ALC/OnQ Hardware Question

[TonyStewart]

sbwright, you're on the right track here.

Where possible i would join the polling loop wires (white and yellow on the switches i believe) together where you have multiple switches in the same gangbox/location. This will cut down dramtically on the total length of your communication bus. I don't the length is a significant issue for the On/Common/Off wires, some voltage drop may occur, but i doubt it woudl be a real issue since there is no reall communication going on there.

I believe that the hubs essentially have a signal amp in them (powered by the 12v) and that the output of that amp is shared across all 9 ports, so it's not 500ft per port, but the total that counts. I hope to get this confirmed somehow..

Between the lot of us i think we are starting to have a pretty good understanding of this system and it's probably a good idea to get the 'designs' vetted by the group before jumping in and spending $$$$.

I havent' messed with my ALC install (12 active switches or so at this time with 6 NIB) in a few months but need to pick this up again to get some more things automated. Anybody hear from Tony Stewart lately?

For Broconne's answer #2 I am assuming that you either have mutiple switches daisychained and connected to each of the 9 ports on the branch hub, you cascade the branch hubs(4) to get the total of 31 puchdowns, or use the 66 blocks? But then again the document says the hub only supports 9 ALC addressable switches, makes me wonder if that is due to only having the 9 physical connection points, something doesn't add up here.

In post #6 the powerpoint slide indicates that the Branch Hub "Extends ALC branch cable additional 1000 feet(total 1500 feet)" this is not consistent with the other documentation and it is not clear as to:

1. 
   1) Is the 1500' for the entire branch length.
   2) Is it 1000' between the branch hub and dist module and 500' for the remaining switch wiring.
   3) Strictly for the wiring between the branch hub and distribution module.

If the 1500' its for the whole branch wiring length that would be great but sounds to good to be true. As with you we need to know the right answer before proceeding.

According to this section of the ALC/Onq Technical Guide any single switch should not be greater than 500' from the interface. Sounds like in a homerun installation as long as each switch individually connected to a Branch Hub or Distribution Module with no longer than 500' of wire would be acceptable

View attachment 2662

Anyone with a current setup and a spool with 500' of wire up to the challenge???

I am here guys. I have been very busy and traveling a good bit. Yes the hubs have amplification, opto isolation and signal buffering/conditioning. While it is rare that electrical noise is coupled into low voltage wiring it does happen. I have seen a number of times when powered hubs saved an install in this regard.

See my other post about the 9 ALC limit for more details. A quick summary is that any one 9 branch hub can support up to 31 devices.

As for the cable length limits. 1500 feet is pushing it and I don't recommend it. The documentation is not clear in this regard. So let me tell you my take and my experience.

Put a hub no farther away than 500 feet from the controller. Then ALL OF THE ALC CABLE RUNS from the hub to the switches must not exceed a total of 500 feet.
Example: 9 runs of 75 feet would total 625 and would be out of spec!

I have visited a number of installs where longer cable runs have been used and problems existed. Things like slow response (or missed signals) can result. This is rare as ALC is tough. But if your luck is like some of mine, your install will be the one that misbehaves. I admit that I have seen longer runs work with no issues. But it's a gamble!

A 66 block has no noise immunity, isolation or amplification. For a very small install, it's a good idea and an excellent way to lower costs. Just be careful with distances and noise from AC wiring.

Tony

 

[TonyStewart]

My current understanding of the whole "HUB" amplification thing is that each hub creates a new output side with a 500' total wiring limit, and it also provides some punch down blocks. I believe the DATA wires on all the punch down blocks are connected together. Punching swithch A1 to port 1 and A2 to port 2 is just a convenience, and you would have the same wiring length limitation if you punched them all to port 1 (via a 66 block, for instance).

The HUB also amplifies the source signal, and can be up to 1000' from the branch source (the controller or the lighting expansion module). This 1000' bonus is only available if you don't have any switches wired directly to the branch source. (OPTION 2 SBWRIGHT)

You would think it would be possible to "cascade" hubs by wiring the hub to one of the ports on another hub and achieve significantly greater total distance from the branch source. However, additional hubs are not intended to be wired to output ports (1-9) on the previous hub, but rather to the input via port 10.

--Bob

P.S. It would be interesting to know what kind of protocol ALC uses for the data. RS-485? Do you think they have a collision detection system?

It is a custom protocol that is available here for download in the ALC lighting docs section. It does have serious collision detection and also is true two way. Meaning that there is feedback that the command was recieved and carried out from the controller as well as knowing when a local switch is pushed manually. This manual feedback is very useful for use as an occupancy sensor and/or path of travel tool.

If you want to play with the protocol, be sure to order the ALC branchtech software (for installs that use a controller from HAI or ELK) or the Scenetech software if you are using the ALC stand alone controller. You will also need the ALC branchtech interface for the HAI/ELK (or for applications you write from your PC) installs.

 

[TonyStewart]

My current understanding of the whole "HUB" amplification thing is that each hub creates a new output side with a 500' total wiring limit, and it also provides some punch down blocks. I believe the DATA wires on all the punch down blocks are connected together. Punching swithch A1 to port 1 and A2 to port 2 is just a convenience, and you would have the same wiring length limitation if you punched them all to port 1 (via a 66 block, for instance).

The HUB also amplifies the source signal, and can be up to 1000' from the branch source (the controller or the lighting expansion module). This 1000' bonus is only available if you don't have any switches wired directly to the branch source. (OPTION 2 SBWRIGHT)

You would think it would be possible to "cascade" hubs by wiring the hub to one of the ports on another hub and achieve significantly greater total distance from the branch source. However, additional hubs are not intended to be wired to output ports (1-9) on the previous hub, but rather to the input via port 10.

--Bob

P.S. It would be interesting to know what kind of protocol ALC uses for the data. RS-485? Do you think they have a collision detection system?

Exactly right. You certainly can let one hub feed 9 more...

Tony

 

[TonyStewart]

I haven't seen a hub in person either. Maybe someone who has can try to see if the circuitry is wired individually to the ports or if the DATA lines are all in parallel.

Which document did you see which questions the hub amplification wiring theory? My understanding came from the HUB installation instructions IS-0235

--Bob

In post #17 above I included a section from the ALC/Onq Advanced Lighting Control - Technical Guide document #1307573. It has recommendations identified for when the total branch circuit wiring must be greater than 500'.

Exactly right. Total wiring limit is 500' not 500' per run......

Tony

 

[TonyStewart]

The best advantage for me using hubs are the switches that can turn off each of the nine legs one at a time for trouble shooting. I have NEVER installed ALC without having to trouble shoot. WHy? not a fault of the "system" but all those tiny little wires and mulititude of connections. I always manage to have to redo at least one butt connection.

The enhanced branch hubs save the day.

The voice of wisdom here. Oh wise you are Ranger Digital.

I will add this. We have held training classes for integrators in home automation for the past 13 years, qualifying me to state this.....
Integrators are no match for the DIY'er. You guys are less trouble and have fewer issues. Sure there some great integrators I have met. But typically they are few and far between.

That said, take care with your connections, color code everything, use wire numbers and you may not see a single issue on installs. The enhanced hubs offer a much cleaner and worry free install...

Tony

 

[TonyStewart]

The best advantage for me using hubs are the switches that can turn off each of the nine legs one at a time for trouble shooting. I have NEVER installed ALC without having to trouble shoot. WHy? not a fault of the "system" but all those tiny little wires and mulititude of connections. I always manage to have to redo at least one butt connection.

The enhanced branch hubs save the day.

The 66 blocks do essentially the same thing (I would argue EXACTLY the same thing) for quite a bit less $. Just to point that out.

See my other posts on this. Summary is that a 66 block is a creative way to make connections on small installs. It does not make a good replacement for a hub though and are certainly not exactly the same thing.

Tony

 

[TonyStewart]

I will look again this evening, but i'm fairly sure that the data lines of all the ports on the hub are paralleled of each other so this means connecting 9 load together on a 66 block and then a single 'trunk' line over to the hub woudl essentially be the same as landing 9 loads individualy on the Hub. The hub is essentially an AMP with some individual punch down block, but they might as well have provided a single set of terminals and have us twist and stick all the cat5 conductors in at the same location. The end result woudl be the same, but a lot less organized.

Can somebody further eleborate on the connection of hubs of a other hub and this port 10 business? It has me a bit confused. If you went hub to hub to thub to hub,etc. Can you essentially put a light as far away from the interface as you want?

Hopefully TS will chime in here shortly and clear up this mess...lol

After being absent for a few weeks I hope I am addressing the right issues.

I assume you are talking about powered hubs 364677-01 or 364736-01. Both of these models use identical circuitry with the only difference being the dip switches on the 736-10 model. And...

Both models are in fact not paralled. At least not in the way that an ohm meter can answer the question. The field wiring side uses opto isolators to provide both a level of coupling and isolation from the noises that come in from the field wiring. Next, you have some pretty decent buffering against noise with a final caveat of amplication.

Many of you are electronics gurus. Look up the chip and other component numbers of the circuitry for yourself. You will find what I am saying is accurate. And better yet, the hubs are not proprietary. Meaning that off the shelf components make up the entire hub. You could even bread board your own hubs if you are so inclined (may cost you more than buying though).

I am not sure what is meant by a port 10? After the 9 ALC landing points and the aux switch landings, the only other port is the connection to the lighting controller interface. Is that what you mean by port 10?

As for the 66 block as a replacement for the powered hub. No way! But it is still a good idea. Hubs increase the price, 66 blocks are clever and cheap. I personally would not plan an install around a 66 block that was over 500' of total cable or beyond 6 or 8 ALC devices.

Hope this helps.

Tony

 

[TonyStewart]

Don't know if this will help or not. Just as I think I understand it.... I learn something new. Attached is graphic from a powerpoint file that might help, if you haven't already seen them. PPT is to big to upload here but would make available if you want. it

Here is a link to all ALC docs and Powerpoints. Select the ALC docs section.

Tony

 

[TonyStewart]

Left: 66 block. cat5e's coming in are punched down on the left. Jumpers on the right of the block connect the on/com/off conductors, also connect the ALC +/- for switches in the same room.

Right, top to bottom: Elk ALC cotroller (currently running in stand-alone mode), distribution hub, branch hub.

EDIT: changed to more recent pic depicting bridging clips in place.

I am trying to catch up and just now saw the picture this this post. So forgive me for not praising you sooner!

I do like this wiring scheme for a number of reasons. The bridging clips act as a form of dip switch. And in this install the quantity of devices is small enough to not worry about cable length (I assume). Let's face it folks, who has unlimited cash these days. The 66 block will work well in some instances.

When I am asked to troubleshoot someones install that has homerun all cables, I carry a few enhanced hubs and install them before I begin. I use hubs that are already punched down and that run to terminal strips which already have leads with aligator clips wired to each Terminal. I find that this test jig is fast to install and supports connecting to multiple ALC cable runs well. I also ALWAYS use Branchtech and the Branchtech interface I mentioned earlier.

More for a troubleshooting jig than a permanent install, the 66 block idea is similar to this and could be adapted as well. Just be alert that a 66 block is a passive device and a hub is powered. As stated earlier, they have different homes!

The picture show a very neat install as well. Most ALC problems come from poor connections. If you wire a system as neatly as this one, then the most common installation problems are eliminated.

Tony

 

[TonyStewart]

Left: 66 block. cat5e's coming in are punched down on the left. Jumpers on the right of the block connect the on/com/off conductors, also connect the ALC +/- for switches in the same room.

Right, top to bottom: Elk ALC cotroller (currently running in stand-alone mode), distribution hub, branch hub.

Just to play devil's advocate, I contend that both the distribution hub and the branch hub are wasted in this setup.

I believe the sole purpose of the distribution hub is to land wires in its punch down blocks. The hub's job is to land wires and amplify signals and provide a "new" 500 foot wiring limit. You aren't landing any wires on the distro hub or the branch hub, and since all your lights are wired to the hub, you are using its 500 foot limit, but not the 500 foot limit of the branch itself. Perhaps it would make more sense for the first 66 block segment to attach directly to the branch, and the second segment to connect to the hub.

Heck if you paid for the distro hub and branch, why not just land your connections on it? I'll bet you could put all the connections from your 66 block onto the two ALC devices instead. Is there some downside to wiring the way ALC intended (i.e. landing to their punchdowns)? (Devil's advocate -- no criticism intended)

--Bob

P.S. I just finished the wiring layout for my house. It contains 11 Hubs in 4 locations for 97 switches and 59 AUXes. To get around the 500'/branch limit I need lots of hubs! The 66blk is $12/12loads (assuming several AUXes). The hub is $40/9 loads (and 10 AUXes) so it is a bit more expensive.

Not wasted! The hub / 66 hybrid is a good merger of the two ideas. Remember that the ALC polling loop is 4.2Volt DC (in it's best pristine state). The cable length limits I addressed earlier are governed by this voltage. You simply can't run 4 Volts too far without the cable impedance impacting the signal level. While cat5 is wonderful cable it is not miracle cable. the twists are more for supporting high speed data than for propogating long cable runs (it is not shielded cable, it's twisted).

Bob, I would be concerned over cable loss with an install of your size. You may get lucky, but I wager that your speed of switch response will immediately suffer (become slow if not noticeably sluggish). You may even have some hard communication errors on long cable runs without a powered hub. here is an example:
When you manually press an ALC switch it will instantly come on - no delay! But..... when you use a scene switch or an ALC switch as a trigger to a programmed event, count on a delay of around 1 second (or worse if you really pushed the limits of the cabling). And.. unless the aux is tied directly to the switch it is a 3 way to (not homerun back to the 66 block), then there will be a delay. So for example, you homerun the Auxes. Then you can see a long delay here too. This is a more bothersome situation as a 3-way is noticeble when it is sluggish. Take the application of a 3way (or a sceneswitch or ALC programmed as a 3-way) in the stairwell. Who wants to be on the third of 4th step before the light comes on.

Understand this. ALC when wired correctly and inside recommended specs has no noticeable delay, 3-ways or otherwise. I am talking about times when you push the limits as I fear is often done with large installs like yours. OnQ did not invent the powered hubs because they had extra parts on hand and decided to use them! They were created for a specific application and meet an exact need. There is a time when they should be used and a time when they are optional. Not knowing the difference is dangerous.

Without powered hubs, You may also have intermittent problems due to occasional EMF, RF and/or AC coupled noise. Maybe not, but it can and has happened on large systems where the wiring is substantial and unconditioned.

Lastly, with large installs like yours your wiring can become a lightning antenna. Hopefully you are not in a state where lightning is an issue. If you are, the 66 block will couple surges right into your expensive electronics. With powered hubs, the opto isolators may save you on a typical static discharge and with a stronger strike may act as a road block to the surge (by blowing up) and might keep the strike out of your expensive controls.

Now some of you forward thinkers are thinking that you can put a surge protector on each of the 66 block inputs. Sorry, but telecom surge protectors clamp at 48VDC (or 24VDC for KSU systems) and are not going to do you any good on a 4 VDC signal.

Tony

 

[AceCannon]

I like the 66 block idea as well. But you must run a whole lot of cable. Be sure to land only one ALC switch per connection of the 66 block. You will still have to unwire the loops from the block to find a bad switch, but this is easier than pulling switches out of the wall.

Hope this helps.
Tony

Thanks so much to Tony for providing expert insight.

My DIY design may be less than ideal, but my overall thought process was:

1- branch hub(s) to amplify the signal

2- 66 block(s) so I could pull a bridging clip at any time to remove a switch from the system for trouble shooting (as the enhanced branch hubs allow)

3 - Note that this plan leaves all other ALC devices intact, on the com bus, even when one switches' bridging clips are removed.

 

[sbwright]

Thanks Tony for the detailed feedback and links, there is no substitute for experience.

 

[rbroders]

Tony,

We bow to your experience, but have you ever experienced a delay on "homerun aux 3-way switches"? Based on my understanding these contact closure signals travel at essentially the speed of light (barring capacitance effects). There shouldn't be any delay no matter how long the wires are!

--Bob

P.S. My install is all powered hubs and no 66-blocks. I have no scene switches and if necessary I can split it into multiple independent ALC systems to reduce the apparent size of each system.

 

[AceCannon]

We bow to your experience, but have you ever experienced a delay on "homerun aux 3-way switches"? Based on my understanding these contact closure signals travel at essentially the speed of light (barring capacitance effects). There shouldn't be any delay no matter how long the wires are!

I realize I am not Tony, but I do have 3 aux switches installed, all wires are homerun to the central wiring closet where they connect to the load bearing switches' low-voltage conductors.

All work immediately, no delay. I feel certain that they indeed are "contact closure" and as such there are really no "signals" except ON/OFF. (yes/no, 1/0, whatever).

 

[felixrosbergen]

Thanks to TS for chiming in with the expert advise.

Acecannon. Did you land all the auxes on the 66 block as well and did you then connect them on the 'wiring room' or 'field' side of the 66 block?

My thoughts are that I have a few 4 and 5 ways and the distro modules provide nice convenient ways to conenct up to 3 auxes to a dimmer whereas i think this may get messy with a 66 block.

Tony, you said the ports on the powered hub are opto isolated. You mean ports 1 though 9 are all isolated from each other OR did you mean the that ports 1 through 9 are isolated from the ALC interface side? If they are isolated how could i have measured 0 resistance say from port 2 to port 6 on the polling loop connections. The AUX's are obviously isolated except for the special ones (port 8 i think).

The surge supressor modules you linked to are interesting, but thats another $40 per branch..ouch..and the question is if they will really save the install? I already have a whole house surge suppressor. It seems that if you take a good hit NOTHING will truly protect you.

Against..thanks for looking into this thread...while the ALC documentation linked is quite usefull it's 'interpretation' seems still tricky. The good thing is that amongst the group of us I think we're starting to get a rather good understanding of the system capabilities.

All we need now are them outlet and inline modules.....

When the winter comes I plan to resume my ALC install to replace the 4 or 5 switches left on the ground floor and start on the upstairs. While for a few locations an ALC switch is hard to justify OCD laws prevent me from having a 95% automated lighting solution in my house. lol