About Ray using a higher current in LED.

[sittinguphigh]

Here is a question/answer from Ray about the new LED.

 
 Hello Ray

DLA has high discussions online about a problem that you changed over you had with the LED.
Going form 29ma LED to 56-60ma LED. This makes it impossible for DLA users to use 128 nodes on one line.
Can you verify if this is true?

I know the technicolor strings have this bright LED in them. DLA tested the light string. We can't use them with 128 nodes.
They use .67 watts per node.

You are not allowed to view links. Register or Login

 

 

Response

 Ray wu:
00:11 Nov 29,2013

 
you are right,but the led is only 3.2V/60ma(arount 0.2w),the LDO on the pcb takes about 0.5w power.
you just need to connect it to a bigger power supply , i supposed.
thanks and best regards
Ray
Me:
07:44 Nov 28,2013

[keitha43]

This confirms what we were saying. The current generation of nodes have similar specs to the Technicolor nodes and as Steve said according to his testing "A 100-count string will use 5.5 amps... which exceeds the 4 amp rating of the SSC and cat5 wire used for power delivery.  Power Injection at the end of the 100-count string will be necessary."

Or just use shorter strands and more ssc's or a Zeus if you want to use the cat 5 to deliver the power without injection and of course the total number of nodes a power supply can power will be fewer than the older style nodes.

[sittinguphigh]

This really makes it a mess. More problems.
You will need two Zeus to do a mega tree.
Or more wires. Means more potential problems.
The power consumption will double.
So it sounds like difference is the LED. Unless you can find a chip takes less power. Or a DLO want ever that is.
What does Ray mean bigger power supple?

[tbone321]

You can always get around that with power injection.  I really hope that in the future of the devices that RJ moves away from cat5 and RJ45 connections between the hubs and the nodes and goes with the terminal blocks that the Zeus uses.  4 conductor twisted pair speaker wire should povide enough signal protection and can be sized well above anything that cat5 can handle and carry much more current and removing the RJ45's from the hub outputs will prevent people from plugging in the wrong devices into the hub outputs and possibly damaging or destroying them.

[sittinguphigh]

You may have no choice. Unless the power source goes around all the hubs and controllers. The consumption doubles the price of doing this. Not good.

[arw01]

I like some sort of connector myself instead of terminal blocks exclusively.  Too easy to slide the wrong color wire or a stray bit of strands to touch their neighbor.  The waterproof pigtails are a little short, perhaps a little longer set of those would work, or a connection block that can be hard mounted to a case. don't know what they have in Roadie land out there for stage lighting since they have to break it down and re-assemble it several times a week I image there are some robust, idiot proof, connection schemes out there.

I know in general the board here tries to keep the costs out, but my time is worth more than an extra few dollars here and there to simplify my hobby in some other areas.

Alan who's going to need some commercial quality strings next year and some scheme to power it all..

[tbone321]

The advantage to terminal blocks is that you can then attach any connector to them that you want if you feel the need for connectors at all on the hub.  I like the flexability of making and changing the length of my cables at will and with the ever increasing current draw of these nodes, it is getting beyond the ability of CAT5 to be able to properly deal with it. 

[Steve Gase]

On one hand I feel vindicated that my numbers were confirmed, and also got the word out so that folks were not completely surprised... 

Still... what a mess.  There will be blown fuses and other problems.

Not blaming DLA for this, but reasons for hobbling the SSCs with 128 nodes have shifted... it should be fewer allowed nodes for these new strings.  If we introduce power injection (with fuses!), then maybe the limits should be lifted altogether so that I could run 500+ nodes from a SSC.

Zach has talked about a Technicolor v2 strings using 5050s... this will change the power profile yet again.

[taybrynn]

Moral of the story  - IMHO - use the approved ss nodes in the wiki -- spending more and getting hAlf the nodes per controller and all the issues -- no thanks ...

[sittinguphigh]

The good thing is we found out early about the problem because of the Technicolor strings. Now we aren't blind sided. We can work on it. Let every one know before buying them. All new RGB led strings TW1804 or WS2811 will have this LED in them.

[keitha43]

Moral of the story  - IMHO - use the approved ss nodes in the wiki -- spending more and getting hAlf the nodes per controller and all the issues -- no thanks ...

The problem is the nodes linked to in the WIKI also pull more MA than they used to. You can no longer drive 128 nodes via the cat5 cable without power injection. I would not be surprised if RJ announced a simple solution to power injection in the future. He is always 3 steps ahead.

[tbone321]

I don't know of a "simple" solution to adding power injection.  The existing simple solution was to use 12V and limit the node count to what the cable being used can handle to eliminate the need for power injection and all of the extra cabling it requires.  The only "simple" solution that I can come up with is up the current capacity of the SSC and DSC an amp or two and ditch the cat5 cable.  I will be doing some tests myself using 4 conductor twisted pair speaker wire to see how far it can go between the hub and the controller before it has signal issues.  If I can get 100 feet out of it, that far is more than anything that I would need and the current carrying capability far exceeds anything that cat5 can deliver.

[taybrynn]

Why is this new led which is half as efficient as the old the new standard?   Makes no sense.

[tbone321]

Three possible reasons, cost, availability, and brightness.  I am not sure if the new LED' are less efficient but if so, I bet that they cost less or the cost of the ones he was using has gone up.  It could also be that the ones he was using are no longer available in the quantity he is needing them at the current price.  They could also be brighter and of course, "brighter is always better ".  Even at the same effeciency, brighter LED's use more current. 

[taybrynn]

IMHO the current nodes were already plenty bright... and now were looking at double the node cost and twice the hardware to run them ... Sigh