LCC Central Power vs Wall Warts?

[dcarrell8]

Ok gentlemen, get out your crayons you might need to draw me some pictures.

I am trying to understand exactly how to power my LCC system.  I had planed on using an ATX central power supply from an old computer.  I have four of these and I was looking at incorporating all of them into my large layout. 

Dick Bronson and others on the LCC group have suggested that this is not a good idea because the Signal grounds take different paths from board to board, and thus DCC signals can introduce noise shared in the same circuitry. 

This is how Dick does it:  "I personally use a 15V 5A central power supply to power my LCC network via Power Points. I also use the same power lines to supply I/O boards. The key difference is that the local I/O boards have optical isolation which effectively prevents any power ground line from connecting to any signal ground line. I.e. no way for the ground to connect over two different paths."

I'm hoping someone here can decipher things for me.
1.) Power points - does this simply mean warts at separate plug-in sockets?
2.) Same power lines - so create a buss from the 15v 5A power supply?
3.) Optical isolation - not sure I understand what this means in terms of hardware.

[Chris333]

I'm so dumb I Googled "LCC" and still don't know what it is.

  Landscape conservation cooperatives   

[dcarrell8]

I'm so dumb I Googled "LCC" and still don't know what it is.

  Landscape conservation cooperatives   

Layout Command Control.  There are several guys in here that use it on their layouts.  A very simplified way to think of it is this... DCC controls Trains.  LCC controls the layout.

https://img1.wsimg.com/blobby/go/db99a7f1-fc76-4747-86a7-74cf11b1cc30/downloads/NMRA_PSR_Clinic_09-2019.pdf?ver=1575959264583

[Chris333]

Well I'm smarter now 

[peteski]

I was also in the dark about LCC, but now I feel more educated too. 
So I looked up LCC Power point and this is what I found: https://www.litchfieldstation.com/product/lcc-power-point/
It seems that the power points supply 1/2 Amp of power.

As far as optical isolation goes, that to me means that the CAN data bus is electrically isolated from the I/O boards. There must be optical couplers (usually an LED and a photo-transisor) which pass the command packets from the CAN cable to I/O card.  So there are no ground loops through the data cables.

But without diving deeper the LCC system's specs, I can't make any additional comments or guesses.

BTW both, the vehicle, and electrical buses are spelled with one "s".  Buss it is an old-fashion word meaning "kiss", and also a brand name for a line of electric components (like fuses).

[kwledbetter]

So what I think he is saying simply he seperates his layout control bus from his DCC power bus.  Good practice to do so as there is possibilities of noise if all connected together either ground.  It's not unique to LCC if you were using anything besides the normal DCC bus (arduino based,  MERG CAN, etc) for layout control it's a good idea.  On large layouts I have even seen people run 2 seperate dcc systems/buses one for trains the other for accessories.  It's just a good idea to help prevent interference gremlins.

[mmagliaro]

That's how I read all this - that the supply to all the LCC devices is completely separated from the DCC for the track.
And by the way, I see nothing wrong with using an old computer ATX power supply.  They are regulated, they put out a lot of current with a very stable voltage, they can be had cheaply, usually have built-in overload protection, and if one ever dies on you, you can change it out by just grabbing another ATX power supply.  What's not to like?

[peteski]

That's how I read all this - that the supply to all the LCC devices is completely separated from the DCC for the track.
And by the way, I see nothing wrong with using an old computer ATX power supply.  They are regulated, they put out a lot of current with a very stable voltage, they can be had cheaply, usually have built-in overload protection, and if one ever dies on you, you can change it out by just grabbing another ATX power supply.  What's not to like?

The "lots of current" part worries me.  It might be possible that with either a "soft short" of few ohms (not a dead short), or even the resistance of the wiring, the power supply's short protection might not kick in, supplying enough current to melt/burn whatever causes the short, or even the wiring.

I also didn't know that LCC still uses DCC signals - I thought it was its own protocol and power distribution system (but I have to admit that I didn't closely familiarize myself with LCC).

[C855B]

... I also didn't know that LCC still uses DCC signals - I thought it was its own protocol and power distribution system (but I have to admit that I didn't closely familiarize myself with LCC).

 

Where did you read this? AFAIK, LCC is its own protocol and bus architecture, based on CAN bus. "Power points" are low-current (0.5A, IIRC) supply points on the bus wiring, which uses regular Ethernet cables, RJ45 connectors on CAT5 cable. Recall that I was investigating LCC for a while and then wised-up.

LCC is a camel, if you recall the idiom. If there's any relationship to DCC, it's a very distant cousin at best.

[peteski]

Where did you read this? AFAIK, LCC is its own protocol and bus architecture, based on CAN bus. "Power points" are low-current (0.5A, IIRC) supply points on the bus wiring, which uses regular Ethernet cables, RJ45 connectors on CAT5 cable. Recall that I was investigating LCC for a while and then wised-up.

LCC is a camel, if you recall the idiom. If there's any relationship to DCC, it's a very distant cousin at best.

Well, here were some "hints" in few posts that implied a relationship (including the initial post).

Dick Bronson and others on the LCC group have suggested that this is not a good idea because the Signal grounds take different paths from board to board, and thus DCC signals can introduce noise shared in the same circuitry. 

From the little reading I did about it I also thought it was a CAN-bus-based system, but the info here was confusing me.

[dcarrell8]

I misspoke in the original post, the noise introduced is not on the DCC side of things but on the CAN bus LCC side.

I asked the original above question in a different group and Dick Bronson answered the following in bold italics:

    1.) So, the Power Point we are referring to is THE SQUARE WHITE BOX sent from RR-Circuits that says power point.

Yes.


    2.) Power lines would be referring to the 15v 5A AC/DC adapter that plugs into the white boxes from an outlet also supplied with the RR-Circuits powerpoint.

No, the supplied adapter is 15V at 1.2A and supplies one Power Point at 500mA per cable out.

    So 10 RR-Circuits white power point boxes can be powered with one 15v 5A adapter??

No, a 15V 5A power supply could power up to 5 Power Points.


    OR... 10 I/O boards (that inject power into the ribbon cables) can be powered with one 15v 5A adapter? each powerpoint requires it's own adapter.

No, the thing being missed here is that typically each CAT5 cable out of a Power Point can supply 500 mA to run about 10 nodes. That means 20 nodes per Power Point, (10 per each direction/connector) or 100 nodes on the above example 5A supply. (from a third party)


    3.) The 12v requirements of the SMD-8 or BOD-8 is isolated from the factory by design, use a separate power supply NOT the above mentioned 15v 5A adapter.

A BOD-8 is powered from the LCC node at 5V. The SMD-8 is an isolated driver and is powered with a external supply. It could be the same supply that powers the Power Point units. (On my layout it is the same source) However it does not need to be the same power supply.


    and finally don't use ATX power for the 12v because you will burn down your layout and short circuit all your LCC stuff in the process! 

Well, lots of folks have used ATX power supplies and not burned down their layouts. This can be explained in several ways:
a. They have fused the individual small gauge output lines.
b. They have used heavy gauge wire for all power distribution wiring.
c. They have not (yet) had any long term short circuits.
d. Their small gauge wiring is so long that its resistance limits the short circuit current to safe values.
e. Murphy owes them a favor for some reason or other.

Dick

LCC has a pretty big learning curve but the possibilities of what it can do are far reaching.  Dick's answer got me moving in the right direction, terminology is key.

~ Dennis

[peteski]

Looks like you're all set to proceed with your plans.  Thanks for the follow up for the rest of us.

[daniel_leavitt2000]

I have questions....

I know NCE uses a polling type of bus where more decoders mean the control point has to wait for everything to check in before the next poll. But aren't the buses for Digitrax, Lenz, ESU and others push based? Wouldn't this negate the entire issue of slow response time?

Why would you need feedback to the control point for track and signals which would have feedback via panel lights anyway?

Why interface two different networks in the first place?

This just seems like a solution begging for a problem. DCC has issues, but I don't see how this fixes any of them.

Now if this supplanted DCC with CAN Bus, I could see the point - simplified decoder programming, functions operate the same across manufacturers, replacing CVs with something - ANYTHING - a bit more user friendly. But this doesn't do any of that.

[dcarrell8]

I have questions....

I know NCE uses a polling type of bus where more decoders mean the control point has to wait for everything to check in before the next poll. But aren't the buses for Digitrax, Lenz, ESU and others push based? Wouldn't this negate the entire issue of slow response time?

Why would you need feedback to the control point for track and signals which would have feedback via panel lights anyway?

Why interface two different networks in the first place?

This just seems like a solution begging for a problem. DCC has issues, but I don't see how this fixes any of them.

Now if this supplanted DCC with CAN Bus, I could see the point - simplified decoder programming, functions operate the same across manufacturers, replacing CVs with something - ANYTHING - a bit more user friendly. But this doesn't do any of that.

I know NCE uses a polling type of bus where more decoders mean the control point has to wait for everything to check in before the next poll. But aren't the buses for Digitrax, Lenz, ESU and others push based? Wouldn't this negate the entire issue of slow response time?
 
If I understand correctly, DCC is a "one way street" "Master Slave" as far as communication goes.  The above mentioned manufactures cannot talk to each other and are proprietary to our devices.  In short, you cannot connect DCC to a push button, it takes some other interface like JMRI to get there, or some other manufacture has to come up with a DCC friendly push button.  LCC is a "peer to peer" form of communication, it takes no central command station. Every node can listen and act on information from another node.

Why would you need feedback to the control point for track and signals which would have feedback via panel lights anyway?

I'm new to LCC so I'm sure I will not explain this very well but I will try.  Nodes can be set up to "listen" to commands from multiple inputs.  A Node can take commands from both a push button or a block detection unit.  These are called events, each event has a unique address, the Node runs events through logic and decides what to do with the command.

Why interface two different networks in the first place? 

I'm not viewing this as interfacing.  They are two separate systems, DCC will run the trains.  LCC will run the layout.

This just seems like a solution begging for a problem. DCC has issues, but I don't see how this fixes any of them.

I personally don't see LCC as a solution to DCC. It's a choice.  If you want block detection, signaling and other things that are layout specific, not involved with driving a train with a throttle, then LCC is a way to get there.  Is it the only way? No.  Is it the best way?  I have no idea, this is my first go at advanced concepts with block detection, signaling and JRMI.

Now if this supplanted DCC with CAN Bus, I could see the point - simplified decoder programming, functions operate the same across manufacturers, replacing CVs with something - ANYTHING - a bit more user friendly. But this doesn't do any of that.


I believe all the above issues you mention is why the developers came up with LCC.  At this point, it only replaces certain accessory decoders designed to control the layout but not DCC as whole.  I view it as an option to control my layout, independent of driving trains.

~ Dennis


 

[railnerd]

LCC is meant to replace things like LocoNet, XpressNet, C/MRI, or NCE Cab Bus.