relays driving me nuts

Hi All,
I'm banging my head against the wall regarding using three relays to sense phase loss on a three phase wye-connected service. Please see the attached PDF for the two schematics of what I've tried. This seems so easy in theory, but the damn electrons won't behave.

Option A:
I tried this first, and it works well until a phase gets turned off. In theory, if I lost a phase, the two relays that derive power from the affected phase would drop out, leaving one relay powered. The PLC would examine the relays' states to determine which phase dropped. In reality, when I disconnect a phase, the one relay that's supposed to stay on chatters on low supply voltage. Some weird voltage leakage is sort-of driving the relay that should be on, but not really.

Option B:
By creating a star point, and grounding it, it seemed that the relay connected to the phase that drops will turn off, indicating directly which phase is lost. In reality, thre-phase loads in the cabinet are creating enough voltage on the "dead" phase to keep all three relays energized.


Here's my question: I've been simulating a phase loss by temporarily replacing the three-pole breaker in the service panel with three single-pole breakers and tripping one breaker. What I can't simulate is an actual phase loss event, where the utility drops a phase (it happens all the time around here). If it's not just me tripping a breaker and effectively isolating that phase at the panel, and instead the whole phase back to the transformer is dead, will I actually have a problem? I think the other loads that are bound to be on the dead phase will effectively pull it to ground, but I'm not sure.

What black magic do the phase-loss relays use to get around this issue? I'd just use an off-the-shelf phase loss relay, but I'd really like to have more info than just "power problem."

This driving me a little nuts and it seems so simple. Ideas, anyone?


Thanks,

rpoet

About option A:

I have some older Telemecanique relays (coil 230Vac) that even pull in (activate) when you only use 130 Vac. I guess that's what going on for this option.

You may want to look at something like these:
http://www.eaton.com/Eaton/Products...Relays/PhaseMonitoringRelays/index.htm#tabs-2

<Disclaimer: I work for an Eaton Electrical distributor>

I NEVER use relays for this - too unreliable. Use a power failure relay - they are not that expensive. I only use the type that indicate loss of phase, phase sequence and assymetry. Do not shirk on assymetry as standard types can be fooled by motor loads. Assymetry should also be adjustable.
I am currently using the Schneider RM4TA02 relay - assymetry is adjustable to 15%.

Your idea of using relays would work if you only had resistors for loads.

I assume you have motors. When single phased the "deenergized" winding will generate. This is essentially how a rotary phase converter works - it converts single phase to 3 phase and does so quite well.

I would second Norms recommendation for a phase loss relay.

Dan Bentler

Typical unit:

http://www.alliedelec.com/search/se...onSearch=D:K8ABPH1,Dxm:All,Dxp:3&SearchType=0

$46, I bet you spent more than that in time trying to figure a way to make standard relays do the job. Been there, done that.

rpoet said:

What black magic do the phase-loss relays use to get around this issue? I'd just use an off-the-shelf phase loss relay, but I'd really like to have more info than just "power problem."

Click to expand...

They look for phase rotation, not phase loss. They do it by looking at the timing between each leg. Latest units use clock circuits. Here's a short article that shows how to do it using resistors and caps.
Shop for a phase rotation relay. I like the universal one from AD.

I actually installed the relay that Keith recommended in another piece of gear this week, and it's working perfectly.

In this case, I was hoping for more information other than "power isn't good," which is what most protection relays offer; having information about which phase has failed, and why would be nice. Since I have space issues in this panel (and one can't see the info LED on the relay when the equipment is assembled), I will be placing a phase-detection relay farther upstream and I will run a large contactor that will control power to the panel. Otherwise, I'm open to recommendations regarding a power quality relay I could talk to over Ethernet or RS232. I imagine they exist, but are $$$.


-rpoet

A power quality meter is not generally designed to detect loss of power - or not the ones I have used. They measure all electrical values and harmonics.
Generally there are several types of power failure detection relays available and some include rotation, loss of phase, assymetry, under/over volts, under/over frequency. The latter two are a PITA as they go off on under and/or over but are commonly used where there are voltage and frequency sensitive UPSs and other electronic equipment as part of the load.
Here is a link to a pretty good article that describes electrical assymetry - some of the article would be relevant for you and some not.
http://www.google.com.au/url?sa=t&r...o4CQAg&usg=AFQjCNEIiLr9PNR2qmaK08cHljPmgMJKuw

Thanks for the advice, all! I'm going to simply detect if three-phase is available in the panel (via a relay tied across two incoming phases) and provide protection in another location with a dedicated phase-loss relay and contactor. That should satisfy everybody's needs, and keep the gear safe from power issues.

Thanks again for the help!

-rpoet