timon wrote:I'll be getting test strings soon both cased and uncased.Ground goes up the center wire, data and ground go up the outside. Much less of a chance for any +12V and data shorts.
You mean data and power go up the outside, I think.
Board is thicker so less likely to crack. It's also a bit larger so there is a little more board area to carry heat off the regulator.
Board was designed to reduce EMR.
Any additional questions?
gavspac wrote:Hiya,just writing to let you add my sorry technicolor pixel story.
I bought 5 strings of 100 and made an led heart which was to go in the street in Brighton UK.
The heart is one long string of 440 pixels with power injected at the beginning and end of each string of 100 pixels. I built the heart in my workshop and amazingly every pixel seemed to work.
After I ordered the pixels I found the technicolor quality thread but it was too late to turn back so with some trepidation I installed them!The installation was interactive and lit up when two people held hands. When not in use for a while it cycled through various patterns.
It worked fine for a week but then it failed.
About 90 pixels in, the string stopped working and after that the pixels were static and random.
When I got to look at it, one of the pixels had overheated and the casing was melted/bubbled. It had damaged 4 pixels afterwards and the one before. I replaced them and found that there was another pixel damaged 15 or so pixels further on.
I bypassed the first two strings and tested the other strings and realised that there were damaged pixels all the way along the 440 pixel string. Damn! Too many to fix in situ with the ladder we had and the wind conditions.
My guess is that when the wire burnt through, the 12v touched the signal wire and went all the way along the string taking some of the pixels out along its length.
Does that seem likely? I'll know more when I take the heart down in a couple of weeks.
Thats my story anyway and here I am broken hearted. Its all a learning experience I guess and I realise if you buy cheap lights you have to be prepared for this kind of thing.
Is there anything I should do in the future to avoid this kind of thing happening?
I thought I should perhaps have had strings on separate circuits so that if some failed there would be a backup. Wonder if fuses would've helped.
I used this power supply: http://www.stadiumpower.co.uk/products/led-drivers/active-pfc-10-300w/hsf300
Turned the voltage down to 11v. Wish I'd turned it down more now!
Nice idea.
Were the pixels open-air, or was there some sort of clear enclosure around them that I can't see in the photos?
Timon, thank you for working to salvage the good ideas of the Technicolor pixels. It's unfortunate that these supposedly better designed pixels seem to be failing especially badly, with apparently both heat/power and water based failure modes. Your redesign faces somewhat of an uphill battle, like the coach in one of those movies who takes a bottom of the league team and tries for the championship. I wish you luck and I am hoping for a good outcome which leads to a group buy, in which I am interested.
I hope the final design will be 12V compatible. Not all supplies can be turned down to 9V, and there's something strange about using a 12V supply which goes to a 12V to 9V regulated DC-DC power supply, which then goes to the 5v regulators of technicolor pixels.
Is there any chance that the revised technicolor pixel will still have an option for using the 1829 chip? There are two reasons that could be interesting. One that matters to me but not everybody is that it has a fast PWM refresh rate, more than 10 times as fast as the notably slow 2811's. The other is that it has a programmable current of 10-41ma per color. That ability would need to be carefully used (and controllers would need to take it into account so as not to accidentally set the higest power), but it could be useful to tune the power dissipation for different situations.
So last night I had 2 more technicolor pixels fail making somewhere between 25-30 out of 200. I replaced them today to get the show back up. Well tonight I had 3 more fail. In the process of looking for repair pixels I realized I have used every single spare pixel I had on these things. The show has ran great this year other than these bulbs. I have 2 window frames I can't repair now so they are down for the year.
In the interest of understanding all of the failures that people are experiencing we seem to be concentrating on the Pixels themselves. Is it possible that a power supply, controller, or combination of both is contributing to the problem? Can a power supply or controller be "over-driving" if that is the correct term, these and other pixels? I have been running 1500 Technicolor Pixels and 12 2812B strips through varying temperatures and weather conditions(Ice, Heavy Rain, Temps from 27-73 Degrees) with no failures. I'm curious which power supplies and controllers people are using with these pixels, and if that could be a contributing factor? I am using Meanwell S-350-12 power supplies and J1sys P12S and P2 controllers.



