Yay!

Yay!

Now that cheers me up fluorescents rule LEDs drool!!

The 600 volt OCV of the ballasts is doing this, they can revise it all the way and it's still not going to work...too bad!
Fluorescent's rule!

So your cheering on the troubles of an American owned company that has a substantial presence here, and that environmental impact of manufacturing and disposing of those tubes with a design flaw? Companies aren't going to go back to inefficient fluorescents, they're going to buy LEDs from a non-American company.

A fluorescent fixture should be equipped by fluorescents. If someone wants to use LED's, the only really working way is to replace the fluorescent fixtures for the LED ones.
Trying to mimic some other technology is alway asking for huge troubles, mainly when the power supply characteristic is practically not known at all Wwith directly mains powered incandescents at least the supply specification is clear, but when replacing the fluorescents or so, the ballast characteristics differ in a very significant way, so it is really technically impossible to match all of them in aany other way than really mimicking the fluorescent characteristic. But then the LED tube will have exactly the same powerr input as the former fluorescent, plus with a power dissipation too high for the size.

It is sad the managers of that American company do not want to believe into laws of physics, this is the result.

The problem is not the 600V. If it is a hole melting above some specific spot in ll tubes, "yay" for some transistor with exceeded power dissipation there, by a design flaw or miscalculation. If it is along the whole tube, "yay" just for the general thermal design

Actually, running on the output of a Fluorescent ballast is much more favorable operating condition than directly on line voltage (each condition with a tube designed for it), as the ballast is getting all the line surges and transients, and the LED tube is getting clean filtered power



Yay for the reminder that LED tubes are unnecessarily complicated devices

Yay for Cree's business practices, recalling defetive products promptly even 2nd time is showing responsible business practice, so actually put out Cree' overall business in good light, more so than the failure put their engineering in bad light in the 1st place

Not yay for the waste tubes, which production is WAY more polluting than production of Fluorescent tubes, and the percent of materials recoverable from recycling them is WAY less than for any Glass and Mercury products



Companies aren't going to go back to inefficient fluorescents, all can i say, bad for them. We are talking about retrofit tubes working on the output of a Fluorescent ballast, so count the ballast output power as the Watts

Ordinary "36W" T8 triphosphor tube : 3250Lm / 32W (on High Frequency) : 101.5 Lm/W

Very same Cree's recalled LED tube : 1700Lm / 21W : 81 Lm/W

Carefull when talking about "inefficient Fluorescents"

Sure they are, when a 21W LED tube is equivalent to 32W Fluorescent. Or maybe its just the usual equivalence fraud, when the 1700Lm is claimed to be equal to 3250Lm (or even to 2500Lm cheaper Halophosphor tube)

http://www.cree.com/Support/Recalls/T8-Recall

So your cheering on the troubles of an American owned company that has a substantial presence here, and that environmental impact of manufacturing and disposing of those tubes with a design flaw? Companies aren't going to go back to inefficient fluorescents, they're going to buy LEDs from a non-American company.

Whoever said fluorescents are inefficient? 😝
Although if you mean by the way companies think that makes sense.

Nothing on lux in those spec sheets Ash? Like 6~7' away from a table surface on which a light source is shining on. How do they compare?

Hey Cree! When's LED tubes version 2.0 coming out?

Nothing on lux in those spec sheets Ash? Like 6~7' away from a table surface on which a light source is shining on. How do they compare?

Hey Cree! When's LED tubes version 2.0 coming out?

That WAS version 2.0!

Quote from: FGS on August 27, 2016, 05:39:29 PM

Nothing on lux in those spec sheets Ash? Like 6~7' away from a table surface on which a light source is shining on. How do they compare?

Hey Cree! When's LED tubes version 2.0 coming out?

That WAS version 2.0!

Oops. I meant 3.0! LOL!

3.0 will do the same, it's the 600 volt OCV that's doing it. Don't want any arguments about that either!
Also, your first post with the gif, I had to delete since it was completely off topic and didn't relate at ALL to this!

At the table surface you measure Illuminance in Lux, not Flux in Lumens. The lamp manufacturer does not know what height the luminaire is above your table, what luminaire optics you have and so on, so they can't even remotely guess the Illuminance. They specify what really relates to the lamp itself, so the Flux

In the same luminaire over the same table, with same distribution lamp (so the diffused 360deg tube), the Flux is proportional to the Illuminance. So if in your setup the light spread to 10 m^2, with 3250Lm lamp you get 325 Lux, then with 1700Lm lamp you invariably get 170 Lux

600V is not a legendary Electric type. It can be handled very well with the appropriate rated components, that come in same packages as the usual 400V stuff, and there is no problem to fit them in a LED tube. Actually i dont think the problem is there anyway : The tubes overheat in use. If the problem would be with voltage withstand, it would overheat as result of some breakdown of a component, so happen at a time of tube burning out

Ash, I'm assuming what has happened with these is the same as what happened last time. The 600 volt OCV from the ballasts were causing the contacts inside the tube to burn up from the article here a quote from that is "Hazard: Electrical resistance between the spring contact and the printed circuit board (PCB) may cause over-heating due to electrical arcing. The printed circuit board can overheat and cause the LED T8 lamp to melt, posing a burn hazard." So I was assuming it's an identical issue here.

I strongly doubt any semiconductor failure may cause local melting. The main reason is, above 300degC (what is not that much) it becomes a hard short conductor, so stops any significant dissipation (that is just the physics; the process itself is perfectly reversible, as the 300SdegC is nothing to cause any damage on chip itself for a short time).
But the known problem makers for red/orange hot spots are ceramic capacitors for low impedance fed circuits and the arcing for high impedance fed circuits.
Because it operates from a fluorescent ballast, the second is really the highest probable cause.

The thing is, the same problem affects fluorescents as well, but there the socket assembly itself is designed to handle such failures safely (if it meets the UL requirements). But the LED tubes have more of such spots and the inn er ones are not that robust as they should be.



And for the operation from a ballast vs directly from mains: That would be true, but only in case you know exactly the output characteristic of the ballast and design for it. That is very likely the case with series reactor ballast, so in 230V areas. But in the 120V areas, where the ballast has to contain some voltage boosting and with electronic ballasts you just can not be sure at all. The related standards describe just the tube characteristic and the required operating point (and that means the rated power), but that still gives many degrees of freedom for the ballast designer. So the only possible way to ensure compatibility is to mimic the known side, that is the lamp. But that means the full 32W (for a 32W fixture) power to be delivered to the LED's. And that is about double to what the users want from this retrofit.
The fact there is no standard about the exact ballast characteristics means, making anything operating on all of them in a reliable way is just impossible.
With a direct mains connection you do not have the inherent surge protection from the ballast, but otherwise there is prettry clear standard on how the supply is supposed to look like, so you know what you are designing for.