Eiko t8 led typeB

Status
Not open for further replies.

Mparn

Member
Location
Missouri
Any idea why a random lamp or lamps would pulse when another series of lights are turned On? The lights in our building is on a multi branch circuit, I'm guessing it has something to do with the shared neutral.
 

ActionDave

Chief Moderator
Staff member
Location
Durango, CO, 10 h 20 min from the winged horses.
Occupation
Licensed Electrician
Any idea why a random lamp or lamps would pulse when another series of lights are turned On? The lights in our building is on a multi branch circuit, I'm guessing it has something to do with the shared neutral.

Whatever the problem is it's not because the lights are on a multiwire branch circuit. Stop focusing on some superstitious fear that isn't based in reality.

The whole North American Continent is fed with power that shares a neutral. There may be some bad wiring causing problems, but electrons are too dumb to know what circuit conductor they are traveling around on, and LED lamps don't care what kind of circuit they wired to.

I think it's far more likely that cheap LED lamps are the problem.
 

Mparn

Member
Location
Missouri
Whatever the problem is it's not because the lights are on a multiwire branch circuit. Stop focusing on some superstitious fear that isn't based in reality.

The whole North American Continent is fed with power that shares a neutral. There may be some bad wiring causing problems, but electrons are too dumb to know what circuit conductor they are traveling around on, and LED lamps don't care what kind of circuit they wired to.

I think it's far more likely that cheap LED lamps are the problem.


Thanks for the reply Dave. I'm getting this random flicker on several rooms. I know it could be a possible wiring issue, but I can't believe that many are wired bad. I have personally wired a few offices by myself and feel they are all correct. I have also tried a Phillips 17w type B with the same flickering results.

Once all lights are on I have zero problems. The only time I have a problem is when a set is turned on, an existing set will have a random flicker on 1 or a few lamps. I can't say it's always the same lamp and sometimes it doesn't happen at all.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
180423-1029 EDT

Mparn:

You clearly have a single shot transient problem that only occurs at some turn on time. To light an LED for a moment it takes some amount of energy.

You have lights that are turned on and off by some sort of switch. What is the actual switching device between power (the hot wire) and the actual lights that flash when some other circuit is turned on? Is it an ordinary mechanical wall switch, mechanical contacts on a relay, or some sort of electronic switch?

How long are wire runs where hot and neutral wires are close together? A 4 ft long LED bulb will take a moderate amount of energy for a momentary flash. Could a 1000 pfd (0.001 ufd) capacitor provide enough energy?

An LED driver probably has a fairly large input capacitor with little current limiting in series. After being off this capacitor has near zero stored charge. If turn on to this capacitor occurs at an AC voltage peak, then you can expect a large input current spike at this turn on.

How could such a large current spike couple to what you think is an unconnected LED? Also there are likely large voltage spikes.

.
 

Mparn

Member
Location
Missouri
180423-1029 EDT

Mparn:

You clearly have a single shot transient problem that only occurs at some turn on time. To light an LED for a moment it takes some amount of energy.




You have lights that are turned on and off by some sort of switch. What is the actual switching device between power (the hot wire) and the actual lights that flash when some other circuit is turned on? Is it an ordinary mechanical wall switch, mechanical contacts on a relay, or some sort of electronic switch?

How long are wire runs where hot and neutral wires are close together? A 4 ft long LED bulb will take a moderate amount of energy for a momentary flash. Could a 1000 pfd (0.001 ufd) capacitor provide enough energy?

An LED driver probably has a fairly large input capacitor with little current limiting in series. After being off this capacitor has near zero stored charge. If turn on to this capacitor occurs at an AC voltage peak, then you can expect a large input current spike at this turn on.

How could such a large current spike couple to what you think is an unconnected LED? Also there are likely large voltage spikes.

.

The switch is an ordinary wall switch. Last week I turned all the lights on that run off the same panel and fired a few compressors from the RTUs. The lights didn't flicker or react in any way. I have plenty of slack between wires feeding the fixture and tombstones. Could that be a problem, too much wire?
If there are large spikes, how would I prevent this?
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
180423-1230 EDT

Mparn:

(1) How many different LED fixtures in one room?
(2) How many different switched LED circuits in one room?
(3) How many switches in one box?
(4) How many different switching locations are in one room?
(5) How big is the room?
(6) Are all the lights in one room on the same phase?
(7) Are all the LEDs controlled by one switch close together in the room or scattered amongst LEDs on other switches?
(8) If you toggle on and off one circuit of LEDs what percentage of time does an LED flash in another circuit? Is it always in the same other circuit?

What does a tombstone have to do with the problem? I think that you can consider a single fixture when disconnected from its supply wires at the fixture to have almost zero likelihood of flashing from adjacent wires.

.
 

Mparn

Member
Location
Missouri
180423-1230 EDT

Mparn:

(1) How many different LED fixtures in one room?
(2) How many different switched LED circuits in one room?
(3) How many switches in one box?
(4) How many different switching locations are in one room?
(5) How big is the room?
(6) Are all the lights in one room on the same phase?
(7) Are all the LEDs controlled by one switch close together in the room or scattered amongst LEDs on other switches?
(8) If you toggle on and off one circuit of LEDs what percentage of time does an LED flash in another circuit? Is it always in the same other circuit?

What does a tombstone have to do with the problem? I think that you can consider a single fixture when disconnected from its supply wires at the fixture to have almost zero likelihood of flashing from adjacent wires.

.

12 fixtures per room, each fixture has 2 lamps.
2 circuits per room, most rooms are divided in half, some thirds
2 single pole switches per box.
Rooms are 30'x30'
They share the same circuit the ones I turn on and the ones that Flickr
They flicker immediately as the lights are powered.
I don't have any dimmers or 3ways.

Tombstones have nothing to do with it, I was just stating my 18 gauge in the fixtures may be fairly long with hot and neutral possibly close together.
 
Last edited:

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
180423-1301 EDT

Mparn:

To get electrical energy to a load you need some sort of closed electrical circuit (excluding radiated radio frequency energy, I think we are only concerned with conductive, capacitive, and inductive components describing your circuit).

Assuming your switches are really open when off, then to get energy to the flickered fixture means there has to be some form of capacitive or inductive coupling to the flickered circuit.

A very remote possibility is a very large transient voltage at the switch box that causes electrical breakdown of the open switch. I view this as a zero likelihood.

I don't believe there is inductive coupling.

Very likely you have capacitive coupling. You might have 20 pfd per foot between some wires. At most it is unlikely you have more than 60 ft of two hot wires from different circuits this closely coupled. Suppose you did, then the capacitance is 1200 pfd. Can I cause a flicker in a fixture with this amount of coupling? Depends. With sufficient voltage it might be possible.

I don't know your voltage. But get a 0.1 ufd capacitor or somewhat larger of sufficient voltage rating for your circuit voltage. Connect this capacitor between neutral and the hot wire to the fixtures that flicker on the output side of the switch that is off. The connection might be made at the switch box or anywhere near a fixture. See if this eliminates or in any way changes the flicker. This creates a capacitive voltage divider that should greatly reduce the source of voltage to a fixture, if capacitive coupling is the problem.

If capacitive coupling is yor problem, then you need to change the way wiring is run.

.
 

Mparn

Member
Location
Missouri
180423-1301 EDT

Mparn:

To get electrical energy to a load you need some sort of closed electrical circuit (excluding radiated radio frequency energy, I think we are only concerned with conductive, capacitive, and inductive components describing your circuit).

Assuming your switches are really open when off, then to get energy to the flickered fixture means there has to be some form of capacitive or inductive coupling to the flickered circuit.

A very remote possibility is a very large transient voltage at the switch box that causes electrical breakdown of the open switch. I view this as a zero likelihood.

I don't believe there is inductive coupling.

Very likely you have capacitive coupling. You might have 20 pfd per foot between some wires. At most it is unlikely you have more than 60 ft of two hot wires from different circuits this closely coupled. Suppose you did, then the capacitance is 1200 pfd. Can I cause a flicker in a fixture with this amount of coupling? Depends. With sufficient voltage it might be possible.

I don't know your voltage. But get a 0.1 ufd capacitor or somewhat larger of sufficient voltage rating for your circuit voltage. Connect this capacitor between neutral and the hot wire to the fixtures that flicker on the output side of the switch that is off. The connection might be made at the switch box or anywhere near a fixture. See if this eliminates or in any way changes the flicker. This creates a capacitive voltage divider that should greatly reduce the source of voltage to a fixture, if capacitive coupling is the problem.

If capacitive coupling is yor problem, then you need to change the way wiring is run.

.

Thanks, I there a reason why this problem didn't occur with the ballast and t12? Will this capacitive coupling cause any future problems, other then the random flicker when lights are turned on?
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
180423-1627 EDT

Mparn:

The probable reason the T12s never flashed was need for greater energy, and possibly less coupling of transient energy.

.
 

Russs57

Senior Member
Location
Miami, Florida, USA
Occupation
Maintenance Engineer
I suggest you look at IEEE 1789-2015 paper.

I suspect your problems have nothing to do with wiring, loads, or shared neutrals. Might prove interesting to see how many others notice this problem. Some are much more sensitive. I am....probably borderline epileptic!

This isn’t an unknown problem and why I advised trusted name. Also mentioned a driver would do better. Problem is most won’t work on 277.

Really hoping problem is some simple/stupid wiring issue. But do be aware some are suggesting there is a health risk. That is not a place you want to be with children involved.

Not trying to worry you. Just saying it might be time to slow down and evaluate things.
 

Mparn

Member
Location
Missouri
Gotcha thanks, I have tried 2 different brands (one being phillips) with the same results. This is a one time flicker on a lamp or few lamps when others are powered on. We've been monitoring the lamps daily for any flickering or strobing. No one seems to find evidence of either.
 
Last edited:

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
180423-1946 EDT

Russs57:

The
IEEE 1789-2015 has nothing to do with the problem that Mparn was questioning. Mparn described a single shot transient event. IEEE 1789-2015 is concerned with a continuous steady state modulation of light intensity. Should this be of concern? Yes. But it is not what the original post was about.

.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
180423-1958 EDT

Mparn:

If the single shot flicker does not bother anyone, then I would do nothing about it. I doubt that it is a resistive leakage problem.

The IEEE 1789-2015 brings up a much bigger and different concern. LEDs designed to provide some approximation of visible white or other color light generally are a blue or UV LED used to excite a phosphor(s) in the visible spectrum. These phosphors generally have moderate time constants of intensity decay. Thus, a square wave drive of the LED emitter that produces a square wave of blue or UV light may have a much more averaged visible light output. The time constants of the different colors from the LED phosphors may not and probably are not the same. Thus, the percentage of flicker from each color may differ.

For other reasons it would be good to drive LED chips from DC.

.

.
 
Last edited:

Mparn

Member
Location
Missouri
I noticed a weak switch today in one of the classrooms. If i held the switch in the middle, the other series would get more of a random flicker. Could it be several of the switches are going bad? It even happened if I would just push in on the switch
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
190426-2032 EDT

Mparn:

With inductance in a circuit and current flowing the opening of the circuit can easily produce thousands of volts to initiate a spark or arc across the opening element. This how an old automotive ignition coil works.

To test the sensitivity of electronic circuits to transient noise I have simply rapidly oscillated a plug to a two bulb 8 ft Slimline fixture.

Will arcing of a set of switch contacts in some fashion couple enough energy into a turned off LED circuit to momentarily flicker the LED? I don't know. But you may have demonstrated that it does.

.
 

Mparn

Member
Location
Missouri
190426-2032 EDT

Mparn:

With inductance in a circuit and current flowing the opening of the circuit can easily produce thousands of volts to initiate a spark or arc across the opening element. This how an old automotive ignition coil works.

To test the sensitivity of electronic circuits to transient noise I have simply rapidly oscillated a plug to a two bulb 8 ft Slimline fixture.

Will arcing of a set of switch contacts in some fashion couple enough energy into a turned off LED circuit to momentarily flicker the LED? I don't know. But you may have demonstrated that it does.

.
Thanks, the flicker happens in the circuit that is already on, when the other is turned on.
 

gar

Senior Member
Location
Ann Arbor, Michigan
Occupation
EE
180427-1005 EDT

Mparn:

At this point you need to define some words and run some independent experiments.

Flicker could means a number of different things. Is it a single shot event? Does it means 120 Hz ripple on light intensity? Not in your case, but it did in Russs57's comment. Does it mean an increase in intensity, or a decrease? Is it a short duration oscillation in intensity? Or etc.

Consider two 120 V switched 100 W bulbs fed from the same phase and sharing a common neutral of 10 ohms. A cold 100 W 120 V incandescent bulb is about 10 ohms, and hot about 144 ohms. With one bulb on its voltage is about 120*144/154 = 112 V. You would see a brightness difference in the bulb compared to a circuit with a much lower neutral resistance.

I have selected the bulb size and neutral resistance such that you can easily experiment. 100 W incandescent bulbs are still around, and a moderately stable resistance at 1 or 2 A and somewhat near 10 ohms is the resistance of a 1500 W space heater.

At the instant the second bulb turns on the voltage across the two bulbs is about 120*10/10 = 60 V. Actually a little less because you have 144 ohms shunting 10 ohms, or 9.35 ohms. So you will see a short time flicker in the dimming direction of the initially on bulb.

Now suppose the two bulbs are supplied from opposite phases. The 10 ohm neutral initially has a voltage drop of 120*10/154 = 7.8 V with one bulb on. At the instant that the second bulb is turned on two opposing currents exist in the neutral. Doing the math just after the second bulb is turned on I get a voltage drop across the 10 ohm resistor of about 54 V. This is in a direction to add voltage to the initially on bulb so that the already on bulb sees about 120 + 54 = 174 V. Thus, the already on bulb brightness momentarily increases a lot, bright flash.

When the turning on bulb reaches its high resistance the 10 ohm resistance has a voltage drop of about 0 volts, canceling currents thru the neutral, and the initially on bulb is slightly brighter than when it was the only load. You may not detect this.

The above is a suggested experiment that I have not played with but may illustrate how neutral impedance can affect components in a circuit.

LEDs have ballasts (power supplies) that may be pretty much like an uncharged capacitor at their input before turn on. Connecting an electronically ballasted LED at a voltage peak and large peak currents may flow.

.
 

Mparn

Member
Location
Missouri
180427-1005 EDT

Mparn:

At this point you need to define some words and run some independent experiments.

Flicker could means a number of different things. Is it a single shot event? Does it means 120 Hz ripple on light intensity? Not in your case, but it did in Russs57's comment. Does it mean an increase in intensity, or a decrease? Is it a short duration oscillation in intensity? Or etc.

Consider two 120 V switched 100 W bulbs fed from the same phase and sharing a common neutral of 10 ohms. A cold 100 W 120 V incandescent bulb is about 10 ohms, and hot about 144 ohms. With one bulb on its voltage is about 120*144/154 = 112 V. You would see a brightness difference in the bulb compared to a circuit with a much lower neutral resistance.

I have selected the bulb size and neutral resistance such that you can easily experiment. 100 W incandescent bulbs are still around, and a moderately stable resistance at 1 or 2 A and somewhat near 10 ohms is the resistance of a 1500 W space heater.

At the instant the second bulb turns on the voltage across the two bulbs is about 120*10/10 = 60 V. Actually a little less because you have 144 ohms shunting 10 ohms, or 9.35 ohms. So you will see a short time flicker in the dimming direction of the initially on bulb.

Now suppose the two bulbs are supplied from opposite phases. The 10 ohm neutral initially has a voltage drop of 120*10/154 = 7.8 V with one bulb on. At the instant that the second bulb is turned on two opposing currents exist in the neutral. Doing the math just after the second bulb is turned on I get a voltage drop across the 10 ohm resistor of about 54 V. This is in a direction to add voltage to the initially on bulb so that the already on bulb sees about 120 + 54 = 174 V. Thus, the already on bulb brightness momentarily increases a lot, bright flash.

When the turning on bulb reaches its high resistance the 10 ohm resistance has a voltage drop of about 0 volts, canceling currents thru the neutral, and the initially on bulb is slightly brighter than when it was the only load. You may not detect this.

The above is a suggested experiment that I have not played with but may illustrate how neutral impedance can affect components in a circuit.

LEDs have ballasts (power supplies) that may be pretty much like an uncharged capacitor at their input before turn on. Connecting an electronically ballasted LED at a voltage peak and large peak currents may flow.

Gar,

Thank you for all your help. Although you are way over my head, you have explained it to where I am able to understand. By "flicker" I mean, a single shot, one time, where the light dims or goes out. It happens really fast but you are still able to see it. You have to actually study it to see that it's dimming and not getting brighter.
 
Last edited by a moderator:
Status
Not open for further replies.
Top