Pixel wires - the downright ugly truth ......

[AussiePhil]

To be fair, this data is taken from two orders placed this year and not from earlier years.

Both sets are two 50 count strings solder joined in the middle with 140mm space to form 100 cnt strings

Set 1 - 18AWG 2818 bullet pixels
Set 2 - 20AWG 2811 bullet pixels

As the positive and negative wiring a a straight connection past each pixel you can measure the loop resistance by shorting out the far end.
The lead resistance was removed from the multimeter reading by using relative readings.

Resistance calculations are based on the data from this site: https://www.omnicalculator.com/physics/wire-resistance

Pixel WG	18	20
Distance Used This is the loop length of a 100 pixels at 4" spacing	20M	20M
Calculated Res - Alu	1.985	3.214
Calculated Res - Copper	1.281	2.074
Measured Resistance	1.995	1.91

Now like me you have done a double take on the measured reading and said wtf, how can 20awg be better than 18awg. Seems one is copper and the other is aluminium wires.

So not only is your resistance and hence voltage drop dependent on the wire gauge it is also dependent on the wire TYPE.

the ugly truth may well be that the 18awg pixels we buy at a small premium may be no better than a quality 20awg when considering voltage drop.

We should all ask what type of wire is being used as well as the wire size
STOP using wire gauge .... ASK for the wire strands and individual wire size

Examples and some photos.

Pixel 18 gauge = 34 strands of 0.10mm wire
Pixel 20 gauge = 21 strands of 0.10 wire

7.5A Fig8 in aus = 24 x 0.20
10A Fig8 - 32 x 020

so some photos, all taken under the microscope.

Pixel 18




Pixel 20


Pixel 20 versus Pixel 18



Pixel 18 v Pixel 20 at a strand level, they are the same strand size just different counts (and types it seems)


The black is an artifact only they are all nice shiny silver

An here is Pixel 18 compared to some 24x020


Cheers
Phil

 

[bpratt]

I've been wondering that myself, aluminium or copper clad aluminium or better pure copper wires.

20 gauge pure copper would certainly be better than 18 gauge aluminium strands in my eyes, but I haven't see any 18 gauge pure copper myself.

 

[David_AVD]

Copper clad aluminium can be harder to solder properly too.

 

[MichaelF5]

can I ask what the OD of the cable is if you have some calipers

 

[AAH]

CCA has a higher thermal coefficient of resistance meaning that the longer it's running at a given current the more resistive it will become. It would actually be nice to see the same resistance measurements after a few minutes of running 100% white through the pixels which in theory should be 5.5A or about 3 times the current capacity of what the 1.91ohm/20m wire works out to. According to my calculations the 20AWG wire is measuring the same as about AWG24-26 or about 1/3 of what it should be. The creative marketing folks in china could well be specifying the total wire size of the 3 conductors to give the wire size rather the single conductor.
The dodgy undersized wire sizes used in pixels is 1 of the reasons I'm strongly against 12V WS2811 pixel nodes. It's not uncommon to see people with strings of several hundred pixels without power injection. A string of 200 pixels equates to 11A and when that's being powered through wire rated for 2A you are asking for trouble.

 

[MichaelF5]

Yes, I'd like to see a bit more Data on these strings, with a bit of thought we could come together with a unified request for Ray to create something that benefits everyone, hearing some of the tests Phil's been running I feel some of the calculators out there are wrong and we need fresh data

 

[AussiePhil]

Yes, I'd like to see a bit more Data on these strings, with a bit of thought we could come together with a unified request for Ray to create something that benefits everyone,

Ray is just the middle man, he buys in from other factories as needed, plenty of vendors out there that we can investigate then ask Ray to source if a single point is desired.

hearing some of the tests Phil's been running I feel some of the calculators out there are wrong and we need fresh data

I suspect the calculators are essentially correct but maybe assuming copper rather CCA etc. We the end users have been blinded by using the quoted size rather than empirically measured wire dimensions, we see 18awg listed for pixels, plug those numbers into the calculators that spit out a genuine answer for that actual size but the wire used is nothing like that size in real life..... human nature kicks in with "thats what the spec says so must be true"

If i can generate some time today amongst yard duties we will get some more testing done.

 

[MichaelF5]

I'd very much appreciate it Phil I'm hoping to everything get a few projects out of the way and then start some testing myself, on a related note, has anyone attempted their own waterproofing Jig etc?