Fuses for LED build

[ryan115]

Has anyone ever had a real world test on the fuses for a parallel string?
I am planning out a build and was planning on using these
http://www.mouser.com/ProductDetail/Littelfuse/0251001MXL/?qs=sGAEpiMZZMseCiJT91fwIpCtAz8CGq9CnfrQNvQWCSQ%3d
but I didnt know whether they would actually blow faster than an LED, looking at the spec sheets it could take around 1 second.
The other option I am looking at are the PLEDs that nuclearheli has implemented...
http://www.littelfuse.com/data/en/Data_Sheets/Littelfuse_PLED6.pdf

 

[2Quills]

My thinking is that if you calculate the appropriate sized resistors for your build in the first place then they will act as current limiting devices which would give your fuses the time they need to blow while resisting over current in the mean time. Everybody seems to be using the exact same resistors on all their builds. Which is probably fine if all you want them for is an easy way to check voltage. But it doesn't necessarily use them for what they are designed for. I could be wrong.

 

[insomniac2k2]

No real world test here, but these lights are rated @ 1amp. It would probably take somewhere near 1.2 sustained amps to actually blow them. You would have to get quite a quick spike to blow an LED before a fast blow fuse. That is a risk that im willing to take. Maybe im a gambler

 

[widmer]

It's not a situation where the LEDs are going to blow (or even sustain damage...) as soon as the current creeps over the spec sheet. The damage will depend on how hot the die gets, and this is the basis for why Cree will specify different max current ratings for emitters where there is 1 vs 2 or 3 dies on a single star.

In the second that it may take a fuse to fry and open the circuit, it is likely that not enough heat will build to do any damage. This is assuming that you're referring to splitting the current coming out of a meanwell ELN series driver. If you're trying to divide up a 6-amp current from a different driver, that might be a different story...

FWIW these are the fuses that I just bought:

http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=F2508-ND
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=F3130-ND

I have a special place in my heart for glass fuses...

 

[Jeff000]

I didnt know whether they would actually blow faster than an LED, looking at the spec sheets it could take around 1 second.
[/URL]

Look at the current time curve on the spec sheet.
The higher the current the faster it will blow. Any current spike that could damage the LED will blow the fuse in 1/100's of a second. 1 second of 20% extra current is nothing. 1 second of 200% current probably wouldn't do anything.
The drivers are so cheap I never bothered with parallel strings. You get more control this way too.
Even 10 extra drivers is only 300 bucks, and if you have that many drivers then 300 is still a drop in the bucket of the cost of the tank.

 

[2Quills]

Yeah I've always heard that the leds can typically handle a little bit of over current for a given amount of time. But for how long I think would depend on your build as widmer suggested. You probably have to look at the particular situation. How much over current are we talking in the event of an open short on a given build? What type of heatsink are you running them on, how well does your fixture dissipate heat? I think those things would all come in to play. I still think that if you went with appropriate sized resistors you would have to worry even less as there would be extra redundancy thrown into the situation.

 

[kcress]

ryan115; To directly answer your question; There have been many cases, in here, where people have blown the fuses with no damage to their LEDs. They do it by bumping things, shorting things, and mis-adjustment of their drivers. I've seen no mention of LED failures in those cases. So, yes, they seem to work fine in these cases. As for a long term issue like one LED shorts in a parallel string I don't recall seeing a case where that actually has happened yet. It's been commissioning issues where the fuses have saved things.

The devices you show are not appropriate for protecting a string from over-current events. NHelli needed them because of his complicated build that has him potting, in epoxy, his LEDs, making them problematic to replace if they fail.

Jeff000; You should not be using more than about 5 or 6 ELNs in one house. You need to use more powerful drivers if you have so many LEDs that you need more than 5 or 6 drivers.

 

[Jeff000]

Jeff000; You should not be using more than about 5 or 6 ELNs in one house. You need to use more powerful drivers if you have so many LEDs that you need more than 5 or 6 drivers.

Harmonics?

 

[2Quills]

Kcress could you tell me how one would determine whether or not harmonics would be in issue with a given build? What method would you use for testing for high harmonic frequency? Not doubting your knowledge or anything but I've spoken with guys over at one of the diy led supply sites that claim to have spoken with meanwell engineers who insist that running 10-12 elns would not be an issue with harmonics. Should we trust them?

Certainly agree that for bigger builds then larger drivers are a much more appropriate and can be a much more efficient way to build. Just wondering how to determine for ourselves and what to look out for. SMPS's do have some form of efi filtering don't they? Don't the capacitors limit where in the peak of the waveform that power is drawn from? If not can you explain what happens when you stack more and more of these drivers up and how it creates higher or more dangerous frequencies?

Sorry for all the questions but there are a lot of people out there who are still unclear on these issues.

 

[ryan115]

Thanks for the help guys...that makes sense. I guess I was thinking about the damage to LEDs being an overcurrent situation that could cause a domino effect before the fuses had a chance to blow, instead of a heat build up issue.

 

[Jeff000]

Kcress could you tell me how one would determine whether or not harmonics would be in issue with a given build? What method would you use for testing for high harmonic frequency? Not doubting your knowledge or anything but I've spoken with guys over at one of the diy led supply sites that claim to have spoken with meanwell engineers who insist that running 10-12 elns would not be an issue with harmonics. Should we trust them?

Certainly agree that for bigger builds then larger drivers are a much more appropriate and can be a much more efficient way to build. Just wondering how to determine for ourselves and what to look out for. SMPS's do have some form of efi filtering don't they? Don't the capacitors limit where in the peak of the waveform that power is drawn from? If not can you explain what happens when you stack more and more of these drivers up and how it creates higher or more dangerous frequencies?

Sorry for all the questions but there are a lot of people out there who are still unclear on these issues.

There are a lot of engineers that do not know how it effects real world things.

I never really thought harmonics would be an issue with the drivers, they seem too small.
Just keep your computers and other electronics off the same circuit, make sure you have a good system ground for your panel, and if you are wiring for the drivers pull a larger neutral.

Harmonics can get pretty complicated, I only know the basics as an Electrician, a electronics guy would know more. I just know how to use upsized neutral, shorter home runs, and install compensating devices. We have a meter that measures the harmonics, but it is about $3500

I think single phase systems (like houses) are more susceptible to issues then 3 phases systems though, but I do not really know enough.

 

[2Quills]

There are a lot of engineers that do not know how it effects real world things.

I never really thought harmonics would be an issue with the drivers, they seem too small.
Just keep your computers and other electronics off the same circuit, make sure you have a good system ground for your panel, and if you are wiring for the drivers pull a larger neutral.

Harmonics can get pretty complicated, I only know the basics as an Electrician, a electronics guy would know more. I just know how to use upsized neutral, shorter home runs, and install compensating devices. We have a meter that measures the harmonics, but it is about $3500

I think single phase systems (like houses) are more susceptible to issues then 3 phases systems though, but I do not really know enough.

Agreed it gets pretty complicated. I've read a ton about linear supplys, smps's, harmonics etc. And it just gets more mind boggling the deeper you get into it. Was hoping for a dumbed down version/explanation of how one would come to the determination of when or where you may reach that threashold of fairly safe to fairly unsure and should stay away.

Personally will be pulling two dedicated circuits on my latest build and lights will be on their own circuit just to be safe. And using fewer/larger drivers for the led build vs the alternative.

I've also heard the idea of running your drivers on power strips with EMI filters to help minimize THD. Is this a good idea for those who've already built a larger fixture with multiple ELN's or no?

 

[kcress]

Hi guys,

I explained the harmonics issue in this thread:
http://www.reefcentral.com/forums/showthread.php?t=1921614

Specifically post 12.

 

[jkopp36]

I can confirm that the fuses will likely save your LEDs. Something went wrong in my testing this weekend (still not sure what) and I blew the fuses on 6 strings of 14 XPEs driven by a HLG-240-54B. Replaced the fuses and all LEDs are fine. Using 1A fuses.

 

[psteeleb]

I can confirm the fuses work

I just started a new build of parrallel strings and have both fuses and resistors. I individually tested all the LED's but made some changes that apparently made for a weak connection using one of the JBJ solderless mounts. When I went to test the set up I saw a quick flash and then nothing, I started testing and tracked it back to the fuse that blew due to all the curent going through the other string. I don't know that it would have blown the LEDs but a fuse is comparitively cheap.

pictuer of fuses and resitors prior to putting together.

this picture is more of the resistor box but the red wires coming in from the left are the "split" positive lines going through the fuses, the white wires are the negative back to the ballast

I'll just add that I may get rid of the in-line fuses and add a 8 terminal type, like the one on the top of the page in the below link
http://www.littelfuse.com/data/en/Data_Sheets/Littelfuse_fuse-clips.pdf

 

[Ermin]

Resurrecting an old post.....

Is 1A fuse a sufficient value to use? If the serial LED string is powered at 1.2A from the driver.....wouldn't that current be enough to blow the fuse already?

 

[zachts]

Resurrecting an old post.....

Is 1A fuse a sufficient value to use? If the serial LED string is powered at 1.2A from the driver.....wouldn't that current be enough to blow the fuse already?

If using LEDs capable of over 1 amp sure, but really, and no one ever followed the rules of caution, if one string fails the current to other strings should still not exceed their rating before the fuses blow.........

LDD drivers are a better and the preferred method now rather than parallel strings in most cases.