Triac issues (lots of them)

[Jerry]

• Hmmmmm, the longer this thread goes the more people come forward with controller failures, most with the same problem. Tells me the the triacs are not robust enough to be long lasting.
  

[jallenl]

I also have had at least 10 triac failures this past season. Happens mainly after some wet weather. Have had 2 failures on a single board on two occasions. Own only 1 gen 3, on its second season with no issues. Believe the gen 3 triacs are rated a little higher than previous version boards. Don't know if they are backward compatible, would be nice if this were the fix. Seems like every weekend I had boards to swap and repair. Still have one that needs to be repaired. All of my failed triacs are on led lights, snubbers have helped a little, didn't have them in 100% of available channels till halfway through the season.

[Robert Burton]

Uncle Al, not sure if I would agree with that on the bad manufacturing. I mean sure..that would make sense if it were all on one generation of controller. But we're seeing this over several generations. Now...wonder. We had a "Wetter" season for the most part up and down the east coast. Soooooo.....

How many of you that had failures were running the controller that had the failures on a circuit that did not have a GFI or GFCI on it? It just hit me that the one controller I had failures on was just that. The other 2 right next to that one....didn't. This was my RED mini tree controller (16 mini trees) next to my white and green mini tree controller.

Just throwing that out there ;-)

[Jeff Messer]

I had a couple of G3 inside the shed that lost channels on them (stay burning). I have 6 controllers in the rain with a 5 gallon bucket over the top of them and never lost a channel on them Yet. The other 25 are inside cabinets or shed and this is where my issue happens.Started thinking it was a heat problem so I put a fan inside one of my cabinets.

Jeff

[Robert Burton]

Well not really looking at how the controller is water-proofed. I think most people know that this has to be addressed. Mine are in the LOR boxes. If that were the issue, I think more of the triacs would have gone out. I am thinking that the "element"...mine was mini trees.... were "wanting" to trip the GFI, but couldn't since it was on a Non-gfi plug (not even going there...fuss over you should be using one...). Just the fact it may have grounded out some how because of that. Heat...yes. I will be interested to see if the "goo" is dried up on mine or not...esp. on those channels that had issues. If that is the case, I will be "re-gooing" all of mine every year. That is a pretty heavy heat sink for what we put on it.

(if you use the heavy heat sink)

I guess we all need to look at what load is on that element. Mine, 400 mini incan's....shouldn't have been a problem.

[plasmadrive]

I have given this triac issue a bit of thought and here is what I think may be happening to those with triac issues.

• First know that a triac goes from zero to full current flow in about .000003 sec. (3us). That is pretty darn fast.
  
• The way these controllers are made the triac switches on the live side of the power. This means the load is connectect to the triac and to neutral.
  
• Neutral and ground are the same point at your rmeter main. What this means is that you can pull current thru ground... Lots of it! This is what GFCIs are to protect against in regards to human contact.
  
• GFCIs will trip in as little as .008 sec. (8ms).
  
• Incans are resistive devices. With incans the more current you draw the more voltage they drop. This has most likely held off a lot of failures in the past that could have been caused by moisture.
  
• LEDs are not resistive devices. They drop a certain voltage and that is about it. Once you reach the threshold where they conduct, anything after that is a short circuit (so to speak). This could have a lot to do with why more triac failures are prevelent now a days and on LED systems.
  

Notice the disparity in the time it takes a GFCI to trip vs how fast a triac can turn on and haul full current?

If for some reason, via water, moisture, twigs or for "whatever" , you turn on the triac and you get leakage current from say.. the hot side of your light string to ground.. the load, (light string), is not even in that circuit at that point. It is just power-triac-ground (return path). That is a short circuit with the moisture and wire being the only resistance in line with that short! That means the triac can conduct a very large amount of current in a very short amount of time thru that path of moisture or "whatever". That current pulse may or may not be enough to harm the triac in one cycle, but it won't do it any good. Keep doing that a few times and you can damage the junction of that triac and cause it to fail prematurely. Chances are that the moisture or current path may evaporate very quickly or be litterally blown away.. opening up that short so fast you can't see it or even tell it was there... but it was there... Triacs are pretty robust and have a pretty high surge current capability.. but they can't do it over and over and/or for long periods of time. (We are talking usec or msec)

Remember, once triggered on, a triac will not shut off until current thru it reaches almost zero. You could use inline inductors to slow this current pulse down, but this is Christmas light fun... not too cost effective to do that. We should all understand that the controllers that LOR sells are not of poor quality or design. They are actually quite well done and within a reasonable cost/value range.

There is still a possibility that a bad batch of triacs exist but it is pretty unlikely given the small amount of failures. They are made in very large batches with a very repeatable process and the technology is very mature.

In any case LOR would have no way to determine if a triac was bad before it was installed unless they tested ever single unit and it failed prior to, or during assembly and testing. Since I have yet to see anyone complain of a triac being bad out of the box, I would conclude... from a technical standpoint, that WE, or our displays are causing the failures... That would not be a quality issue on LOR's part.

Before you ask, NO, I have nothing to do with LOR.. I have a lot of experience with this technology however..

I post this to try to help everyone that doesn't already, understand what the true issues are regarding triac failures, especially with LEDs. Nothing more..

I hope this is helpful to some and didn't bore others to death..

Edited January 20, 2013 by plasmadrive

[Robert Burton]

Plasmadrive - I totally agree that LOR's stuff is VERY well built. Everything (for me) makes sense. I sometimes get leaves, or pine straw blow onto these mini trees. Of course when it is wet, well that doesn't help.

I would be real curious to see if some of the elements are bushes, grass, trees that might add to to that resistance. (Folks please post...)

I just don't remember this many triacs blowing last year. My only reason for mentioning the software thing. But again, just throwing it out there.

[plasmadrive]

Hi Robert,

Software pretty much has zero to do with triac failures for these simple controllers. The triacs turn on from zero to full on where ever in the cycle they are told to turn on. They don't really care where in the cycle that is..

Mostly the moisture on those other elements are usually the cause as opposed to the elements themselves.. But they can contribute.. it is a lot of a guessing game..

Craig

[Uncle Al]

Uncle Al, not sure if I would agree with that on the bad manufacturing. I mean sure..that would make sense if it were all on one generation of controller. But we're seeing this over several generations. Now...wonder. We had a "Wetter" season for the most part up and down the east coast. Soooooo.....

How many of you that had failures were running the controller that had the failures on a circuit that did not have a GFI or GFCI on it? It just hit me that the one controller I had failures on was just that. The other 2 right next to that one....didn't. This was my RED mini tree controller (16 mini trees) next to my white and green mini tree controller.

Just throwing that out there ;-)

The failures I had were on Gen2 controllers purchased in 2011. Both were on GFCI's controlling 100-600 incans per channel. One was controlling wire-frames the other roof lights. All the other channels in these boxes were controlling similar circuits. I'm still thinking it was just defective triacs.

[-klb-]

It seems to me that one of the DIY types was investigating a range of ST semiconductor production lots for suspected high failure rates. I will say that with controllers going back to 2007, all my triac failures are 2010 and newer. I believe that the shortages of controllers and late shipping in 2011 was documented by LOR to be in part directly related to high failure rates in the triacs. My understanding is that they have taken action to get a better supply. So hopefully we won't see the same issues, or at least not the same rates in the 2012 and newer controllers.

Just because a technology is mature does not make it immune from corner cutting, and the risk of people not following all production quality control practices.

[plasmadrive]

It seems to me that one of the DIY types was investigating a range of ST semiconductor production lots for suspected high failure rates. I will say that with controllers going back to 2007, all my triac failures are 2010 and newer. I believe that the shortages of controllers and late shipping in 2011 was documented by LOR to be in part directly related to high failure rates in the triacs. My understanding is that they have taken action to get a better supply. So hopefully we won't see the same issues, or at least not the same rates in the 2012 and newer controllers.

Just because a technology is mature does not make it immune from corner cutting, and the risk of people not following all production quality control practices.

This is true. I would be surprised if they have that many failures in a batch.. but it can happen.. I wonder LOR has to say about that.. Is that something they said publicly?

As well, if there were bad batches, they should have shown up one before now.. unless it is something like contamination or bonding issues..

[Jerry]

Just for the record, mine was on a pole 2' off the ground, Yes it was connected to a GFI breaker, powering 400 incand. mini lights.

NOW, here's the is the not so funny part. The last time I ran the show all worked just fine. I don't cut the power to the controllers when the show is not running (I may in the future). We got rain and wet snow (never an issue before) and 2 days after the storm I noticed the GFI was tripped. The show was not running during the storm or the 2 days afterwards (the show season was over). When I reset the GFI and went out to check things I found one channel on at about 50%. So my triac went bad just sitting there with no show running for days. We had over 3" of rain during Super Storm Sandy and my Holloween show was up and didn't have any problems other than a tripped GFI breaker.

[plasmadrive]

Jerry,

That is an interesting problem.. Do you know if you had any power hicups during the storm. OTHER THEN A BAD BATCH of Triacs.... (which has not been ruled out)

here is a "What if".

Lights or plug got wet and the line side from the triac had a water path to ground... when the power glitched (or a data hicup, which is possible) the triac fired for a short time, like one half cycle... current path to ground directly from the triac. .. and there you have a possible high over current issue causing the triac to fail. Something had to have happened to cause the GFI to trip right? Even if the triac had been triggered on, or shorted for that matter, if there was no path to ground for power, the GFI would not have tripped...worse would have been the lights came on.... so the GFI tripping is telling you that something conducted to ground.. Why the triac supplied power is a mystery of course but the triac itself did not cause that GFI to trip. (It can't whether it was good or bad).

[-klb-]

If power were interrupted, the power on state of the channels is undefined, and they may briefly be on during power up. So a power outage might be a contributing factor.

I'm still going with some batches of triacs are better than others, as more than half of controllers I have are from before 2010, and have zero failures. But all are equally mixed across uses.

[Jerry]

Jerry,

That is an interesting problem.. Do you know if you had any power hicups during the storm. OTHER THEN A BAD BATCH of Triacs.... (which has not been ruled out)

here is a "What if".

Lights or plug got wet and the line side from the triac had a water path to ground... when the power glitched (or a data hicup, which is possible) the triac fired for a short time, like one half cycle... current path to ground directly from the triac. .. and there you have a possible high over current issue causing the triac to fail. Something had to have happened to cause the GFI to trip right? Even if the triac had been triggered on, or shorted for that matter, if there was no path to ground for power, the GFI would not have tripped...worse would have been the lights came on.... so the GFI tripping is telling you that something conducted to ground.. Why the triac supplied power is a mystery of course but the triac itself did not cause that GFI to trip. (It can't whether it was good or bad).

All my controllers are in LOR boxes (they should know what works best, right) . I attribute the GFI trip to water getting in a power supply plug, all are the 3 blade type. Don't know if enough water got in there if it would short the hot side to the ground side inside the plug, and trip the GFI. If so, this condition shouldn't affect the triac , I wouldn't think.

As for power hickups , my computer is on a quality surge protector , both the power in side and the incoming DSL side. We do have a cheap clock that must be very sensative to power hickups , it goes off every time our house lights even blink. It didn't trip off this snowstorm.

This leads me to think it's all due to a triac issue of it's own and not an outside issue blowing it. Why else would it go bad just sitting there ?

[plasmadrive]

All my controllers are in LOR boxes (they should know what works best, right) . I attribute the GFI trip to water getting in a power supply plug, all are the 3 blade type. Don't know if enough water got in there if it would short the hot side to the ground side inside the plug, and trip the GFI. If so, this condition shouldn't affect the triac , I wouldn't think.

As for power hickups , my computer is on a quality surge protector , both the power in side and the incoming DSL side. We do have a cheap clock that must be very sensative to power hickups , it goes off every time our house lights even blink. It didn't trip off this snowstorm.

This leads me to think it's all due to a triac issue of it's own and not an outside issue blowing it. Why else would it go bad just sitting there ?

Ok.. the computer really hasn't anything to do with the triac issue so lets take that out of the equation. The only exception to that is if it told that channel to turn on. unlikely though..

I have one of those clocks too, so I get that part... lets take power glitch out fo the equation as well.

Triacs do not just blow for no reason or without some sort of cause.. most of which is listed previously, (there are always exceptions of course).. GFIs also do not trip for no reason unless they fail or are bad. Let us assume (we need to make some assumptions with the given data) that the GFI is good.

For the GFI to have tripped, something had to cause either an over current flow thru the hot lead which would have caused it to trip like a breaker, or there had to be an imbalance of hot to nuetral current. Let us assume that latter..

• There had to be a power source and a return path that were unequal thru the GFI. It could have happened prior to the LOR controller in which case it would not be relavant to the triac issue but pretty unlikely I think.
  
• That means there had to be something that caused that imbalance of current.. It could have been that the triac conducted, for whatever reason, and the current path to ground was thru the moisture from the lot side of the light strings or the plug. That could indeed cause the triac to short due to a high current pulse. Remember the LOR controller protocol only sends changes in commands. If you had a batch of noise that caused the channel to turn on, it would remain that way until told to turn off. (If I understand the LOR protocol correctly) This is just a "what if".. We have no evidence of what really happened or in what order.
  
• Lets say the triac shorted first.. the only thing that would do is turn on the lights, not trip the GFI. If the triac shorted first, then the water issue occured, that could trip the GFI.
  

So there are a couple of possible scenarios here. Neither of which can be pinpointed with the data at hand. I know this is somewhat over blown on my part, but as a manufacturer of power electronic "stuff", I see too many times where the blame for a failure goes where it doesn't belong.. so analizing this is kind of theraputic to me..

It could still be a bad batch of triacs.. but in your case.. something else happened.. whether relevant or not.. it did happen.. the GFI tripping is proof of that.

I am like you in that I keep my system power all the time. all my controller strings are talking via ELLs to the computer.

[Robert Burton]

Mine is sitting on the my work bench, waiting for me to doctor on it. I will do that next week when I get back from a business trip. Almost feel like take a string of lights and trying to replicate the cause. Replacing 2 already...why not 3.

[Santas Helper]

I'm still going with some batches of triacs are better than others, as more than half of controllers I have are from before 2010, and have zero failures. But all are equally mixed across uses.

I'll agree with the "some batches" as manufacturers are not perfect when manufacturing these things. I have early model controllers, over 20 of them, and have had maybe 6 triacs go bad over a 6 year term in all +20 controllers.

Maybe it's just me but it seems (observing complaints from others recently) that older model controllers might have been more robust in the triac area. But that's just me thinking.

Then again it could be the climate in different areas with humidity, rain, snow and such.

Edited January 24, 2013 by Santas Helper

[Jerry]

Thanks guys. Replacing the triac when it gets here. Hope this isn't going to be a frequent duty .

[plasmadrive]

Mine is sitting on the my work bench, waiting for me to doctor on it. I will do that next week when I get back from a business trip. Almost feel like take a string of lights and trying to replicate the cause. Replacing 2 already...why not 3.

I doubt you find find anything to point to the cause, but if you do, please let us know.