led floodlights flashing

[Solanaceae]

Are the LEDs with simple capacitative ballasts or more complex drivers? This could also be a mode of driver failure. I saw led track lighting at a place in Florida flashing and striving wildly, turns out a craptastic driver failed.

 

[LIGO-ELEC]

LED Lights not on Dimmer, Still Flash...

LED Lights not on Dimmer, Still Flash...

In the Entrance Hall of our Facility, there are four rows of PAR38 LED's floodlights, 32 in all, none of which are on Dimmers - only regular switches. I've had six of the LED bulbs, randomly located through the four rows, flash upon first turning the lights on in the morning, . They will eventually stop flashing after about less than a minute from when turned on, and not do so again until after they have been turned off for a while (gotten cooled off?). I've gone back to the Manufacturer and received the same response about Dimmers, but no explanation as to why our PAR38 LEDs would flash even though they are not on Dimmers. As a test, I replaced one of the four rows with another Manufacturer's PAR38 LED Floodlights and have had no issue for the past six weeks of their operation. I have since taken the conservative route and am replacing all of the remaining LED bulbs with another manufacturer's LED bulbs hoping that this would mitigate this nuisance flashing (Gotta keep our receptionist happy!).

In my opinion, even though Dimmer compatibility may be a more prevalent root cause of LED Bulbs flashing, it is not the only cause. There is something else that is being missed.

 

[mgookin]

In the Entrance Hall of our Facility, there are four rows of PAR38 LED's floodlights, 32 in all, none of which are on Dimmers - only regular switches. I've had six of the LED bulbs, randomly located through the four rows, flash upon first turning the lights on in the morning, . They will eventually stop flashing after about less than a minute from when turned on, and not do so again until after they have been turned off for a while (gotten cooled off?). I've gone back to the Manufacturer and received the same response about Dimmers, but no explanation as to why our PAR38 LEDs would flash even though they are not on Dimmers. As a test, I replaced one of the four rows with another Manufacturer's PAR38 LED Floodlights and have had no issue for the past six weeks of their operation. I have since taken the conservative route and am replacing all of the remaining LED bulbs with another manufacturer's LED bulbs hoping that this would mitigate this nuisance flashing (Gotta keep our receptionist happy!).

In my opinion, even though Dimmer compatibility may be a more prevalent root cause of LED Bulbs flashing, it is not the only cause. There is something else that is being missed.

Welcome to the forum.

It sounds like you have an issue resulting from the drivers getting up to temperature. Something flaky is going on.

Some people have fixed some LED problems by putting a resistive load on the circuit (incandescent bulb).

ETA: I just looked at your profile. Sounds like quite a job you have there. Do you get to play with the big stuff or are you more of a facilities manager?

 

[LIGO-ELEC]

Welcome to the forum.

It sounds like you have an issue resulting from the drivers getting up to temperature. Something flaky is going on.

Some people have fixed some LED problems by putting a resistive load on the circuit (incandescent bulb).

ETA: I just looked at your profile. Sounds like quite a job you have there. Do you get to play with the big stuff or are you more of a facilities manager?

The Incandescent Light Bulb solution is intriguing...! I'm going to have to try that. Thanks!

I actually am the Electrical Engineer for the Site. However, I steer away from working on the Lasers and Interferometer equipment. I jokingly state that I deal with the Big Electrons and let the Electronics Engineers deal with the Little Electrons...

It is almost like reporting to the Engineering Section of the Starship Enterprise - on a daily basis. I keep expecting to run into Mr. Spock turning a corner.

 

[gar]

160908-2010 EDT


LIGO-ELEC:

I would take all the bulbs that malfunction and run them individually on the bench. Use a Variac to supply different test voltages, and some means to adjust the bulb initial temperature. Liquid CO2 thru an appropriate nozzle is a good means. Must use a siphon tank, or turn a non-siphon upside down. These tanks are about a 110# empty, and pressure is toward 1000 PSI. You can also use dry ice in a cardboard box, or an ordinary refrigerator. Liquid CO2 thru a nozzle is the quickest means for cooling.

If voltage and temperature changes don't cause a single bulb on the bench to malfunction, then it is necessary to try other circuit variations to try to simulate the cause.

Name the manufacturers for the working and problem bulbs. We need this information to judge what manufacturers are producing good product.

.

 

[LIGO-ELEC]

160908-2010 EDT


LIGO-ELEC:

Name the manufacturers for the working and problem bulbs. We need this information to judge what manufacturers are producing good product.

.

The PAR38's that are malfunctioning are Phillips EnduraLED 20° Flood, 16W, 4200K, Dimmable, Cat #40817-9. I am replacing these with EIKO [FONT=&quot]LED17WPAR38/NFL/840K-DIM-G4A [/FONT]

I wasn't sure if it was Ok to name the manufacturer, so I avoided that in my initial post.

 

[gar]

160909-1059 EDT

LIGO-ELEC:

It is my own opinion that it is important to identify manufacturers so others can know what to look out for.

With all the new kinds of devices being created to save energy or provide other features it has become a jungle.

I have played with CREE bulbs, I may have one Phillips, and I have several Feit bulbs.

My son won't buy a CREE because he had too many failures. I have had one CREE fail due to an internal thermal problem. This bulb will run for a number of minutes, then quit, and after a cool down time automatically lights up again. A different CREE bulb that I have on about 16 hours per day for the last two years was sputtering the night before last. This lasted for about 15 minutes, corrected itself, and has been OK since then. I have a number of CREE bulbs with no problems. The CREE bulbs I have tested have good dimming characteristics with both an adjustable sine wave, and with phase shift dimming. When a CREE is operated directly off the line, no dimmer, the RFI (radio frequency interference) is not bad.

My son has had no problem with Feit bulbs. However, I have tested some Feit bulbs that put out a strong RFI signal around 800 kHz, middle of the AM broadcast band.

With any bulb (incandescent, CFL, or LED) that is dimmed with a phase shift dimmer will produce noticeable RFI. Some, possibly many, CFLs, and LEDs, don't dim very well with any means of voltage control.

.

 

[LIGO-ELEC]

160909-1059 EDT

LIGO-ELEC:

It is my own opinion that it is important to identify manufacturers so others can know what to look out for.

With all the new kinds of devices being created to save energy or provide other features it has become a jungle.

I have played with CREE bulbs, I may have one Phillips, and I have several Feit bulbs.

My son won't buy a CREE because he had too many failures. I have had one CREE fail due to an internal thermal problem. This bulb will run for a number of minutes, then quit, and after a cool down time automatically lights up again. A different CREE bulb that I have on about 16 hours per day for the last two years was sputtering the night before last. This lasted for about 15 minutes, corrected itself, and has been OK since then. I have a number of CREE bulbs with no problems. The CREE bulbs I have tested have good dimming characteristics with both an adjustable sine wave, and with phase shift dimming. When a CREE is operated directly off the line, no dimmer, the RFI (radio frequency interference) is not bad.

My son has had no problem with Feit bulbs. However, I have tested some Feit bulbs that put out a strong RFI signal around 800 kHz, middle of the AM broadcast band.

With any bulb (incandescent, CFL, or LED) that is dimmed with a phase shift dimmer will produce noticeable RFI. Some, possibly many, CFLs, and LEDs, don't dim very well with any means of voltage control.

.

Agreed - At LIGO, we are extremely sensitive to RFI due to our mega-super-sensitive instruments. We actually have to turn off all light fixtures in the LASER/Interferometer areas of the site (as well as all other non-essential Electrical and Electronic equipment and devices) when we are in observation mode in order to eliminate their electronic noise from coupling into our measurements. At this time, I am also testing out some ECOLITE LED replacements to a few 4' Flourescents at one of the Laser/Interferometer labs on site to see what the effects may be. I'll post any findings on here as things progress.

I do appreciate all the input from everyone on this. It is an annoying issue for sure.

Paul

 

[gar]

160909-1545 EDT

LIGO-ELEC:

The simplest RFI solution is to directly use incandescent bulbs off of the AC circuit, no phase shift dimmers.

Another approach is to get special LED bulbs without the supplied power supply (some would call that driver a ballast). An LED is basically a DC device. It is a semiconductor diode that emits light only when current flows in one direction, the forward conduction direction.

Then you supply a DC current source to the LED consisting of a line frequency transformer, a diode rectifier, and a linear current controller (the simplest is a resistor). To a large extent an LED can be considered a constant voltage load and thus must be fed from an approximately constant current source. This system can be virtually RFI free.

.

 

[Electric-Light]

LED systems can be a notorious RFI emitter. So bad that they've been known to cause interference that have to FCC knocking on the door.

Those that use sophisticated electronic control to set calibration for color temperature to compensate for their inherently high manufacturing variance are particularly bad.

 

[peter d]

LED systems can be a notorious RFI emitter. So bad that they've been known to cause interference that have to FCC knocking on the door.

Those that use sophisticated electronic control to set calibration for color temperature to compensate for their inherently high manufacturing variance are particularly bad.

And despite that, they are rapidly replacing obsolete lighting systems. :thumbsup:

 

[K8MHZ]

LED systems can be a notorious RFI emitter. So bad that they've been known to cause interference that have to FCC knocking on the door.

Proof?

I keep up with the FCC's notices of violations when I can, and I have never heard of them doing any such thing. Many ham radio operators wish they would, though.

CFL's, on the other hand, used to be so bad that they had a warning on the box not to use them near radio receivers, especially in marine applications.

 

[Electric-Light]

Proof?

I keep up with the FCC's notices of violations when I can, and I have never heard of them doing any such thing. Many ham radio operators wish they would, though.

CFL's, on the other hand, used to be so bad that they had a warning on the box not to use them near radio receivers, especially in marine applications.

Not that CFLs or fluorescent are innocent but this explains something some LED ballasts do to control LEDs that fluorescent ballasts don't.
http://www.emcrules.com/2011/07/radio-interference-from-led-lighting.html

some concerns:
http://npstc.org/download.jsp?table..._Efficient_Lighting_Report_Final_20150630.pdf

https://ecfsapi.fcc.gov/file/10808183306390/LED Interference 8-4-16.pdf

 

[K8MHZ]

Not that CFLs or fluorescent are innocent but this explains something some LED ballasts do to control LEDs that fluorescent ballasts don't.
http://www.emcrules.com/2011/07/radio-interference-from-led-lighting.html

some concerns:
http://npstc.org/download.jsp?table..._Efficient_Lighting_Report_Final_20150630.pdf

https://ecfsapi.fcc.gov/file/10808183306390/LED Interference 8-4-16.pdf

You have not provided any proof that the FCC 'came knocking on the door' because of noisy LED's.

Your links are about consumers not liking the RFI from LED's. I can tell you that the FCC has received complaints for RFI from plasma TVs, electric fences, touch control lamps, surveillance cameras and many other consumer grade devices and has done nothing about it.

Your NPSTC (which is not a government organization) link has more complaints about fluorescent lighting than LED.

Here is the latest enforcement efforts done by the FCC. Please, show me ONE instance of the FCC doing anything about noisy lights.

https://transition.fcc.gov/eb/Headlines.html

I am not saying they are not noisy, but your assertion that they are so noisy the FCC is concerned is false.

 

[Electric-Light]

I am not saying they are not noisy, but your assertion that they are so noisy the FCC is concerned is false.

https://transition.fcc.gov/bureaus/oet/ea/presentations/files/oct14/21-RF-Lighting-Oct-2014-BH.pdf

You're going to have to do the research if you insist on specific individual incidents. I've explained steps you can take to address those under your control. I'm also providing remedies instead of giving up and assuming nothing can be done about for interference originating from interference devices including, but not limited to LEDs.

My understanding is that GE recalled inventory and offered replacement for affected fluorescent ballasts already in service.

http://www.wsj.com/articles/SB10001424052702304434104579378994224188328
https://www.lbagroup.com/blog/700-mhz-lte-uplink-interference/

 

[iwire]

You have not provided any proof that the FCC 'came knocking on the door' because of noisy LED's.

You're going to have to do the research if you insist on specific individual incidents.

In other words you cannot back up this claim you made here.

LED systems can be a notorious RFI emitter. So bad that they've been known to cause interference that have to FCC knocking on the door.

You made that statement, it is not up to K8MHZ to prove your statement is accurate.

Its a simple thing, provide a link or a source of that info you have shared with us.

 

[peter d]

In other words you cannot back up this claim you made here.



You made that statement, it is not up to K8MHZ to prove your statement is accurate.

Its a simple thing, provide a link or a source of that info you have shared with us.

I think you're far more likely to get some drivel and statistics about practically obsolete T8 systems. :roll:

 

[iwire]

LED systems can be a notorious RFI emitter. So bad that they've been known to cause interference that have to FCC knocking on the door.

Lucky I don't have LEDs in my own shed or I would not be able to listen to the radio.

 

[gar]

160929-2403 EDT

iwire:

I ran an RFI test in the AM band this evening. The LED bulb was a CREE 9.35 W 120 V 60 Hz part number SA19-08127MQFD-12DE26-1 and some other numbers.

I tuned my Panasonic all band radio to a somewhat quite spot in the AM band. The receiver was AC powered. Thus, some RFI can come from conducted noise on the AC line. Other noise is radiated. Dimmers in the house produced a moderate amount of noise. All dimmers and other noise makers were turned off. Now background noise was distant stations and atmospheric noise. This particular model of CREE has a current waveform which is very close to a sine wave with only small blips.

This CREE produced no significant power line conducted noise to the receiver. I had to bring the bulb itself within about 8 to 12 inches from the receiver antenna to produce much noise from the CREE. This is radiated noise or magnetically coupled noise.

By contrast the Feit bulb that I have previously described produces so much signal atound 800 kHz that my car has to be about 80 to 100 feet away from a street light that is powered from my power company transformer to get away from the Feit noise. A very big difference in the noise output of two different LED bulbs.

.

 

[peter d]

Lucky I don't have LEDs in my own shed or I would not be able to listen to the radio.

Nice shed.