Avoiding Strip Failures This Year

cdnbob

New elf
Joined
Jul 26, 2014
Messages
33
Location
Ottawa
Greetings.

I'm looking for some suggestions to avoid strip failure problems that I had last holiday season. Any suggestions about why I had such a bad experience and how to avoid it again would be welcome.

Last Christmas (2015) I added a 12 strip mega tree to my display. Advance testing went well, as did initial installation. So I started the season with a working mega tree. But then the strips started to fail. At first just the upper part of one strip. Then another. And another. After a month, all 12 were in various states of full or partial failure. After taking it all down, I tested the strips in warm conditions, and 2 of them worked again, but not the rest.

The common symptom is that part way up the strip the voltage goes funny. These are 5v strips and it can be the data line or the +5 line that shows wrong voltage. (I test the voltage by poking a probe through the silicon sheath - only after taking them down.) Typically the data line shows a constant high voltage above an LED and normal voltage below (a very small fluctuating voltage is what my tester shows for an operating data line). So I'm guessing that the LED failed and is shorting the data line and breaking everything from there on.

Visually, the LEDs work fine up to a point, and then beyond that they are either out or 1 solid colour. And they aren't all the same past the "break point", so some my be out some one colour, some another. And turning power of and on can change the colour above the break point.

I bought 3 of these strips in 2014 and had no problem with them. One of those 2014 strips was wrapped spirally around my light post in the front yard, very exposed to winter weather and worked throughout both winters (as a static light strip after the holiday season). So I thought they'd be a good choice for the mega tree too.

These are the strips I used: http://www.aliexpress.com/store/product/BLACK-PCB-5m-DC5V-WS2812B-led-pixel-srip-IP68-30pcs-WS2812B-M-with-30pixels-reverse-protection/701799_926778326.html

I replaced the stock connectors with Ray's pigtails.

There was no obvious signs of water inside the strip sheath when I took them down.

Weather was odd last year. Very mild up to Christmas Eve, then turning back to normal (sub-zero C all day). No failures before Christmas Day, but by New Years about a third were failing. Some rain before Christmas and some snow after. I don't think anything would have warmed up enough to melt the snow up to New Year's Day, so I'm having a hard time believing water got in. That said, the silicone filler at the end of the strip looks much more robust on the 2014 strips than on the 2015 ones. But at IP68, I wouldn't expect all 12 to suffer water intrusion.

So what do you think? I appreciate any input.
 

fasteddy

I have C.L.A.P
Global moderator
Joined
Apr 26, 2010
Messages
6,648
Location
Albion Park NSW
There are different failure modes that cause different issues.

1: One colour missing from the LED is normally a Bad LED and usually the red colour that fails, it can also be the chips output.

2: The chip can fail and then will not pass data past that to the next chip so effectively everything after that is off.

3: A bad solder joint on the components which can have many different effects depending on what component has a bad solder joint

4: A bad solder joint between the 0.5 metre strip sections, this normally will not work past the bad joint but if the connection has a high resistance then you may get varying results with pixels being stuck on and acting random

So it may be a combination of issues that you are seeing. Generally due to the cheap price of the strip, then not the highest of quality components are used and in many cases the chip itself is a copy and not the original and the LED has a very small wafer in it that's driven harder.

There is also the issue of product consistency and as most of these lights are built to a tight budget then most times the parts sourced are based on price and not quality consistency. So sometimes it can be hit and miss.

Very steep temperature variances can cause issues with solder joints cracking due to the thermal stresses, this is especially true for places like North America where the winter temps get very low but if the lights are then in the sun then there is a fair bit of thermal shock that can occur due to the temp difference between being in shade and being in the sun.

So hopefully that may help you gain a better idea of the different failure modes seen in strip and what is the main failure mode in your strip.
 

ShellNZ

Senior elf
Joined
Jun 17, 2012
Messages
892
Location
Karaka, Auckland, NZ
Sorry, I dont have any suggestions as to why you had so many failures, just wanted to say that my entire 4000px matrix is made of that strip and have only had a couple of pixels die in the last 2 years. Most problems I have come across when doing Xmas lighting has stemmed from grounding, everything grounded together?
 

bdt

Apprentice elf
Joined
Mar 31, 2013
Messages
88
Location
Clearfield Utah
I've had my share of failures with my strips as well. Most of my failures appear to be bad solder joints. Usually I can rub the LED or joint and it/they will work with the right pressure. The strips are cheap enough that its to be expected.
 

darylc

404 darylc not found
Joined
Dec 8, 2012
Messages
1,146
I think there are also bad batches from time to time.

For instance last year I built a matrix with 48 lengths of 2m 60 pix/m 5v 2812b strip.

Ordered it in 2 lots, the first lot perfect. The second lot same part number, same source, 1 month apart every single strip has failed leds, often 5-10 failed leds along the 2m length.
 

fasteddy

I have C.L.A.P
Global moderator
Joined
Apr 26, 2010
Messages
6,648
Location
Albion Park NSW
Strips are designed to be mounted to a hard surface, once they are mounted like this then you normally get good reliability unless there is an issue with LED or chip quality.
If you use strip without a solid backing and unroll and roll up all the time then you can expect solder joints to start failing
The strip I use has been installed for 2 years now and is on for 4 hours every night and has not had one failure in the 2 years of use
 

OzAz

Senior elf
Global moderator
Joined
Dec 29, 2015
Messages
997
Location
Newcastle
Fasteddy said:
There are different failure modes that cause different issues.

1: One colour missing from the LED is normally a Bad LED and usually the red colour that fails, it can also be the chips output.


That's similar issue with one strip on my new arches, worked the first few times of testing then last time one segment of 3 leds not displaying red :(
 

cdnbob

New elf
Joined
Jul 26, 2014
Messages
33
Location
Ottawa
Thanks for all the input. Much appreciated.

Temperature could certainly be an issue. All was fine leading up to Christmas when we had mild weather - above 0C even at night. At Christmas the temps dropped dramatically to near normal for us: -5C in the day and -15C at night. But at first the strips were good. I'd been concerned about what would happen when it got cold, but the first two or three days all was well. When the problems started to show, the failures would be seen right at the start of the evening when the lights came on. I never saw a case where LED in the strips would fail during the evening as the temperatures were dropping.

I ran things on a timer so at about midnight the lights go off and come back on at sunset the next evening. Might it help if I kept power on to the strips? Would enough heat be generated to provide some insurance against the thermal stress of the outside temperatures?

Would the power coming on by a timer contribute to the problem? The working strip from the year before was powered by taking 12v from a supply in the garage and using a 12v to 5v buck converter at the strip to give it +5v. For the mega tree I had +5v coming right off the power supply to the strips.

Can the chip fail (or solder joint) and short voltage to the data line? That's what I was seeing mostly in the failed strips. Just before the first failed LED, the data line had what looks like "proper" voltage (very low, fractions of a volt), but after the first failed LED it was up over 1 volt, sometimes well over 1 volt.

The strips were mounted on half-inch conduit, so they had a solid backing. The conduit was secured to avoid dramatic flexing or swaying in the wind.

Re grounding, yes things are grounded together. I am using and ECG-P12S controller for all 12 strips - 6 on each half of the controller. The P12 can't support the current for 6 strips, so +5v to the strips also comes off the power supply directly. Two power supplies, one for each side of the P12, and each side and the strips for that side are grounded to the power supply ground.

It's encouraging that others are using these strips with good results. My first 3 certainly work well, and handle the cold too, so far. Ray suggests it was a bad batch and is offering to replace them. So I have to decide if there are better strips to select or to try again with these. I'm thinking I'll give these another try.

Thanks again for the input.
 

fasteddy

I have C.L.A.P
Global moderator
Joined
Apr 26, 2010
Messages
6,648
Location
Albion Park NSW
Keeping power on the strips may help a little with the cold but im tending to think its more a batch issue as ray has suggested, in the end its a good outcome for you
 

cdnbob

New elf
Joined
Jul 26, 2014
Messages
33
Location
Ottawa
I attached the strips to the conduit using zip ties. The top was zipped tightly to prevent slipping, but that zip tie was clear of the strip itself and was over the end cap, so no risk of damage to the strip. The zip ties along the conduit were not too tight to prevent damage to the strip. I could slide the strip around the conduit if I needed, and that came in helpful when alignment didn't work out exactly as expected when hanging the conduit. The conduit had a hook bolted on the top so it could hook onto the support pole, and it essentially hung from the pole. But they wanted to twist a bit, so having loose zip ties allowed me to make alignment adjustments to accommodate the twist.
 

fasteddy

I have C.L.A.P
Global moderator
Joined
Apr 26, 2010
Messages
6,648
Location
Albion Park NSW
cdnbob said:
I attached the strips to the conduit using zip ties. The top was zipped tightly to prevent slipping, but that zip tie was clear of the strip itself and was over the end cap, so no risk of damage to the strip. The zip ties along the conduit were not too tight to prevent damage to the strip. I could slide the strip around the conduit if I needed, and that came in helpful when alignment didn't work out exactly as expected when hanging the conduit. The conduit had a hook bolted on the top so it could hook onto the support pole, and it essentially hung from the pole. But they wanted to twist a bit, so having loose zip ties allowed me to make alignment adjustments to accommodate the twist.

When using conduit it is important that when picking them up for install and removal that you do it from the centre and not from one end. The reason for this is that the conduit will bend and if the strip is on the upper bend of the conduit then it can damage the strip due to the different bending ratios of the strip and the conduit. At very least when handling strip mounted to conduit then face the strip downwards so any bending radius of the strip is less than the conduit so it will not stretch and potentially damage the strip.

I doubt this was your cause to your failures but I thought it might be appropriate to mention this
 

cdnbob

New elf
Joined
Jul 26, 2014
Messages
33
Location
Ottawa
The advice on handling the conduit is well taken.

As it happened, I was mostly right when installing last year. I laid the conduit on the ground with the strip up. I grasped the upper half of the conduit and raised it up and climbed the ladder a few steps at the same time until I could extend the top end with the hook up to the top of the pole to engage the eye. This kept the upper half pretty straight, and any bend in the lower half was such that the strip was on the inside of the curve and wasn't being stretched. There was some flexing at the very top as I worked to engage the hook in the eye, but I don't think the strip was stressed much by this. From practice with the conduit on its own I learned that having much flex in the top of the conduit made engaging the hook in the eye a very difficult task, so I kept such flex to a minimum.

I'm satisfied that stretching the strip wasn't my problem last year. If it was, then I think the problems would have been apparent immediately, and not all strips would have been affected. The most likely place that the strip might have stretched would have been near the top, and the points of failure on the strips were in the lower half in almost all cases. That said, I will look to support the conduit more carefully this year to ensure that stretching will not be an issue.
 

ShellNZ

Senior elf
Joined
Jun 17, 2012
Messages
892
Location
Karaka, Auckland, NZ
All of my house is strip on conduit and with me on the roof, son on the ground passes the conduit up to hubby on first floor roof who then passes it up to me, it gets scrapped along the gutter as it gets dragged up, sometimes its 2 conduits joined by strip and it twists and turns on its way up. Somehow......still goes lol :)

If I had any dead pixels it was generally due to the way they were in storage. The strip is ziptied to the conduit. If the little head of the ziptie I think pushed directly onto an LED I think that is what caused it to hiccup, e.g. lost the red most of the time I think. Thats me though, I stack all my conduit so just changed where I had the little end of the ziptie so that it wouldnt squash the strip it was being stacked on.
 

plasmadrive

Full time elf
Joined
Jan 12, 2012
Messages
248
Location
Elk Grove, CA
I had some strip failures last year caused by moisture. I had Ray custom make the lengths for me and mostly I had vertical mounted strips fail because moisture would gather at the bottom inside the sleeve. Not sure whether is was condensation, or leakage, but there was quite a bit of corrosion at the bottom of some of the strips. All of those failed. I did have many vertical strips that had no issue, so I am leaning toward them not being sealed correctly.

I did not have the issue on Horizontal strips at all.
 
Top