ghosting issue with pixels

Jimbo

Full time elf
Joined
Feb 27, 2014
Messages
144
Hi, I am having an issue with ghosting on my mega tree and spinners.

On the mega tree (which is 2d, 12 x 4m 2118 strip with 7 power injection points) when the power injection is in on the first strip or the 3rd strip it ghosts and plays the wrong colours. However if i unplug these the tree works fine, but the first four strip don't have enough power to show correctly.

I have 5m cable running dmx and the negative from the p2.

I am guessing it could be 1 of 2 things.

1. The cheap power supplies from ray wu.

2. The cable length for the dmx run is not thick enough.

Anyone else have this issue?
 

lizardking

IT IS STILL ALL BENS FAULT
Joined
Dec 26, 2014
Messages
416
Location
Gold Coast
jimbo i had the same issue with my mega tree i had too many power injection points check to make sure your negatives are not in a loop scenario
 

lithgowlights

Dedicated elf
Joined
May 6, 2010
Messages
1,023
2811 or 2118? 2811's are an absolute pest - they will either work perfectly or will do nothing but play up, and I had the latter here. I needed to run power to the strings with normal alarm cable, but the data (and reference ground) needed to be sent over a cat5 twisted pair. Anything longer than 1 meter refused to work for me, but others have it working over alarm cable at 10+ meters like my 2801's here.
 

Jimbo

Full time elf
Joined
Feb 27, 2014
Messages
144
2811's.

It's weird. Even if I have just one power injection plugged in at the end and plug either the first 2 power injection it does it.

The spinners also work fine then ghost and work fine.

I have seven power injections off 1 power supply for the mega tree. I run it at 11 volts so only uses 80% capacity.

Not sure what Is meant by loop. How do i check?
 

i13

Dedicated elf
Joined
Jul 5, 2013
Messages
1,172
It may not be the problem but 5 metres is probably pushing the limit for the data line. Thickening it may not help as doing so increases the capacitance in the wire and that can lower the signal quality. The only solution would be a null pixel if this is a part of the problem. Don't double the data wire.

I'm not really sure what the loop is either.

Also when it is suggested to only use cheap power supplies to 80% of their capacity, that is referring to the current/wattage, not the voltage. Adding up the current with 55.5mA per pixel (worst case scenario and on full white), you have a total of 26.6 amps in the tree which is 319.7 watts so you'd be running the power supply at 91.3% capacity. If I were you, I'd turn it up to 12V and test it without turning everything on completely white.
In reality, pixels often don't draw that much current so full white may not be a problem. It would be easy to test how much current each length of strip draws.
 
Top