JDB3
Senior Member
- Location
- San Antonio, Texas
I doubt that the bulbs (4 of them) have been changed, since they have been in the house a short time. Thanks,
lights dimming
- Thread starter JDB3
- Start date
- Status
I doubt that the bulbs (4 of them) have been changed, since they have been in the house a short time. Thanks,
dmxtothemax
Senior Member
- Location
- Brisbane, Queensland, Australia
OK
NEXT STEP
where does the wire from this receptacle go to next ?
Does it go to a j box in the ceiling
Or does it go to a switch
Or straight to the fan hood ?
NEXT STEP
where does the wire from this receptacle go to next ?
Does it go to a j box in the ceiling
Or does it go to a switch
Or straight to the fan hood ?
Could there be device in the house causing this, by any chance?
Had a house many years back where the lights would intermittently pulse. Never the same circuit with no apparent rhyme or reason. Guy I was working for determined that the FPE panel was the cause. Changed it out. Lights started pulsing again. I get the assignment of showing up one morning to stay all day and observe what was going on in the house and when the pulsing occurred. I'm checking receptacles, breaker connections, etc. nothing. Finally, on a trip back into the main house, the lights start pulsing. Start asking what each of the two people in the house just did. Guy, was in his office, had turned on and started to use a plasma printer. Lights pulsed. Turn printer off. Lights stop pulsing. Guy threw out printer.
Had a house many years back where the lights would intermittently pulse. Never the same circuit with no apparent rhyme or reason. Guy I was working for determined that the FPE panel was the cause. Changed it out. Lights started pulsing again. I get the assignment of showing up one morning to stay all day and observe what was going on in the house and when the pulsing occurred. I'm checking receptacles, breaker connections, etc. nothing. Finally, on a trip back into the main house, the lights start pulsing. Start asking what each of the two people in the house just did. Guy, was in his office, had turned on and started to use a plasma printer. Lights pulsed. Turn printer off. Lights stop pulsing. Guy threw out printer.
Michael15956
Senior Member
- Location
- NE Ohio
A person would think the home's occupants should have figure that one out.
Bet your glad that you didn't "determined" if was the FPE Panel.
rippledipple
Member
- Location
- New Jersey
- Occupation
- Electrical contractor
neutral
neutral
neutral
you have to get poco to check neutral leg outside where it comes in, its dirty!!!!!
Sierrasparky
Senior Member
- Location
- USA
- Occupation
- Electrician ,contractor
What is a Plasma printer. Sounds expensive and was thrown out!
JDB3
Senior Member
- Location
- San Antonio, Texas
Really guys. Please proof read and correct your posts because it drives me nuts trying to figure out what you are saying. This is getting worse and worse with mobil devices.
Sorry, after coming home after working in the field (where the jobs are). Straightening up the truck, doing outside chores, & other things. And then getting on the computer, I may not use correct terms for everyone, though it seemed that others had no problem with what I meant to say.
By the way, does your "device" have spell check?
Sorry, after coming home after working in the field (where the jobs are). Straightening up the truck, doing outside chores, & other things. And then getting on the computer, I may not use correct terms for everyone, though it seemed that others had no problem with what I meant to say.
By the way, does your "device" have spell check?
I would first check the meter connections. Take their seal off and check. I had home owner tell me their lighting had been dimming for number of years. They lost Mircowave, dishwasher also.
Found out that the neutral in the meter box never was turned in from day one. This was 10 years old and had been inspected.
Found out that the neutral in the meter box never was turned in from day one. This was 10 years old and had been inspected.
readydave8
re member
- Location
- Clarkesville, Georgia
- Occupation
- electrician
Loose?:?
hbiss
EC, Westchester, New York NEC: 2014
- Location
- Hawthorne, New York NEC: 2014
- Occupation
- EC
That would have been a LASER printer, and an old one at that. I've seen that several times, every time the heating element cycled lights dimmed. Good choice to throw it out, probably cost more in electric than the cost of a new one.
Yes they do and that's often the problem if you have it set to automatically replace words. I understand everything you have going but it's important to communicate clearly. All I can say is I hope nobody sends out proposals and customer correspondence the way I see some posts written here.
-Hal
The guy called it a PLASMA printer. I never saw it. It may very well have been a LASER printer; but, who knows. Printer was the issue. Problem solved. Zipidee doo da day.
JDB3
Senior Member
- Location
- San Antonio, Texas
Got back to the house with the lights dimming today. I took another master electrician (who is in charge of the electrical & lighting @ a Home Depot) with me today.
We checked the voltage at the ceiling box connections & at one of the sockets (as was previously suggested) of the light kit. With all the bulbs on, and with the one bulb out (to check the voltage in the light socket). The voltage stayed constant. Put the bulb back in, and the light kit dimmed after a few seconds.
The builder went to store to get fluorescent or LED bulbs to replace the ones that came with the fan/light kit.
He just called & stated that there is no dimming now.
BAD sensor (or what-ever it is called) to prevent putting to large of wattage of bulbs in.
Another fan / light kit that the owner says does it all the time (except when we are there) never had any dimming. He is also replacing the bulbs in that one!
He is going to contact the utility company to have them monitor the voltage at the house to find out about other lights dimming.
We checked the voltage at the ceiling box connections & at one of the sockets (as was previously suggested) of the light kit. With all the bulbs on, and with the one bulb out (to check the voltage in the light socket). The voltage stayed constant. Put the bulb back in, and the light kit dimmed after a few seconds.
The builder went to store to get fluorescent or LED bulbs to replace the ones that came with the fan/light kit.
He just called & stated that there is no dimming now.
BAD sensor (or what-ever it is called) to prevent putting to large of wattage of bulbs in.
Another fan / light kit that the owner says does it all the time (except when we are there) never had any dimming. He is also replacing the bulbs in that one!
He is going to contact the utility company to have them monitor the voltage at the house to find out about other lights dimming.
gadfly56
Senior Member
- Location
- New Jersey
- Occupation
- Professional Engineer, Fire & Life Safety
Maybe a silly question, but what kind of lights? Incandescent, fluorescent, halogen, or LED?
Whoops! Posted before I got the the end of the thread. My bad.
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
160901-2345 EDT
I have read parts of this thread from time to time. Many questions and comments are not clear, incomplete, misleading, or quite obviously don't apply.
If lights are dimming, varying in intensity, or flickering, and the voltage on the circuit to the lights is not varying in correlation with the intensity varaition, then nothing that preceeds the point of voltage measurement has anything to do with the intensity variation problem. Don't bother the power company. I am assuming that there is not a 500,000 W radio transmitter next door.
A DVM with MIN/MAX capability can probably detect short time voltage variations that would be seen as light intensity variations. My older Fluke 27 is not real fast on MIN/MAX. A reference lists 100 mS.
See the following for DC response http://www.fluke.com/fluke/uses/comunidad/fluke-news-plus/articlecategories/dmms/voltagemeasurements
Fluke does not make an explicit statement about AC, but I assume it is the 100 mS. The 87 has a shorter time, but that is a peak value.
To do 1/2 cycle AC measurements would require a relativly specialized meter. At 60 Hz 1/2 cycle is 8.3 mS.
.
I have read parts of this thread from time to time. Many questions and comments are not clear, incomplete, misleading, or quite obviously don't apply.
If lights are dimming, varying in intensity, or flickering, and the voltage on the circuit to the lights is not varying in correlation with the intensity varaition, then nothing that preceeds the point of voltage measurement has anything to do with the intensity variation problem. Don't bother the power company. I am assuming that there is not a 500,000 W radio transmitter next door.
A DVM with MIN/MAX capability can probably detect short time voltage variations that would be seen as light intensity variations. My older Fluke 27 is not real fast on MIN/MAX. A reference lists 100 mS.
See the following for DC response http://www.fluke.com/fluke/uses/comunidad/fluke-news-plus/articlecategories/dmms/voltagemeasurements
Fluke does not make an explicit statement about AC, but I assume it is the 100 mS. The 87 has a shorter time, but that is a peak value.
To do 1/2 cycle AC measurements would require a relativly specialized meter. At 60 Hz 1/2 cycle is 8.3 mS.
.
Fulthrotl
~Autocorrect is My Worst Enema.~
- Occupation
- E
100 MS should give you real world indications.
the 1 MS off a fluke 87, etc. is going to drive you batty.
the threads pretty scattered, i might have missed it, but
are the lights giving grief LED's?
Peter was having enough problems with an installation of
high end lights he bought a data logger to hunt for it.
edit: my money is there is a floating neutral somewhere in
the house, and the dimming is a function of the load balance.
Last edited:
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
160902-0743 EDT
The Fluke 87 in AC mode is really an averaging device that measures the waveform RMS value averaged over some time constant that is longer than 1/2 cycle. The selected averaging time is possibly based on a 60 Hz signal.
Inherrently an RMS measurrement implies an average. RMS is the instantaneous value squared, then the mean (average) of the instantaneous squared values (this is ideally over an integral number of 1/2 cycles to be meaningful, but a number of 1/2 cycles like 15 and not synced with zero crossings will work fairly well), and last the square root of the average results in the RMS value.
I have to experiment with thre 87 in its short time MIN/MAX, but I believe it will give me the maximum instantaneous peak, and the minimum (possibly meaning negative) instantaneous peak.
Positive and negative peaks are not much value in solving a lamp dimming problem. However, these measurements might be useful in looking at peak voltage with and without a transient voltage limiter.
.
The Fluke 87 in AC mode is really an averaging device that measures the waveform RMS value averaged over some time constant that is longer than 1/2 cycle. The selected averaging time is possibly based on a 60 Hz signal.
Inherrently an RMS measurrement implies an average. RMS is the instantaneous value squared, then the mean (average) of the instantaneous squared values (this is ideally over an integral number of 1/2 cycles to be meaningful, but a number of 1/2 cycles like 15 and not synced with zero crossings will work fairly well), and last the square root of the average results in the RMS value.
I have to experiment with thre 87 in its short time MIN/MAX, but I believe it will give me the maximum instantaneous peak, and the minimum (possibly meaning negative) instantaneous peak.
Positive and negative peaks are not much value in solving a lamp dimming problem. However, these measurements might be useful in looking at peak voltage with and without a transient voltage limiter.
.
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
160902-1225 EDT
I did a quick test using a #2 AC 120 V coil motor starter as a transient generator, and a Flike 87 meter on MIN/MAX 1 millisecond. On the 1000 V range it read overload as a result of random cycling of the motor starter. This is to be expected.
.
I did a quick test using a #2 AC 120 V coil motor starter as a transient generator, and a Flike 87 meter on MIN/MAX 1 millisecond. On the 1000 V range it read overload as a result of random cycling of the motor starter. This is to be expected.
.
JDB3
Senior Member
- Location
- San Antonio, Texas
If lights are dimming, varying in intensity, or flickering, and the voltage on the circuit to the lights is not varying in correlation with the intensity varaition, then nothing that preceeds the point of voltage measurement has anything to do with the intensity variation problem. Don't bother the power company. I am assuming that there is not a 500,000 W radio transmitter next door.
While the general contractor, another master electrician with experience with sales,installation, trouble shooting, & I were there, we just saw dimming at the one ceiling fan light kit, and like I stated, after the GC changed the bulbs to LED, it stopped!
The trouble started after the owners were occupying the house for a couple of months (so they said). Even the ceiling fan light kit, that they said did it on a constant basis (lights dimming), did not dim while the three of us were there.
This house is feed by its own individual pad mount transformer.
This is why we want the utility company to monitor things on their side.
While the general contractor, another master electrician with experience with sales,installation, trouble shooting, & I were there, we just saw dimming at the one ceiling fan light kit, and like I stated, after the GC changed the bulbs to LED, it stopped!
The trouble started after the owners were occupying the house for a couple of months (so they said). Even the ceiling fan light kit, that they said did it on a constant basis (lights dimming), did not dim while the three of us were there.
This house is feed by its own individual pad mount transformer.
This is why we want the utility company to monitor things on their side.
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
160903-1151 EDT
JDB3:
I don't believe that you ever explicitly stated that the problem fan light originally had incandescents installed. Your last post, #38, now states that some kind of bulbs have been replaced with LEDs. We have to assume the originals were incandescent. You saw dimming with the original bulbs. This is the only fixture where you have personally seen dimming, and now with LEDs installed you do not see dimming. Apparently dimming at this fan-light fixture occurred fairly consistently.
We need a definition of what dimming means. Does it ever mean brightning? Are durations very short, long, or quite random? Apparently long enough for you to take lightmeter readings. That is relatively long.
You indicated in your first post that dimming at the fan-light was a change from 17 to 12 foot-candles, or about a 30% drop, 100% to 70.6%. From a test I have run in the past for a standard incandescent bulb this change corresponds to a voltage drop from 120 V to about 102 V. This is a huge drop. In a later post you indicated no voltage change on the branch circuit fairly close to the light when dimming occurred. Thus, not a power company problem, nor a neutral or hot line problem prior to the point of the voltage measurement.
We do not know what caused the dimming problem in the light fixture because you have not investigated what is in the fixture. Having replaced the original bulbs (unknown type) with LEDs has apparently eliminated the problem. A particular sample of a new CREE 9.4 W visually shows virtually no intensity change from 120 to 102 V. At the moment I don't plan to run a more controlled experiment on intensity variation for a change in sine wave voltage.
Your customer has claimed other lights in the home have problems in that they dim at times for no reasion. You have not observed this problem with other lights. At this point in time I don't think you have sufficient reason to ask the power company to monitor the input supply, that costs them money, nor would I trust them to identify or correctly idemtify a problem. You need evidence that points to a power company problem before you ask them for monitoring.
You should make some tests at the main panel with some known significant load(s).
At my home today I have a 50 kVA pole transformer located at the street. This is a single phase system with a center tapped secondary with a ground rod at the pole. The transformer primary is supplied by a delta source. Thus, there is no daisy chained ground wire from one pole to another. I do not presently have a wire from that pole ground rod to use as voltmeter probe.
From the present pole transformer location to the pole where my original transformer was located is a wimpy service cable. Length about 90 ft. From the original pole transformer location to my main panel I have 50 to 70 ft of 0000 copper.
The following measurements at my main panel are from this morning. The measurements imply that the service wiring is the major impedance, and the transformer is only a small factor. I used a 16.2 A 120 V test load consisting of a 1500 W photoflood and 1000 W single unit kitchen heating element in parallel. I used two voltmeters, but one is sufficient. I used a mercury relay to switch the load on and off. This relay could handle a much larger load current. Tests were run by separately loading both phases. Note that meter resolution is 0.1 V, and that I don't have large voltage changes with this amount of loading. 100 A would be a better test.
We should expect that each of the service wires has about the same resistance.
As an approximation I assume the transformer impedance is small compared to the service wires.
Measurement results:
My results from loading phase B follow:
Loading phase B with 16.2 A:
Phase A changed from 124.0 to 124.6 V when phase B was loaded with 16.2 A. A positive change of, +0.6 V. Change direction is expected.
Phase B changed from 123.5 to 122.4 V when phase B was loaded with 16.2 A. A negative change of, -1.1 V. Change direction expected.
What does this mean? When phase A has no change in load, then it produces no change in the neutral current. But when phase A has no current change, and phase B has a resistive increase in load current then there is change in neutral current that produces a change in the neutral voltage of such a phase that it is additive to the original phase A to neutral voltage. Thus, phase A to neutral increases.
If we assume that the impedance of each hot conductor and the neutral are the same, and the change in voltage drop along each conductor is proportional to the change in current thru the conductor, then the change in voltage drop along the hot and the neutral will be the same.
The unloaded phase A's conductor is essentially a voltmeter probe to the transformer. So the voltage change across the neutral (transformer center tap to main panel neutral bus) due to the phase B load is the same as the voltage change seen at the main panel from phase A to the neutral at the main panel. Since the change in voltage across the neutral ia about 1/2 the change in voltage of phase B's load change, and is small (in an expected range), then we can assume that the neutral and phase B hot line are in good condition. Reversing the measurement will test the phase A hot line.
My results from loading phase A follow:
Loading phase A with 16.2 A:
Phase B changed from 123.5 to 124.1 V when phase A was loaded with 16.2 A. A positive change of, +0.6 V. Change direction is expected.
Phase A changed from 124.2 to 123.1 V when phase A was loaded with 16.2 A. A negative change of, -1.1 V. Change direction expected.
To get an idea of whether or not the power company has a primary voltage variation problem, then the main panel voltage needs to be monitored. To do this I would use two TED 1000 power monitors that also record voltage.
The TED 1000 model actually has 1 second resolution. Later units are really 2 to 5 seconds. See a 24 hour plot at my website with 1 second resolution. http://beta-a2.com/EE-photos.html
The photo at P 26 is measured at the main panel. The photo at P 29 is measured at the socket the freezer is plugged into, and is only of the freeze. The freezer is more than 100 ft of wire from the main panel, but 90 ft or so is #6. The remainder is #12 and there are two series breakers. The voltage drop spikes result from the long line from the panel. Since motor inrush current does not last 1 second, but more likely about 1 or 2 tenths of a second we can expect the actual voltage drops to be greater than shown.
.
JDB3:
I don't believe that you ever explicitly stated that the problem fan light originally had incandescents installed. Your last post, #38, now states that some kind of bulbs have been replaced with LEDs. We have to assume the originals were incandescent. You saw dimming with the original bulbs. This is the only fixture where you have personally seen dimming, and now with LEDs installed you do not see dimming. Apparently dimming at this fan-light fixture occurred fairly consistently.
We need a definition of what dimming means. Does it ever mean brightning? Are durations very short, long, or quite random? Apparently long enough for you to take lightmeter readings. That is relatively long.
You indicated in your first post that dimming at the fan-light was a change from 17 to 12 foot-candles, or about a 30% drop, 100% to 70.6%. From a test I have run in the past for a standard incandescent bulb this change corresponds to a voltage drop from 120 V to about 102 V. This is a huge drop. In a later post you indicated no voltage change on the branch circuit fairly close to the light when dimming occurred. Thus, not a power company problem, nor a neutral or hot line problem prior to the point of the voltage measurement.
We do not know what caused the dimming problem in the light fixture because you have not investigated what is in the fixture. Having replaced the original bulbs (unknown type) with LEDs has apparently eliminated the problem. A particular sample of a new CREE 9.4 W visually shows virtually no intensity change from 120 to 102 V. At the moment I don't plan to run a more controlled experiment on intensity variation for a change in sine wave voltage.
Your customer has claimed other lights in the home have problems in that they dim at times for no reasion. You have not observed this problem with other lights. At this point in time I don't think you have sufficient reason to ask the power company to monitor the input supply, that costs them money, nor would I trust them to identify or correctly idemtify a problem. You need evidence that points to a power company problem before you ask them for monitoring.
You should make some tests at the main panel with some known significant load(s).
At my home today I have a 50 kVA pole transformer located at the street. This is a single phase system with a center tapped secondary with a ground rod at the pole. The transformer primary is supplied by a delta source. Thus, there is no daisy chained ground wire from one pole to another. I do not presently have a wire from that pole ground rod to use as voltmeter probe.
From the present pole transformer location to the pole where my original transformer was located is a wimpy service cable. Length about 90 ft. From the original pole transformer location to my main panel I have 50 to 70 ft of 0000 copper.
The following measurements at my main panel are from this morning. The measurements imply that the service wiring is the major impedance, and the transformer is only a small factor. I used a 16.2 A 120 V test load consisting of a 1500 W photoflood and 1000 W single unit kitchen heating element in parallel. I used two voltmeters, but one is sufficient. I used a mercury relay to switch the load on and off. This relay could handle a much larger load current. Tests were run by separately loading both phases. Note that meter resolution is 0.1 V, and that I don't have large voltage changes with this amount of loading. 100 A would be a better test.
We should expect that each of the service wires has about the same resistance.
As an approximation I assume the transformer impedance is small compared to the service wires.
Measurement results:
My results from loading phase B follow:
Loading phase B with 16.2 A:
Phase A changed from 124.0 to 124.6 V when phase B was loaded with 16.2 A. A positive change of, +0.6 V. Change direction is expected.
Phase B changed from 123.5 to 122.4 V when phase B was loaded with 16.2 A. A negative change of, -1.1 V. Change direction expected.
What does this mean? When phase A has no change in load, then it produces no change in the neutral current. But when phase A has no current change, and phase B has a resistive increase in load current then there is change in neutral current that produces a change in the neutral voltage of such a phase that it is additive to the original phase A to neutral voltage. Thus, phase A to neutral increases.
If we assume that the impedance of each hot conductor and the neutral are the same, and the change in voltage drop along each conductor is proportional to the change in current thru the conductor, then the change in voltage drop along the hot and the neutral will be the same.
The unloaded phase A's conductor is essentially a voltmeter probe to the transformer. So the voltage change across the neutral (transformer center tap to main panel neutral bus) due to the phase B load is the same as the voltage change seen at the main panel from phase A to the neutral at the main panel. Since the change in voltage across the neutral ia about 1/2 the change in voltage of phase B's load change, and is small (in an expected range), then we can assume that the neutral and phase B hot line are in good condition. Reversing the measurement will test the phase A hot line.
My results from loading phase A follow:
Loading phase A with 16.2 A:
Phase B changed from 123.5 to 124.1 V when phase A was loaded with 16.2 A. A positive change of, +0.6 V. Change direction is expected.
Phase A changed from 124.2 to 123.1 V when phase A was loaded with 16.2 A. A negative change of, -1.1 V. Change direction expected.
To get an idea of whether or not the power company has a primary voltage variation problem, then the main panel voltage needs to be monitored. To do this I would use two TED 1000 power monitors that also record voltage.
The TED 1000 model actually has 1 second resolution. Later units are really 2 to 5 seconds. See a 24 hour plot at my website with 1 second resolution. http://beta-a2.com/EE-photos.html
The photo at P 26 is measured at the main panel. The photo at P 29 is measured at the socket the freezer is plugged into, and is only of the freeze. The freezer is more than 100 ft of wire from the main panel, but 90 ft or so is #6. The remainder is #12 and there are two series breakers. The voltage drop spikes result from the long line from the panel. Since motor inrush current does not last 1 second, but more likely about 1 or 2 tenths of a second we can expect the actual voltage drops to be greater than shown.
.
Knuckle Dragger
Master Electrician Electrical Contractor 01752
- Location
- Marlborough, Massachusetts USA
- Occupation
- Electrical Contractor
All great advice you have recived.
9 times out of 10 when there is fluctuating throughout a residential home it is a neutral or hot leg issue.
I would insist on the local POCO check the power out in the street and at the meter if you are not allowed to touch it.
9 times out of 10 when there is fluctuating throughout a residential home it is a neutral or hot leg issue.
I would insist on the local POCO check the power out in the street and at the meter if you are not allowed to touch it.
- Status