Is having more that 2Amps of power supplys really a problem ??

Discussion in 'C-Bus Wired Hardware' started by Charlie Crackle, May 18, 2006.

  1. Charlie Crackle

    Charlie Crackle

    Joined:
    Aug 3, 2004
    Messages:
    815
    Likes Received:
    8
    Location:
    Melbourne
    Is having more that 2Amps of power supplys really a problem ??

    Customer in the country and has 3 phase power. The power supply company regulary loose a phase.

    The lighting is split across all 3 phases. The total load on the cbus from input units is 1.4 Amps. output units power supplies are spread across 3 phases 600 ma per phase.

    The problem is when the power supply company loose a phase I loose 600Ma of power supply putting the power supplied < power required which makes cbus do wierd things. (some switches work and some don't)

    I was going to increase the power supplies per phase to 800ma that way if a phase is lost there is still plent of power for cbus.

    Problem here is when all 3 phases are on cbus will have 2400ma which is greater than the MAX cbus power supplies of 2000ma.

    I assume the issue here is the short circuit current ?? I assume you could have 10000ma of power supplies as long as there were no faults they would still only supply the 1400ma the input units draw.

    My alternative solution is putting a phase detection drop out relay and get it to power 400ma of cbus power supply when it detects a phase dropout. only problem here is which phase to wire this extra power to....:)

    Charles
     
    Charlie Crackle, May 18, 2006
    #1
  2. Charlie Crackle

    JohnC

    Joined:
    Apr 6, 2005
    Messages:
    554
    Likes Received:
    1
    Location:
    Sydney
    Hmmm... interesting question !

    The 2A limit is to do with the maximum current rating of the Cat5 cable. There will be no adverse "C-Bus effects" if you have too much current potential, but you are correct in assuming it is a limit in short-circuit conditions (of the C-Bus Network).

    I am sure you've considered completely separating ALL C-Bus side from the Load side (which is spread across 3 phases). Then, all you need is that magical device that picks which phase is still alive, and powers that (very small current) for the "C-Bus Side" of ALL the devices. I reckon this is a better way of wiring it anyway, as any short circuit / RCD fault on the load side (tripping a breaker) will also not affect the C-Bus Power that is available.

    The added advantage of this is that you don't need as many C-Bus units with P/Supplies in them (=cheaper), and also don't need to worry about over-current on the Network Cable.

    I'll leave it to the experts to tell you what device will do that job - I don't think a simple phase drop-out relay is sufficent, you need something smarter than that :)

    I reckon the correct technical term would be a "connect this 240V load to the best phase" device

    John
     
    JohnC, May 18, 2006
    #2
  3. Charlie Crackle

    Newman

    Joined:
    Aug 3, 2004
    Messages:
    2,203
    Likes Received:
    0
    Location:
    Adelaide, South Australia
    That's only part of the story. There are 2 factors that limit the amount of power supplies you can have on a C-Bus network:
    • The short-circuit current rating of the cable.
    • The AC impedance of the network at communication frequencies
    If you have more power supplies than necessary then there's the risk of damage to the insulation on the conductors in the cable when a short occurs.

    As you add more power supplies you are also reducing the AC impedance of the network. This affects the reliability of communications. Pushed too far the network might not communicate at all. Power supplies have lower AC impedance than a key unit for example. After all, they're the ones generating the power in the first place. I don't have the data to hand to give you exact figures.

    This is why you can't just hook up a DC power supply or battery to C-Bus - the communication pulses get squashed by the low AC impedance those sorts of power supplies have.

    If you wired all the power supply sides of the output units to a common phase then you could use a UPS for that circuit?

    You could probably do a poor-man's backup using 3 drop-out relays and 3 power supplies, such that if one phase drops out the power supply kicks in that's connected to a different phase. Of course if you lost 2 phases then you'd be in a bit more trouble!
     
    Last edited by a moderator: May 18, 2006
    Newman, May 18, 2006
    #3
  4. Charlie Crackle

    JohnC

    Joined:
    Apr 6, 2005
    Messages:
    554
    Likes Received:
    1
    Location:
    Sydney
    Thanks Newman - I hadn't considered the Impedence side of things...

    That's a good idea re: the UPS !
     
    JohnC, May 18, 2006
    #4
  5. Charlie Crackle

    Darpa

    Joined:
    Apr 30, 2006
    Messages:
    426
    Likes Received:
    0
    Location:
    Australia
    The only problem with using a UPS, is that anything worth under $1000 is only going to be able to run things for less than an hour (rough guesstimate, lol), and will definately only be single-phase-powered, which then begs the question...

    WHICH BLOODY PHASE DO I CONNECT IT TO????? :p

    Oh, the joys of problem solving... half the time you end up with more questions and problems than you started with, lol


    Darpa
     
    Darpa, May 18, 2006
    #5
  6. Charlie Crackle

    RossW

    Joined:
    Oct 10, 2005
    Messages:
    118
    Likes Received:
    0
    If it's an impeadance thing, as much as anything else, why not just stick a decent inductor in series with the output of the supply?

    Possible other suggestions would be to use a (not terribly complex) current-limiting series element on the end of (n) DC supplies.

    Imagine, in effect, three transformers (one on each phase, or a 3-phase transformer if you have access to one), into either 3 full-wave bridge rectifiers (or a 3-phase 6-diode bridge) to provide you with a decent cross-phase supply. Each phase would be comfortably able to do the entire job by itself, when parallelled each would be just idling along, and the current-limit OUTPUT stage would protect the cables. High-impeadance DC supplies are not difficult, they're more "the norm" than the exception in telco applications.

    (Suggestion: it cost me $200 for dual, redundant 48V telco supplies including distribution breakers and battery backup, each supply good for 300W (WAY more than you need here!) - for 2A output, you could just about get away with a power-OP-AMP to regulate the supply down to 32V and current-limit, and probably still be cheaper than a commercial DIN-rail supply!)
     
    RossW, May 18, 2006
    #6
  7. Charlie Crackle

    Charlie Crackle

    Joined:
    Aug 3, 2004
    Messages:
    815
    Likes Received:
    8
    Location:
    Melbourne
    Newman,


    I contemplated this, but I remembered that dimmers require the power supply to be on the same phase as the dimmer...

    Wouldn't powering the power supply(part on a dimmer) on a UPS (which could be out of phase buy who knows how many degrees (when it goes on battery) , cause problems with the dimmers.

    If I hand my time gain. I would have used relays and dimmers with no power supples and then just used DIN power supplies powered on a UPS.

    I know it would be a no no, all warrantees would be void and you would have to work out how to open with out destroying. Is it just a matter of removing 1 component on a relay/dimmer that would disable the power supply in them?


    Charles
     
    Last edited by a moderator: May 19, 2006
    Charlie Crackle, May 18, 2006
    #7
  8. Charlie Crackle

    JohnC

    Joined:
    Apr 6, 2005
    Messages:
    554
    Likes Received:
    1
    Location:
    Sydney
    Interesting - I seem to remember asking on this forum why the dimmers weren't internally wired (ie: why separate A+N for control and load sides), beacause it's a waste of time having to install those bridging cables on every bloody dimmer. I was told that the "separation" was a design feature so that you could separately power your C-Bus side !

    I guess that is actually the "proper" solution. Keep all the C-Bus on 1 phase, and the loads split up however you like across all 3 phases. There must be a way, so who's the clever bod who can solve this ?
     
    JohnC, May 19, 2006
    #8
  9. Charlie Crackle

    Newman

    Joined:
    Aug 3, 2004
    Messages:
    2,203
    Likes Received:
    0
    Location:
    Adelaide, South Australia
    For Dimmers, yes, this will cause problems. It is not a problem for relays however. This is because, for the current DIN dimmers, the zero crossing detection signal comes from the power supply. Consequently, if you disable the internal power supply you will lose the zero crossing detection signal for the dimmer.
     
    Newman, May 19, 2006
    #9
  10. Charlie Crackle

    Ross

    Joined:
    Aug 3, 2004
    Messages:
    174
    Likes Received:
    0
    Which therefore suggests that if you are designing a large install with multiple phases, due diligence is required before laying a single cable. Easy.
     
    Ross, May 19, 2006
    #10
  11. Charlie Crackle

    JohnC

    Joined:
    Apr 6, 2005
    Messages:
    554
    Likes Received:
    1
    Location:
    Sydney
    As an aside...

    Ross,

    My understanding is that it wouldn't make any difference or require special planning for a 3 phase install, since each individual dimmer would usually be connected to it's own power-side thingy (the left-hand block of 4 modules).

    It wouldn't matter if you had dimmers on different phases, providing you didn't "cross-wire" the power-side from a dimmer on 1 phase onto the output-side of a dimmer on another phase. I've done lots of large installs on mulitiple phases, and have never had to worry about what phase is connected to what.

    So, we need to be careful that we aren't mixing things up and confusing the issue... the original discussion is about C-Bus NETWORK Power Supplies, and working out a way to keep THOSE in a way that they could be UPS'd or maintained in the event of a dropped phase. The "power-side" within individual dimmer modules could be on different phases for different dimmers.

    Hmmm - I'm explaining it badly I think... what I mean is that a 5508D1AP doesn't have a C-Bus NETWORK Power Supply inside it, but it still has that left-side 4 module block. So, inside there are the electronics to determine the zero-crossing point for the 8 dimmer channels housed in the right-side 8 modules. Sort of a "power supply" for that particular dimmer, but NOT providing any power back onto the C-Bus Network.

    So, it only matters that those 2 sections of that particular dimmer are on the same phase... it does NOT matter if the C-Bus NETWORK Power Supply that is powering the C-Bus Network Comms is connected to that same phase, another phase, or even distributed across different phases...

    The difficulty in Charlie's case is that he seems to have used dimmers and relay units with inbuilt C-Bus Power Supplies (eg: 5508D1A), and those devices are spread across different phases. Therefore, if it drops a phase then the amount of C-Bus Network Power available will drop below the recommended level. However, if the Network power was supplied by a number of 5500PS instead of using dimmers/relays with inbuilt power supplies, then there would not be an issue of a dropped phase compromising the Network Voltage (therefore killing C-Bus Comms for the remaining devices).

    I hope I made some form of sense with that ???

    PS : I finally decided to "solve" Ross's signature tag...

    LOL :D

    01101010011011110110100001101110
     
    JohnC, May 19, 2006
    #11
  12. Charlie Crackle

    Ross

    Joined:
    Aug 3, 2004
    Messages:
    174
    Likes Received:
    0
    Hi John,
    Yes, I am on track with your point of view and the issues with triac dimming, your post just validates my original comment.

    FTR
    I reliable Cbus network should if at all possible be powered by a seperate circuit breaker. Nothing else, nothing more. That phase used should also contain all dimmed lighting loads. If a project is designed in this manner, there are no issues as far as I can see.

    This is a big mistake. You leave your installation open to all sorts of issues. One point of network power supply is the big G.O. Its purely a design issue, the rest is candy.
     
    Ross, May 19, 2006
    #12
  13. Charlie Crackle

    JohnC

    Joined:
    Apr 6, 2005
    Messages:
    554
    Likes Received:
    1
    Location:
    Sydney
    I think you lost me completely, Ross...

    I do understand and agree with what you say about the "preferred method" being to have all your C-Bus Network Power coming off 1 breaker (phase) - that is exactly what I said previously.

    What I was trying to say in the last post is that there is no specific requirement for the dimmers themselves to have integrated power supplies, nor to run them from the same phase as the C-Bus Network Power. It will definitely work, even if it's not considered "preferable" - of course if you had a choice it'd definitely be recommended to keep all the PS on the same phase.

    Most sites I've worked on have considerably more load on the dimmers than the relays. In a large showroom I manage (400+ channels) there is no way that we can have all our dimmers on 1 phase, because we'd unbalance the 3-phase to the absolute bug***y !

    On that site, let's say there are an average of 300 x 0.5A dimmed loads - there's no way we can design a system (nor have a compliant electrical installation) that potentially could draw 150A off 1 phase, and maybe 20A off each off the other 2 !

    So, the way we usually do it is to do all the C-Bus power using Relay Units with Internal Power Supplies or 5500PS's. We connect all the C-Bus power to one breaker / phase. We then analyse the loads and spread them across phases - first the dimmers (1/3 of the dimmer modules per phase - so they are balanced), then the balance of the loads using Voltage-Free relay channels (picking up whatever phase we need to keep it reasonable balanced).

    Is this incorrect? I'm not trying to argue, but rather trying to get an understanding of whether there are any issues regarding the way we have been installing to date.

    Cheers, John
     
    JohnC, May 19, 2006
    #13
  14. Charlie Crackle

    Ross

    Joined:
    Aug 3, 2004
    Messages:
    174
    Likes Received:
    0
    Hi John,
    I tried I really did.

    If I am understanding you here....
    Based on your last post would not 300 + loads constitute the use of say 38 x 8 channel dimmers?

    Would this not constitute the consideration of several networks?

    Could these networks, lets just say 3 for discussions sake, be distributed over each phase.?

    Would this not be a much more desirable and reliable installation considering the electrical isolation of each network?

    I don't see any conflict on the advice I gave you.
     
    Ross, May 19, 2006
    #14
  15. Charlie Crackle

    brj

    Joined:
    May 12, 2006
    Messages:
    26
    Likes Received:
    0
    5500 Ps

    Would it be considered best practice to completely separate the power supply to the bus from the relay/dimmer units and to completely rely on the 5500PS to provide all power requirements.

    Would there be any downside in providing CBUS power needs in this way?

    Also what are the telltale signs when an installation isn't appropriately serviced with power supplies? Or would it simply be unreliable signalling across the network?

    Thanks.....
     
    brj, May 19, 2006
    #15
  16. Charlie Crackle

    Ross

    Joined:
    Aug 3, 2004
    Messages:
    174
    Likes Received:
    0
    Hi brj,
    I would have to say that doing this is unnecessary and will add expense to the project. Additional, separate power supplies are only required to boost a network that has more components that use power than those that can supply it. Or on a backbone multi network project where you just need a power supply for that network since in most cases the backbone is usually just the conduit to the remote networks. Good network design should in all but the odd instance, never require separate power supplies. Good network design should also provide some redundancy in the advent of a component failure. But that?s another topic.
    Additional switchboard space, additional cabling = more unnecessary time and expense.
    Erratic network activity in many forms, the worst case being that nothing works.
     
    Ross, May 19, 2006
    #16
  17. Charlie Crackle

    brj

    Joined:
    May 12, 2006
    Messages:
    26
    Likes Received:
    0
    Relay units with power supplies then?

    So following on from some earlier comments by others, would it make more sense then to have the power supplied by relay units as oppose to dimmer assemblies to avoid potential complications? And to avoid the dedicated power units?


    Thanks.....



    BRJ
     
    brj, May 20, 2006
    #17
  18. Charlie Crackle

    PSC

    Joined:
    Aug 3, 2004
    Messages:
    626
    Likes Received:
    0
    Location:
    Brisbane, Australia
    Brj,

    Every install is different, the clients requirements / expectations should always be taken into consideration.

    As mentioned in earlier posts in the event of a power failure you may want certain units to be powered by a UPS. In this case separate power supplies are the way to go. The separate C-Bus power supplies powered by a UPS will ensure that the network characteristics remain intact.

    As for purchasing output units with built-in power supplies refer to you Clipsal price book and do the math. As Uncle Ross said it comes down to dollars, some output units (with built-in power supplies) are more economical than others.
     
    PSC, May 20, 2006
    #18
  19. Charlie Crackle

    brj

    Joined:
    May 12, 2006
    Messages:
    26
    Likes Received:
    0
    Thanks

    Thanks Peter - so at the end of the day - it's much of a muchness with UPS powered dedicated power units (split across phases) potentially more reliable but with some more cost (maybe) and taking up more DIN space.....

    I'm figuring there's probably 4 or 5 right ways of doing this all with relative pros and cons........


    Cheers
     
    brj, May 20, 2006
    #19
  20. Charlie Crackle

    wanricky

    Joined:
    Sep 12, 2005
    Messages:
    219
    Likes Received:
    0
    Back to basic,

    network bridge.

    Break things up to two networks and boost up the power for each network.
     
    wanricky, May 20, 2006
    #20
Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.