Voltage Relay Pickup Problems

Status
Not open for further replies.

big john

Senior Member
Location
Portland, ME
Had an interesting problem involving an Over/Undervoltage protective relay that was preventing an ATS system from working.

Everything pointed back to the relay not actuating, went and bench-tested it and the relay would pick up at 102VAC when slowly ramped.

In service it was getting instantaneous 114VAC applied, but the contacts would not pull in reliably. It's obviously a weak coil because if I give the relay the slightest tap, it will close and maintain.

I've read foot-notes for voltage relays that warn when the "actuating quantity" is too close to the pickup voltage, this may happen, but I was hard pressed to explain why it worked on the bench and not in service, and I don't think it impressed the customer much.

Any ideas?
 

LEO2854

Esteemed Member
Location
Ma
Had an interesting problem involving an Over/Undervoltage protective relay that was preventing an ATS system from working.

Everything pointed back to the relay not actuating, went and bench-tested it and the relay would pick up at 102VAC when slowly ramped.

In service it was getting instantaneous 114VAC applied, but the contacts would not pull in reliably. It's obviously a weak coil because if I give the relay the slightest tap, it will close and maintain.

I've read foot-notes for voltage relays that warn when the "actuating quantity" is too close to the pickup voltage, this may happen, but I was hard pressed to explain why it worked on the bench and not in service, and I don't think it impressed the customer much.

Any ideas?

Was the bench test in your shop on a different electrical system?

I'm thinking that on site there is a chance that the transformer may be loaded and causing the voltage to drag down because of being loaded to the max.

Also, how old is the ATS? Is it located in a corrosive environment?
 

big john

Senior Member
Location
Portland, ME
Different electrical systems, but the equipment is calibrated, I gotta trust the voltage readings. The bench test was with a CMC-356 which is gonna be as accurate as it gets, and in the field I was using a Fluke 87V.

Definitely old gear. General electric stuff from the '70s.

It's just more than a little disconcerting to do a standard relay test on a piece of equipment, say it's good, and then install it and watch it fail in service.
 

JoeStillman

Senior Member
Location
West Chester, PA
Is it electronic or electro-mechanical? Was the relay oriented in the same physical position during the test? I'm imagining it installed vertically on a backplane in the gear, but laying on it's back on the workbench.
 

big john

Senior Member
Location
Portland, ME
Is it electronic or electro-mechanical?
Instantaneous electromechanical: No timing disk.
...I'm imagining it installed vertically on a backplane in the gear, but laying on it's back on the workbench.
I'm not that lucky.

The difference I know of it it was bench tested on perfect 60Hz utility power but failed on generator power: Not sure how much of a frequency problem would have to exist before it started messing with the coil strength, though.
 

SG-1

Senior Member
Ok, Big John it's time to come out of the dark ages of electro-mechanical & start using some modern devices like a Bender 420. :happyyes:
http://www.bender-us.com/products/voltage-relays.aspx#Description

Unless your like me, and figure the first electrical disturbance will fry it's brains out :eek: or not.

Before you decide the device is bad you may want to take a look at the sine wave produced by the generator if you have the equipment. There is something the device does not like going on.
 

big john

Senior Member
Location
Portland, ME
...Before you decide the device is bad you may want to take a look at the sine wave produced by the generator if you have the equipment. There is something the device does not like going on.
Yeah, that's really the direction I'm leaning. There's a frequency relay in place, unfortunately I assumed that was working properly and would've tripped if there was a problem.
 

ATSman

ATSman
Location
San Francisco Bay Area
Occupation
Electrical Engineer/ Electrical Testing & Controls
Your Fluke 87 should be able to measure generator frequency.
Is the voltage relay part of the ATS because if it is then the relay may be the VFS
type (Voltage-Frequency Sensor) and frequency sensitive.
 

Phil Corso

Senior Member
BigJohn...

Typically, an Under/Over Voltage disk-type relay has two contacts; on called left, the other right. Each contact's response was different depending on Model! Also for some types, pickup voltage was influenced by frequency.

Can you provide the ANSI device number related to function, i.e., only voltage monitoring. Or, could it have been used to monitor the neutral-to-ground voltage of a Y-connected generator or transformer?

Regards, Phil Corso
 

big john

Senior Member
Location
Portland, ME
...Can you provide the ANSI device number related to function...
It's a GE 27 undervoltage relay. It has NO and NC contacts and they're using both so it's not really just undervoltage, because it also depends on correct voltage being present in order to close some loads.

It's a GE NVG11A that just uses a group of three telephone-relays to provide instantaneous pickup for all three phases.
 

SG-1

Senior Member
I could not find an instruction manual for that device on the web. GE NVG11A
Can you see each of the telephone relays when it is in operation ?
Is just one of them unhappy ? or
Do they all pick up & drop out together ?
 

topgone

Senior Member
I could not find an instruction manual for that device on the web. GE NVG11A
Can you see each of the telephone relays when it is in operation ?
Is just one of them unhappy ? or
Do they all pick up & drop out together ?

It's supposed to be NGV not NVG. Here's the link.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
150123-1104 EST

big john:

Your relay circuit consists of an adjustable series resistance for setpoint adjustment. This feeds a bridge rectifier with a capacitor in parallel with the bridge input.

I think the capacitor can be ignored for your purposes. It stated purpose is reduction of short time transient voltages. The combination of the adjustment resistor and the shunt capacitor is a low pass filter, and I expect that this is substantially above 60 Hz. Thus, no frequency affect on the fundamental frequency component, 60 Hz.

The bridge rectifier DC output is not filtered other than by the inductance of the relay and its shunt diode.

There is a Zener diode in series with the relay coil. This improves the threshold sensitivity and stability of the relay function. But this could produce substantial errors in calibration with a badly distorted input waveform.

Current does not flow to the mechanical relay coil until the Zener conducts. This occurs when the instantaneous voltage input to the Zener and relay exceeds the instantaneous voltage drop across the Zener and relay.

Could you have a sufficiently bad AC waveform I don't know.

As an experiment I would take one of the relays and open a lead to the relay coil, insert a Simpson 260 in DC A or mA, as appropriate, in series with the coil to measure its average DC current. On your calibration bench use two AC DVMs, one a Fluke 87 RMS reading and the other a Fluke 27 average reading, to measure pullin and dropout voltage points along with the Simpson DC coil current measurements. Using a low power Variac, about 100 W (fine wire), run the same experiment from the generator power source. If there is sever waveform distortion it should show up in the measurements.

According to the relay instructions the pullin is not more than 10% above dropout voltage.

A single one of your voltage relays is only useful for dropout or overvoltage detection, but not both functions in the same circuit. This has nothing to do with the available contacts.

Relay pullin and dropout are based on average DC current to the relay coil.

.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
150125-2134 EST

A simple experiment with a Fluke 27 (average reading) and a Beckman 4410 (RMS reading) AC voltmeter.

The experiment is voltage measurement on the output of an inexpensive UPS with a computer load, and with and without AC input.

Code:
             Fluke 27 (average)         Beckman 4410 (RMS)

AC In ON           122.4                      122.2           direct AC line

AC In Off        108 to 110                 119 to 120        battery backup
During battery backup voltage is not very stable.

This experiment is useful only to show how the average and RMS readings could be different between AC line and generator output. But i don't expect generator output to be as bad as the UPS output. On the other hand is the unknown effect from the Zener diode.

.
 

big john

Senior Member
Location
Portland, ME
GAR, I appreciate the detailed reply. As it stands it looks like everything still points to generator PQ problems to explain the difference in pickups.

Since these aren't my relays, I can't dismantle one, but everything was tested with calibrated true RMS meters, so again, I have to trust my voltage measurements or else everything is suspect.

Unfortunately, whether I go back and do any additional testing is in the customer's hands at the moment, so I won't know anything for sure until they give me the okay.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
150126-0953 EST

big john:

If at least you make voltage measurements at the generator with RMS and average reading meters, then there may be useful information. The readings of both meters should be obtained with your test 120 V sine wave, and just note any offset between the meters and the direction. The change in offset between the meters from your test panel to generator output will be some indication of waveform distortion. A portable scope could be very useful.

A Simpson 260 could be used as the AC average reading meter, but I might expect, by chance, a bigger offset between the Simpson and the RMS meters because of the lower accuracy of the Simpson. But the Simpson with a taut band suspension is quite repeatable. What we are looking for is a change in the offset from the sine wave source to the generator output.

As an experiment I created an a test circuit using a P&B KUP 24 V DC relay, 1N4005 shunt diode, a 1N4747 12 V Zener, a bridge rectifier from a 28 V adjustable transformer. This type of relay is not good for drop-out detection because of its large on-off hysteresis.

With a DC supply in place of the bridge the input voltage was 29.0 V for pull-in resulting in 17.2 V across the relay coil.

With an AC sine wave to the bridge of 28.9 V RMS full pull-in occurred. There was much chatter as this voltaage is approached before solid pull-in. The input voltage is in the ballpark of what is expected. The chatter implies there is not enough inductance in this relay to get good DC current averaging.

.
 
Status
Not open for further replies.
Top