Replacing bad triacs

[Tim Fischer]

Just now, Darrell said:

All the prior comments indicate that snubbers should only help with jerky dimming.  I'm still left with controllers that have channels with a constant "bleed" that will light an LED string at 10% which is enough to be visible in the yard display when that string is suppose to be off.  Can triacs "age" or degrade with time/use to allow some bleed?

That's just been my experience. I'm definitely not an electronics expert.

If putting a load on the channel helps, it seems like that's what you can do for a fix. An appropriately high resistor value (sorry I can't help with that but hopefully others can) for example. Speaking of which, what resistance value are your current snubbers using (or maybe I should ask, what are you using for snubbers)?

 

[Darrell]

My snubbers are DIY ones from about 10+ years ago.  I used a rubber 110v repair plug and soldered a resistor across the poles.  I think the resistor is a 1k ohm resistor.  The open end where the cord should be was closed with a liquid sealant before re-assembly.  This design was from old posts on earlier versions of this forum.  Reading some older posts on other boards, I am reminded that Gen-1 and Gen-2 LOR controllers often needed snubbers for LED strings to present enough "load" that the triac circuit would see the load and fully trip the triac.  Snubbers were for more than just smoothing out the dimming curves.  I guess that over the years some of the snubbers I had went bad.  This gave the appearance of the triac or channel going bad over time.  I hate getting older and forgetting what I already knew at one point.  Now I need to go back and retest all my Gen-2 controllers AND my snubbers.

[Tim Fischer]

Good info. I guess I never needed them other than on a few circuits that I have a lot of LEDs on, and that was just for smooth dimming. All my controllers are Gen 1.

The snubbers I do use are very similar. I'd have to put a meter on them to see what value resistor I used (or I think in my parts kit I have a few spares labeled "snubber" too, along with some more marked "DMX Termination" lol) 

[Darrell]

I tore apart one of the old snubbers.  Here's a better picture of how I made them.  This shows that it is a 47k ohms resistor.  It also shows the rust inside after 10 years in the weather.  I had thought they were better sealed.  The 1k ohm resistor in the 1st picture must be from some other project I've forgotten about.  Using the multimeter directly on the resistor shows it is blown.  5 out of the 12 snubbers I have quick access to read as bad with the multimeter.  I guess this is one more thing I will have to check and test as I set up again next year.  Most of the lights have already been packed away.  There always something ....

Thanks to everyone on your comments and suggestions.  It definitely forced me to keep looking and has saved me from tearing apart my controllers to unnecessarily replace triacs.

[Steven]

1 hour ago, Darrell said:

The open end where the cord should be was closed with a liquid sealant before re-assembly.

That's completely unnecessary and counter-productive. If the end is sealed then when water wicks through the prongs it won't have any way to dry. Just leave end ends open. Water may get in when it rains, but it will quickly dry if left open and won't cause any damage.

[rmturner54]

I guess iam lucky.  I have 2ea gen1 AC controllers probably 10 yrs old. Never used snubbers or needed them. I have 6 channels of 8 strand of red, green, blue, 50 leds each. So thats 350 to 400 leds each. They are full wave. Only changed 1 triac in all that time. It was obviously bad. Would not turn off.

But I agree with not sealing the dyi snubbers.

[Tim Fischer]

When I built mine I just squirted some hot glue in where the cord is supposed to go in. I didn't really make an effort to totally seal them, but having a big 1/4" hole in the back seemed wrong too.

Maybe these days I'd 3D print a little plug to close the hole without making it watertight.

[Darrell]

Here's a status update.  I ended up replacing 7 triac's spread between 5 different Gen-2 controllers.  Repeating the testing before and after the repairs - a bad triac can allow an LED string to light up at about 15% on all the time.  If tested with an old style incandescent lamp, the triac will show as good.  The incan lamp will fully turn on and off on command but if an LED string is plugged in, it will stay on at 15%.  It will still fade up to 100% and back down to the 15% level.  After changing the triac, that same channel will now work properly with 100% on or fully off - no power leakage.  I did end up with one bad channel that replacing the triac did not work or I screwed up the replacement.  It is now completely off all the time.  Luckily, I now have enough spare channels again that I'm not going to worry about trying to fix it again.

One the issue of the snubbers - I tested about 50 and approx 1/4 of those were blown.  As I took them apart to replace the resistor about 1/2 had the rust as shown in the earlier comments and photo.  However, even with the rusty ones, the problem was really a blown resistor.  Even the snubbers that did not have much rust inside still had a blown resistor.  These snubbers were all about 12 years old.  I don't know what caused the resistor to blow, but for the cheap cost of a new resistor after 10+ years, it was an easy fix.  Just one more thing to be aware of.

 

[rmturner54]

Thanks for sharing the results. My suspections following this was triacs. I have replaced 1 triac (AC) control and 2 Mosfets (DC) in all the time I have been doing this.