Flickering stars, when they shouldn't

cdnbob

New elf
Joined
Jul 26, 2014
Messages
33
Location
Ottawa
I have 9 of these: [url=http://www.aliexpress.com/store/product/DC12V-WS2811-star-type-pixel-module-6pcs-5050-SMD-rgb-led-inside-1-44W-IP68-ONE/701799_1602954057.html]http://www.aliexpress.com/store/product/DC12V-WS2811-star-type-pixel-module-6pcs-5050-SMD-rgb-led-inside-1-44W-IP68-ONE/701799_1602954057.html[/url] in a series - first 6 are back to back in line, then an 8ft/2.5m separation to 3 more back to back. The last 3 are flickering, mostly the last one. When set to white they seem to flicker to red mostly. This is more prevalent if some of the first 6 are on, especially when set to white.

I think this is a power issue but I'm confused since the length doesn't seem too long and there's only 9 of them - total 13 watts, so 1.08 amps at 12v.

I run a 24v feed to the first one then use a dc-to-dc converter to drop it down to 12v. The first one is about 30ft/10m from the controller/power source. I measure 12.1v out of the converter at the first star. Converter has a max draw of 5 amps.

Is this a power problem or could it be something else. What mistake might I be making in my power calculations?

Bob
 

battle79

Full time elf
Joined
Dec 8, 2012
Messages
465
Location
Chirnside Park, Victoria, AU
The issue is not Amps, but voltage.

By the time your 12V gets to the last three stars it has gone through all the cabling to get there and is likely only 10V or even lower. This effects the ability of the chip to decipher the data signal, as it uses ground as a reference to compare data to.

Voltage drop is caused by the resistance of the cable as well as the resistance in the PCB in the star (assuming it has in and out terminals for power). This can be overcome by running a power feed from the same power supply to the last three stars to inject power, or by increasing the size of the cable to reduce the resistance.

Generally it is more practical to just run another cable.

Regards,
Rowan
 

mborg10

Michael Borg
Joined
Dec 28, 2011
Messages
1,446
Location
claremont meadows
I can't click on your url but if it is the same stars as I have I found they don't send their data too far. I had issues after 1m of extra cable. My 94 stars are all linked together and I have to inject about every 20 stars.
 

cdnbob

New elf
Joined
Jul 26, 2014
Messages
33
Location
Ottawa
Data range. I was stuck on the voltage side of the question, but the extension between the 6th and 7th could be too far for the data signal, if the 6th star doesn't send the data well. Maybe a null pixel in the middle would help.

I'll double check the voltage coming off the last star, but I'll be surprised if that's the issue. Indoor testing before they went up showed good voltage at the start of the last 3 (end of the extension).

Michael, are you injecting data every 20 stars, or just power? Wouldn't injecting data mean you use a separate output on the controller for each injection point?

Bob
 

mborg10

Michael Borg
Joined
Dec 28, 2011
Messages
1,446
Location
claremont meadows
No just injecting power date is from one output but all my stats are connected directly to each other. I wanted to connect the last star to a candy cane and the distance of 3m was too far to get a signal. A null pixel at 1m saw the canes work and flicker sometimes so I just used a different controller for the candy canes
 

logandc99

Dedicated elf
Joined
Jul 13, 2012
Messages
1,153
My stars are connected directly to each other also so the only distance between them is the total length of the two pigtails and they work fine. However, whilst I can get them work if they get data following after my strip light, I cannot get strip light to work after the stars no matter what distance I have and how many strip lights I tried so there is something peculiar about these stars that makes them not want to play nice with other led types. Like Michael, I solved my problem by rearranging my setup so the on the left side of my roof, the stars are connected at the end of the strip light (which works fine) but on the right side of the roof the stars have their own output from the P2 with nothing else connected after them.
 

mborg10

Michael Borg
Joined
Dec 28, 2011
Messages
1,446
Location
claremont meadows
When I originally went to order these stars they were only available in dc24v and I asked Ray if he could do dc12v as a special order. After my order he then listed dc12v on his site. I wonder if they would have worked better as dc24v
 

fasteddy

I have C.L.A.P
Global moderator
Joined
Apr 26, 2010
Messages
6,648
Location
Albion Park NSW
These would have been better as a 24v light due to the additional current requirements of these.

I wonder when voltage is checked that people are checking the voltage at the last pixel when they have their lights set to white which is the maximum load
 

sck1ng

Apprentice elf
Joined
Oct 23, 2013
Messages
77
I had similar problems with mine cdnbob.. I eventually gave up on using the LEDs in the stars.. I opened the Stars up and removed the LED module that was in there and glued 3 pixels in each.. Not as bright as the original module but have had no flickering issues..
 

sck1ng

Apprentice elf
Joined
Oct 23, 2013
Messages
77
741b8cb5a911e7f75abbd1ccde7c16ba.jpg


This is what mine looked like with the three pixels in them. If I had more time and had to do them again, I would have put 5 pixels in there.

https://vimeo.com/115034254
 

logandc99

Dedicated elf
Joined
Jul 13, 2012
Messages
1,153
Fasteddy said:
These would have been better as a 24v light due to the additional current requirements of these.

I wonder when voltage is checked that people are checking the voltage at the last pixel when they have their lights set to white which is the maximum load

I'm not sure if voltage was the problem in the situation I was having. I had power injected into the start of the stars, power between the last star and strip and power at the end of the strip. It was as if the data coming off the stars couldn't be read properly by the strip light.
Strip going into stars was fine but stars into strip was a no go.
 

sck1ng

Apprentice elf
Joined
Oct 23, 2013
Messages
77
logandc99 said:
I'm not sure if voltage was the problem in the situation I was having. I had power injected into the start of the stars, power between the last star and strip and power at the end of the strip. It was as if the data coming off the stars couldn't be read properly by the strip light.
Strip going into stars was fine but stars into strip was a no go.


I agree logan.. This was definitely a data issue.
 

fasteddy

I have C.L.A.P
Global moderator
Joined
Apr 26, 2010
Messages
6,648
Location
Albion Park NSW
logandc99 said:
Fasteddy said:
These would have been better as a 24v light due to the additional current requirements of these.

I wonder when voltage is checked that people are checking the voltage at the last pixel when they have their lights set to white which is the maximum load

I'm not sure if voltage was the problem in the situation I was having. I had power injected into the start of the stars, power between the last star and strip and power at the end of the strip. It was as if the data coming off the stars couldn't be read properly by the strip light.
Strip going into stars was fine but stars into strip was a no go.

The issue with many chinese vendors including Ray is that they source many things from lighting markets and factory seconds to keep price down and this means that quality can vary as well as the chips used may be coming from different clone factories, so a 2811 may in fact not be compatable with another 2811 that is made by another manufacturer as some are finding out.
 

cdnbob

New elf
Joined
Jul 26, 2014
Messages
33
Location
Ottawa
I haven't checked the voltage at the last star when the stars are all on full white, but I've been running them with a sequence that only lights one or two at a time, and the last 3, particularly the 7th and 9th, will flicker or come on with the wrong colour even when there are no other lights on in the set.

I do have one other of these stars in front of a 150 pixel strip, and the strip works fine. There's no gap between them in this case, which makes me think it's that gap between the 6th and 7th in the main set that's the issue. I'll try to insert a null pixel in the middle of that gap after Christmas.
 

cdnbob

New elf
Joined
Jul 26, 2014
Messages
33
Location
Ottawa
So I inserted a null pixel into the extension between the 6th and 7th stars. But I goofed. I forgot the the pigtail polarity of the stars is the reverse of the Technicolor pixel strings. (Stars use GDV+ and the Technicolor strings use V+DG, reading clockwise from the alignment tab.) So the null pixel was inserted in reverse polarity, and it is now dead.

But so is the first star downline from the dead null pixel. I don't understand why. When I tested the stars and first discovered the reversed polarity in the pigtails, I did give the stars the wrong polarity, and they survived that mistake just fine. I checked the extension, and I did not reverse the polarity end-to-end when I inserted the null pixel. Only the null pixel should have seen the reversed polarity. The downline stars would have been fed correct polarity.

Any ideas what happened to kill that star?

Bob
 
Top