Results 1 to 7 of 7

Thread: Multiple power supplies questions???

  1. #1
    Join Date
    Oct 2018
    Location
    Sherwood Park, Alberta, Canada
    Posts
    44
    Post Thanks / Like

    Default Multiple power supplies questions???

    I have been wondering about this, If I need to use say two different power supplies for a mega tree, do I connect the voltage negatives together???

    I'm asking because I read someplace that the data line needs to be connected to the negative, which doesn't make sense to me. Wouldn't the data line ground out???

  2. #2
    Join Date
    Jan 2015
    Location
    Folsom, CA, USA
    Posts
    1,732
    Post Thanks / Like

    Default Re: Multiple power supplies questions???

    You need to connect Data- along with Data+ throughout the data line, so, inherently it connects Power1 V- to Power2 V-. However, I never connect either V- or V+ of one power supply to another.
    In Lights Therapy

  3. #3
    Join Date
    Oct 2018
    Location
    Sherwood Park, Alberta, Canada
    Posts
    44
    Post Thanks / Like

    Default Re: Multiple power supplies questions???

    Thanks mrGrumpy,

    I understand that the data out from the pixel has to be connected to the data in of another pixel. I will keep the power supplies separate then. Thanks again.

  4. #4
    Join Date
    Jun 2011
    Posts
    1,876
    Post Thanks / Like

    Default Re: Multiple power supplies questions???

    Quote Originally Posted by mrGrumpy View Post
    You need to connect Data- along with Data+ throughout the data line, so, inherently it connects Power1 V- to Power2 V-. However, I never connect either V- or V+ of one power supply to another.
    You do need to connect the power supply 0V (common) together.. Not the V+. The data is not differential so it will require common and Data.

  5. #5
    Join Date
    Dec 2012
    Location
    Hudson MA
    Posts
    3,213
    Post Thanks / Like

    Default Re: Multiple power supplies questions???

    Data and v- from the output pixel to data and v- on the destination pixel plus psu2 v- an d psu2 v+to the destination pixel. You do NOT need to tie v- at the psu's


    2018 - Moving and going to visit my Daughter in New Zealand. Most likely I will be dark or nearly dark, Some static stuff that is simple to put up.

  6. #6
    Join Date
    Jun 2011
    Posts
    1,876
    Post Thanks / Like

    Default Re: Multiple power supplies questions???

    Quote Originally Posted by MartinMueller2003 View Post
    Data and v- from the output pixel to data and v- on the destination pixel plus psu2 v- an d psu2 v+to the destination pixel. You do NOT need to tie v- at the psu's
    Very true, you do not NEED to connect the 0V (common) AT the power supplies. They can connect pretty much any point along the common path.. For example, my Mega tree has 4 power supplies. When I inject power via "Ts" along the way, that is where the commons tie together. Works fine.

  7. Thanks rmcgee5 thanked for this post
  8. #7
    Join Date
    Nov 2014
    Location
    Fox Valley Area, Wisconsin
    Posts
    122
    Post Thanks / Like

    Default Re: Multiple power supplies questions???

    I would like to pass along a thought.

    Digital logic relies on the zero axis reference. This ZERO axis is the NEG (V-) of the power supply. Our displays work on 5v TTL logic.

    Unless the power supplies are bonded V- to ground (a bad idea), completing the NEG bonding to multiple power supplies ensures a clean ZERO axis.

    As long as the power supplies are bonded together via the NEG (V-), you will ensure a clean floor for the TTL signals to happily dance.

    Another thought: ensure power supplies are NOT bonded to ground. Test this with a meter. Test both POS (V+) and NEG (V-) to the metal case of the PS in question - it should show anything above ZERO (i.e. no continuity). There can be some resistance (MOV devices) but not ZERO.

    My research showed many power supplies and suppliers show ADDING bonding to the NEG (V-) to ground. There are applications where this is a good thing. For our use, this can be as simple as an annoyance (corrupt data) or as devastating as destroyed pixels.

    GFCI protection does NOT prevent this. Most commercial GFCI devices trip around 4-5ma. This simply means that AC power leaking: 120v @ .004a = .48 watts. That does not sound like much, does it. 3ma @ 120v across your chest can kill you.

    The misnomer about GFCI is it protects everything plugged into it - NOT TRUE. It LIMITS the amount of current that can leak before it acts and opens the circuit.

    The best representation: a GFCI acts like a balancing beam. 120v on one side and Neutral on the other. As long as it remains in balance, it does nothing. As soon as it goes out of balance (set trip level) it acts and opens the circuit.

    What this says is SOME current can leak out. A true 1 to 1 ratio, zero leakage, is possible, however, not practical for consumers. Most people would be spending more time resetting the GFCI than using it.

    Another misnomer: A GFCI device requires a GROUND to operate correctly. NOT TRUE. That GREEN WIRE is a EQUIPMENT GROUND. It does nothing for the GFCI device. The device senses the HOT and NEUTRAL balance only. The 'TEST BUTTON' will fail to operate if the device does not have a GROUND, that is about it. The device will do its job in the absence of a Green Ground Wire or a broken ground path. So if your TEST button does not trip, that does not mean the device is defective. Additional tests by an electrician can confirm this.

    Defective GFCI devices are usually those who will not "SET" - that pesky RESET button does not work. Yeah, time for a new one!

    It is also a note that all GFCI devices are not built the same. Hospital grade GFCI devices must trip between 1.5ma to less than 3ma (back when I was testing them). The cost for these devices is significantly higher too.

    So, simply put - it is not a good idea to bond the NEG (V-) to ground and expect your GFCI outlet to prevent any controller problems.

    If you have a "DATA GHOST" or corruption that appears after a heavy rain or melting snow, I start looking @ the power supplies - I help others and found some of these were bonded to ground. Since the 120v plug is usually on the OUTSIDE of the enclosure, not in a weather proof case, in close proximity to the controller, the leakage is not enough to trip a GFCI - but the leakage is enough to corrupt the DC power. It is a domino effect: this in turn can cause data corruption. Removing the NEG (V-) to ground solves this MOST of the time. Water in connectors is #2 on the investigative list of things to look for.

    Last thought (history lesson) Ma Bell bonded the +48v to GROUND. The actual RING signal is 100v @ 10Hz, all the bells sounded at the same rate.

    The reason: 48v is high enough to cause death. Bonding the POSITIVE to ground alleviated liability and made it safer for workers too.

    Today this is the current practice for all land lines in the USA.

    Eddie
    -Eddie

    The missus wants to ride!

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •