WS2811 Technicolor Strings Have High Failure Rate

[arw01]

Sorry, fellows I didn't go the bleeding edge, i thought WS2811 would be bleeding edge, but those of you who have a string that fritzes out with a misty rain, what about drying them back out, making sure they are working, spraying the nodes and entry points with some corrosion x which should keep the water from contacting whatever small spot it's finding.

[Made2Rock]

The "Dimmer" tool for xseq (xLights) raw files is now complete:
 
SeqXlater-xseq.pl
 
 
The conductor file version was also updated -- it had a problem where the channel number comparison was off by one.
 
SeqXlater-seq.pl
 
 
Both are found here:
 
You are not allowed to view links. Register or Login

 
Note: I found it necessary to do the xLights file format because it can deal with more than 4x4096 universes.  Conductor format is limited to 16384 channels because Conductor uses multiple units to handle every new group of 4 universes.

Got them, Thanks Steve

[keitha43]

Back in the day when the original ip66 nodes came out, I tried hairdryers and such and in almost all cases the nodes wouldn't start working again. So I replaced them. I went through about 70 nodes that season doing repairs.

[arw01]

Back in the day when the original ip66 nodes came out, I tried hairdryers and such and in almost all cases the nodes wouldn't start working again. So I replaced them. I went through about 70 nodes that season doing repairs.

Even after being in rice for a week, they would not come back alive?

[keitha43]

I never tried rice. Just hairdryer and indoors a week. They had already been soaked in watered down plastidip before they ever were set up outdoors that year. I had one replacement string of the newly created ip68 that went the rest of the season without a failure so I made a deal with Ray to send all the rest of the ip66 to him and he sent new ip68 after the season. But I had to pay the return shipping to China. Boy that was expensive.

[dduck]

I my r&d,  I found no conformal coating on the pcb and wire.  Measure the resistance on the data and gnd wires and see what the value is.  The normal value should be the high megaohm.  I plan on testing some of my modded nodes to see how they hold up.  I would like to get hold of some of the final prototypes for analysis.

Just checked with Ray.    This was his response on my problems:

very sorry for the problem.
yes,please adjust the 12v power supply to 10v around to reduce the heat.
i checked your order, all of your order has been made with conformal coating, so please no worry about the moisture problems.
anyway,there would be some failed nodes, we could not ensure 100% working,hope you understand

FYI I received my TC's mid November.

My plan is to:
1) dry out these strings
2) check the resistance.
3) replace the bad nodes.
4) keep my voltage at 10V, intensity at 50%

[Made2Rock]

I was so pleased (cough) with the performance of the megatree I took it down.

Ray has so far failed to stand by his product which he has admitted to being defective. I guess this is the chance you take dealing with him and I like many others have been burnt.

Reducing the voltage will help reduce the amount of power you are asking the TC node to dissipate. You can also find a utility to reduce the intensity of the LED created by Steve Gase. I would strongly recommend using that. From my testing you can reduce the intensity by 50% and it will still bright enough. Together this may help these !@#$@$^% nodes to survive.

Joe

[jnealand]

I have made a number of orders with Ray over the last couple of years and have not had any problems with him.  I think I would rephrase your statement to be "I guess this is the chance you take dealing with new unproven technology".  I'm an old fart and my motto is that you should never buy version 1 of most products.  In my life I have seen more new products fail in their early deliveries than succeed.  new products always seem to have great promise, but .....   You pay your money you take your chances.  Just my well seasoned opinion.

[Steve Gase]

I have made a number of orders with Ray over the last couple of years and have not had any problems with him.  I think I would rephrase your statement to be "I guess this is the chance you take dealing with new unproven technology".  I'm an old fart and my motto is that you should never buy version 1 of most products.  In my life I have seen more new products fail in their early deliveries than succeed.  new products always seem to have great promise, but .....   You pay your money you take your chances.  Just my well seasoned opinion.

well, I don't see these as version 1.


the injection-molded casing solution was based on 2 past generations (the commonly-deiscussed IP66 and IP68 nodes)... it specifically was intended to solve all the past issues.


the use of WS2811 was intended to leverage the success of a commonly-used, and now more-standard chip, already used successfully by many other strings.


the vendor has been doing this product for years, and for the most part a known and trusted entity.


the use of TM1804 chips was a move to reduce the risk even further -- risk, not from the chips, but risk due to the newness of using SSCs with WS2811.


EVERYTHING in this product was intended to use proven, successful techniques and materials to create a product that would withstand additional years of use -- beyond the IP68 product.


The testing by Zach with freezing, with submersion, etc. provided the basic validation. Early testing of Ray's WS2811 strings by RJ sealed it.


So... there was definitely an understanding that this was a final, ready for anything product.  We were charged for a production product, and not given a discount to try samples in a "beta" test.

[Steve Gase]

Update... 


I still don't have my tree up.  I am sure that I am being too methodical for the time left in this season, but I am trying to avoid any short-cuts that might make matters worse.


I received my buck converters over the weekend, and I did some testing with both WS2811 (50ct) and TM1804 (100ct) strings.


Inside my shop I have set the buck converters to 8.5v and I've run the lights at 50% intensity and they look pretty good at all (50%) white. 


I am slowly building (50) buck-converter-pigtails and tuning them into the 8.5v range.


I have altered all sequences to be pre-dimmed for the megatree, so that when I connect the tree I will have everything ready.  I have made the dimming step part of my procedure...  (1) NC to .xseq file, (2) dim the tree lights, (3) convert to controller file format.


Once all pigtails are ready I will resume my assembly of strings onto EMT.

[Hauvega]

The construction of the prototype strings might be different than the production strings.  Changes could have been made to reduce cost or reduce production times.  This could have happened during production.  Conformal coating process has a cost and time to complete. 

[Made2Rock]

I have made a number of orders with Ray over the last couple of years and have not had any problems with him.  I think I would rephrase your statement to be "I guess this is the chance you take dealing with new unproven technology".  I'm an old fart and my motto is that you should never buy version 1 of most products.  In my life I have seen more new products fail in their early deliveries than succeed.  new products always seem to have great promise, but .....   You pay your money you take your chances.  Just my well seasoned opinion.

well, I don't see these as version 1.


the injection-molded casing solution was based on 2 past generations (the commonly-deiscussed IP66 and IP68 nodes)... it specifically was intended to solve all the past issues.


the use of WS2811 was intended to leverage the success of a commonly-used, and now more-standard chip, already used successfully by many other strings.


the vendor has been doing this product for years, and for the most part a known and trusted entity.


the use of TM1804 chips was a move to reduce the risk even further -- risk, not from the chips, but risk due to the newness of using SSCs with WS2811.


EVERYTHING in this product was intended to use proven, successful techniques and materials to create a product that would withstand additional years of use -- beyond the IP68 product.


The testing by Zach with freezing, with submersion, etc. provided the basic validation.  Early testing of Ray's WS2811 strings by RJ sealed it.


So... there was definitely an understanding that this was a final, ready for anything product.  We were charged for a production product, and not given a discount to try samples in a "beta" test.

Exactly. In fact I didn't get any RGBs last year because they were in my opinion still working out the bugs. This is the new generation

Joe

[tbone321]

Which is why I waited yet another year.  Ray is a good guy but he really needs to crack down on his suppliers to get this right.  I don't mind paying a few more cents per node to be sure that they actually work.  Having to drop the voltage and run them at 50% output to hopefully prevent them from failing is completely unacceptable and him just saying I'm sorry isn't any better.

[jnealand]

Exactly. In fact I didn't get any RGBs last year because they were in my opinion still working out the bugs. This is the new generation
Joe

I didn't post to pick on you guys, I posted to defend Ray.  A new product is a new product even if it is built from a kluge of older products.  This is a 1st generation product made from older components in a new form factor that looked very good.  But to blame Ray for the failures that he personally did not cause does not help solve the problems and blames the very person that will help to solve the problem in the future.  Take out your frustrations on the product and not the man.  Yes someone in manufacturing probably caused at least some if not all of the issues, but we do not know who that person is or who he works for.  You can't blame a car dealer for a manufacturer's engineering problem or an incompetent employee on the assembly line.  I'm sure that Ray is just as frustrated by this issue as everyone else is.

[Steve Gase]

Ray,

There is a lot of counterdicting information in the boards about the Technicolor strings and their problems.

Could you answer the following questions?

1) Are the TM1804 strings using a higher voltage, like the WS2811?

2) Will reducing intensity help avoid the issues?

I am just now getting ready to put up the 50 TM1804 strings, but if I can tweak something first to avoid problems, I'd like to do that now rather than destroying pixels and ruining the strings.

Any advice?

Thanks,
Steve

Dear Steve,
very sorry for the delay reply.
 
1: tm1804 node only consume about 0.3W,but the 12v ws2811 NODE consume about 0.6w,because these is a LDO on the pcb, that is why the 12v ws2811 technicolor met a heat problem, i will re-design the pcb , the new pcb will come out next week.
2: please do not let the tm1804 nodes work all the time,24hours a day,please only let it work in the night. and adjust the power to 11.2v around from 12v.
3: if the whold string not work, please cut off the first node to try again,because the first one already is easily burnt.
 thanks and best regards