working toward a stationary engineer’s license. As an “instructor,” my skills and comfort level needed to exceed that of the student—and they do. I’ve been working with this stuff since I was a kid. I was trained by engineers, electricians, and linemen.
The circuit was located a few feet from the breaker panel, and we had our own breaker and fusing on the “circuit board.” For analysis, I didn’t want to lead in with anything beyond what the student had in his tool belt. Once the circuit was built, I asked him to measure the currents in each phase and the neutral conductor with his inductive clamp on ammeter as well as the ac voltages WRT neutral and the dc voltages across the bulbs (see the table).
As he closed the jaws of the ammeter around the neutral conductor, he was startled. The magnitude of the current was nearly twice that of a phase leg. I figured I’d add to that bewilderment and I asked him to measure the frequency of the current that he was seeing. His jaw dropped.
“I’m measuring 180 Hz!? How can that be? The power company only gives us 60 Hz!” he said.
“Yes, you are. I agree with that,” I replied.
I had what I needed—a bewildered student driven by curiosity. We then went to the white board, where I briefly explained the rules for three-phase power. Any phase current can be represented by a sum of zero, positive, and negative sequence currents.
I drew the current waveform into the bridge rectifier and discussed peak-charging the large electrolytic capacitors. The peak current is a sum of harmonics. The odd third multiples of the fundamental are called triplens. They add in the neutral conductor in phase.
I drew the waveforms including the fundamental component at 0°, 120°, and 240° phasing. Then I drew the third harmonic incident on each of those waveforms. The student instantly understood.
“That’s where the 180-Hz current came from! They all add up in the neutral!” he said.
We went back to the shop, removed the LOTO gear, and re-energized the circuit for a bonus round. This is reinforcement time. With the switch closed in the neutral line, we see high triplen currents. What happens when you open the neutral switch, akin to cutting the neutral bus to splice in a new branch?
You interrupt a current flow. There is then a potential difference between the bonded neutral conductor on the branch and the load neutral. We measured this potential difference as 42 V RMS, 180 Hz. Also, the light bulbs get dimmer.
What then happens when you open the switches that disconnect the capacitors from the bridge? The triplens go away and the neutral current goes to zero. At that point, it doesn’t matter whether the neutral switch is open or closed. This is roughly analogous to adding PFC to uncorrected loads, at least in terms of waveforms and neutral currents.
I brought in my Fluke 43-A power meter and showed the student the currents and voltages and phase relationships. It all agreed with the stuff on the white board. Figure 2 shows a reproduction of the phase current and neutral current waveforms that we saw.
The last test I wanted to run but didn’t have the time to get to was a negative sequence. To establish one, we would have had to reconfigure the bridges as one lumped diode (positive terminal = cathode, negative terminal = anode) and rewire things a little bit. This then would have established half-wave rectifiers into the loads. The even harmonics would have caused a negative sequence that rotated against the fundamental phasor.
Helping this student was a lot more fun than I expected. My first notion was to cut to the math. I’m glad I didn’t. I approached the curious student as I would approach any other problem in the engineering world—I gave it my best shot and tried to make sense out of it in a mutually acceptable language and arena. It took a few extra hours, but it was nice to be able to help.
I bring this up to the electronics community as a means to reinforce old knowledge in a new way. The switched mode power-supply (SMPS) designs that we work on can be realized as a 500MCM bonded neutral conductor in a duct feeding two dozen subpanels of lighting circuits.
I’d want to know what was going to happen before I ratcheted down my cable cutter on that conductor, as would anyone else working on that branch. I simply couldn’t leave that as a mystery to someone who would service those systems. It’s an old school notion, but I always try to offer a comprehensive solution, from the prototype PFC circuit to the final revision, up until decommissioning from service.