1
\$\begingroup\$

I've got a request to fix a snafu made by engineering when they designed this layout. The end-user wants 3 lights that monitor the voltage status on each of their phases to ground (neutral is bonded to ground in the panel).

Since the incoming utility is 208/120Y, they just connected a 120VAC light from each phase to ground and called it a day.

However, we include transformers (labeled XRFMR) for control power in our cabinets for (somewhat) isolated control power. Most of you will see the issue right away here... Whenever they lose A or B phase, the voltage will backfeed through the primary of the transformer and keep the phase light illuminated.

I've been wracking my brain here on simple solutions, but I can't come up with anything that gives us individual lights for each phase. There are numerous phase-loss relays out there that monitor all 3 phases and give you a single NO/NC contact, but nothing with a contact per each phase.

Any ideas?

Example Circuit

\$\endgroup\$
  • 1
    \$\begingroup\$ Can't you connect the lights from phase X to neutral, where X is the phase to be monitored? \$\endgroup\$ – TheComputerGeek010101001 Dec 17 '15 at 22:18
  • \$\begingroup\$ What's a CPT? Remember you have an international audience. \$\endgroup\$ – Transistor Dec 17 '15 at 22:58
  • \$\begingroup\$ changed CPT to control power transformer \$\endgroup\$ – Darksabre Dec 18 '15 at 4:40
  • \$\begingroup\$ @user94774: Yes, that would work normally, except we have a transformer with primary taps on the A and B phase. Since a transformer is basically a long run of wire coiled around a metal, you can still get conductivity (backfeeding) from A to B (or B to A) if the other phase drops out. This would keep the light illuminated, even though the phase is really dead. \$\endgroup\$ – Darksabre Dec 18 '15 at 4:44
  • \$\begingroup\$ Connecting the lights to equipment safety ground is not good. For what you're trying to do there, connect them phase to neutral. Never attach to equipment safety ground unless you are installing a main service panel, the one place you bond to ground. This is totally separate from any advice to connect them phase to phase. \$\endgroup\$ – Harper Jan 28 '18 at 4:53
3
\$\begingroup\$

Connect three lamps across the phases, rather than to ground.

One lamp between phase 1 and 2, the second between 2 and 3, and the third between 3 and 1.

If a phase loses power, one of the other two phases may backfeed on it, but since it will mostly be in phase then the lamp will go out between the phase that's lost power and the phase that's backfeeding it.

\$\endgroup\$
  • \$\begingroup\$ Yeah, I thought of that as well, but if you lose one phase, you lose both lights that share that phase, correct? \$\endgroup\$ – Darksabre Apr 18 '14 at 15:46
  • 1
    \$\begingroup\$ No, you lose one light. It doesn't tell you which phase failed, but narrows it down to two possible phases. At that point you'll have to do additional testing, but at least it does tell you there's a failure. \$\endgroup\$ – Adam Davis Apr 18 '14 at 18:59
  • \$\begingroup\$ Hmm, I could see how that would work for Phases 1 and 2 since they're coupled through the transformer primary. But what if I lost Phase 3? I imagine I'd lose the second and third lights in that case. \$\endgroup\$ – Darksabre Apr 21 '14 at 13:16
  • \$\begingroup\$ @Darksabre Yes, in your case that is what will happen. If phase 1 or 2 goes out, you will lose the one light between the two phases. If phase 3 goes out, you'll lose both lights connected to that phase. So you'll have a little more information than otherwise. Having thought about it more, I don't see a nice off-the-shelf solution for you, though. Were it me, I'd wire a 3 phase plug in the location where you need the indicator, then use a power analyzer along with the lights. When any light goes out, have them look at the power analyzer to figure out the exact issue. \$\endgroup\$ – Adam Davis Apr 21 '14 at 16:29
  • \$\begingroup\$ ...and arrange the three lights in a triangle representing the wye, with a little wye graphic between them so it's visually clear what's happened... \$\endgroup\$ – TDHofstetter Aug 24 '14 at 20:35
0
\$\begingroup\$

I'm not sure if you want to do this sort of hack in an industrial setting, but you can produce individual phase indicators fairly simply. Get a commercially available monitor which has indicator LEDs for the 3 phases. Open it up, remove the LEDs, and wire their leads to 3 optocouplers (you might need to replace the LED limiting resistors as well). Then use the optocouplers to drive relays. It's not pretty, and it's a kluge, but if you're careful with workmanship it ought to be fairly reliable.

\$\endgroup\$
  • \$\begingroup\$ Heh yep - while it sounds like fun, I don't think UL would approve of that type of setup :( \$\endgroup\$ – Darksabre Apr 18 '14 at 15:47
0
\$\begingroup\$

If you don't care about sequence test, you could detect any fault imbalance with 3 equal R's , one to each phase and the common Y junction to a light of low current (back-back LEDs or photocoupler switch)

You can also use two lamps and a cap of equal impedance in a Y connection. The two lamps will be chosen to be half power when balanced. (Dim) the cap gives phase shift such that any imbalance causes one light to be brighter than the other.

\$\endgroup\$
0
\$\begingroup\$

My initial thought was that a diode (actually three) wired in series would solve your problem -- a diode being a one-way valve that only lets electricity go one way -- it's how a car alternator can generate 3 phase AC and then (with 6 diodes) produce DC. But a diode would only get one side of the sine wave -- and would let the other one through, which got me to thinking:

The question I have is why is that light backfeeding? If you loose the phase, if the wire is cut/disconnected, the wire still isn't grounded, and even if it was, there shouldn't be a voltage between it and ground, so assuming your ground is sufficient (is it?) how could the current backfeed from the ground connection? Now what if something upstream was connecting two of the phases together -- e.g. A is both disconnected from A and connected to B.

In that situation, the light wouldn't be backfeedng -- it'd be running off B, just like B is. Could a defective 3 phase motor do this -- I don't know.

\$\endgroup\$
0
\$\begingroup\$

I know it's a bit old question, but I've run into a similar situation. Maybe you could add a phase detector between A & B. In case of failure, the phase detector will lit.

Edit: This can be combined with the solution proposed by @Adam Davis in order to obtain the exact faulty phase.

\$\endgroup\$
0
\$\begingroup\$

I suggest the easiest answer is to use neon bulbs with correct (100-200K ish) resistor very small but high-over-rated voltage incandescent indicator bulbs with resistors. they will light when phase is active. Feed back will be so minimal that it is nearly non existent. Not real complicated so you can see it will work and last. Don't trust diode array setups or LEDs as they are not electrically durable over the long hall and are killed by spikes.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.