Power injecting directly to a Pixie2D

[randyf]

I'm in the process of re-purposing a Pixie2D to handle pixel lights on my patio (I realized last year that I will want more where it was deployed, so acquired different devices).  In it's new location, it will be expected to drive 2 strings or 100 globe pixels, and there is little chance that it will be able to handle 2 strings.  So I intend to inject power from a different power supply.  My original thought was just to do that inline, but today I was thinking that maybe I can actually feed the box with an appropriate amount of power.  In theory, this is quite doable, but there may be some gotcha I haven't thought of.

What I intend to do is to unsolder the 12V+ output, and connect them to a cable that exits the box and can be connected to an external PS (with the ground soldered to the ground feed at the output - note, depending on responses, I might do two lines to each separate output).   As I see this, it should be no different than injecting in the output line, but removes the need for an extra set of pixel connectors (and also prevents higher power pixel lines from being used without the extra PS).

Does this seem reasonable, or maybe one of the LOR folks (e.g. Matt) might have some concerns (besides a possible warranty issue that I'm willing to accept)?

[Jimehc]

Install a Power Injection "T" after first sting of Pixels on each port

Edited July 30 by Jimehc

[randyf]

17 hours ago, Jimehc said:

Install a Power Injection "T" after first sting of Pixels on each port

That was the original plan ("My original thought was just to do that inline").  That would require two pigtails, two tees, and then modifying the two tees (need to pull the DC+ pin so that it doesn't feed injected power back to the controller).  But it had occurred to me that I should be able to accomplish the same by lifting the DC+ line at the string outputs and connecting one injection cable directly to the controller.  But there may be issues I either haven't considered or may be unaware of (e.g. need to connect a jumper ground between the outputs).  Also note that for this particular use, I need to inject immediately out of the box, as the controller won't be able to handle more than one string of 50 pixels (the strings I will be using draw more than the box is designed for).

[TheDucks]

1 hour ago, randyf said:

That was the original plan ("My original thought was just to do that inline").  That would require two pigtails, two tees, and then modifying the two tees (need to pull the DC+ pin so that it doesn't feed injected power back to the controller).  But it had occurred to me that I should be able to accomplish the same by lifting the DC+ line at the string outputs and connecting one injection cable directly to the controller.  But there may be issues I either haven't considered or may be unaware of (e.g. need to connect a jumper ground between the outputs).  Also note that for this particular use, I need to inject immediately out of the box, as the controller won't be able to handle more than one string of 50 pixels (the strings I will be using draw more than the box is designed for).

Feed from the far end. Besides Voltage Drop, wire capacity is the other issue for PI. Since these strings are designed for that draw, should be no prob

If you don't mind voiding warranty, just snip the PLUS ONLY  between strings (Data NEEDS that common ground intact)

[randyf]

7 minutes ago, TheDucks said:

Feed from the far end. Besides Voltage Drop, wire capacity is the other issue for PI. Since these strings are designed for that draw, should be no prob

If you don't mind voiding warranty, just snip the PLUS ONLY  between strings (Data NEEDS that common ground intact)

While also feasible to feed from the ends (will be implemented as 4 parallel strings with the controller in the middle), it requires two injection points, pigtails, and longer cable feeds from the PS.  Feeding to the box (in theory) only requires one short run from the PS and no pigtails (not counting the PS connection, but I need that anyway - and yes, will only lift the DC+ line, but may also bond the ground lines for both outputs so only one DCin line is needed).

 

Edit: Yes, there will also need to be a jumper between the end of the first and beginning of the second strings, but that has already been considered in the entire run, but is also in a consideration of injecting at the beginning and end of the strings

Edited July 31 by randyf

[randyf]

Update: I decided to try injecting directly to the box.  Worst case I kill the box and a string of lights; but then my injection need goes away.  Good news it looks to have worked just fine.

The technique was as has been mentioned above:  Unsolder the 2  "+" wires on the output; bond the two "-" pins (a.k.a. "ground" - and there are two through holes to make it easier); then run the "+" and "-" lines from an external power supply to the lifted wires and bonded point respectively.  I also verified that this really was working, by disconnecting the external PS and see that the strings are not powered.

 

As a supplemental note, this box is intended to run exterior lighting, but is placed in a location where it is protected from the weather.  However, it is still effectively outside, and since it can get quite hot in Arizona in the summer, I've chosen to also leave the cover off so that there is some possibility of keeping it from overheating.  Again, if this fails, I'll need to find a way to get the controller (possibly a new one) in a cooler location and run extensions (PS will still likely need to be close, but since it is way overpowered for the use, long large-gauge wires might be fine).  I'm not sure if anyone is interested in the long-term stability.