T tapping the data line

[jlarson]

I was wondering if they made any "T" adpators that you would not have to daisy chain your data lines?

[-klb-]

It is easy to forget, but there is the network repeater:

http://store.lightorama.com/rsnere.html

It has two ports on the base network, and one port on the repeated network. So you can use it to tee a network..

But, at that price many people figure out how to straight line it instead...

Also, remember that the USB-485B has two ports. So it can go in the middle of your network.. Some items hooked up to one port, and the rest to the other...

[Paul Roberson]

You could use a RS485 Network Repeater. Again a bit pricy for a T, but cheaper then ELLs


Oh well did not type fast enough. :shock:

What klb said.

[jsholla]

The biggest problem with a RS-485 network is contolling signal reflections due to network topology, cable lengths and proper network termination. There is plenty of literature on internet that talk about the theory of RS-485 for those of you interested in learning about charactistic impedance, etc. With that said, the best network topology is a daisy chain with both end of the network terminated to using 120 Ohm resistor. In most cases the RS-485 adaptors used at the computer end of the network is usually terminated at 120 Ohms, while the other end is not. I made a simple terminator using an RJ-45 connector and placing a 120 Ohm resistor between the middle two pins, which are the two signals. Then plugging the terminator into the last RJ-45 connector on the network.

I've included an article which explains the basics of RS-485.


Attached files RS485_RS422_Basics.pdf

[Dan C]

Welcome to the forum fellow Nebraskan!

In all honesty Network cable is not that expensive.
you can order 100' cables for under $10 from places like Monoprice.com. or even a 1000' roll for under $80.

So for me, that is a lot cheaper than Easy light linkers or a repeater.

[Guest guest]

So you're saying that you can't use a splitter like telephone lines use to take one line and have it have go out to 2,4,6, or 8 lines being split from the main phone jack using those phone line splitter jacks?

I know you can use these to have like a cordelss phone, Fax machine and regualr phone 9and in teh old days) a modem all connected to a single phone line jack.

I honestly do not understand, if such a splitter is available for cat5 cable, as long as the outlets are all wired in parallel that this could not work since each controller has its own ID, the only difference is that the controllers would not be daisy chained in series, but I would think going from a parallel splitter this should be possible, after all, the jacks in the USB485B are in parallel and you can "split" your controllers on that, for example I can put 3 on one side and 2 on the other.

So if a T connector is wired in parallel and plugged into the USB485B or even an MP3 Director unit, I just don't see why this would not work, as I can find no difference in using a parallel USB485B's output cat5 jacks as opposed to a spliiter that would hopefully be wired in parallel, allowing each controller to be plugged into each cat5 jack independantly and still function the same as if daisy chained to each other.

Either way still uses the same number of cat5 cables.

So why couldn't this work?

I'd really like to know. LAN cabling has always been a bit of a puzzler with me as to why some things will work, but others won't.

[Steven]

Orville wrote:

So you're saying that you can't use a splitter like telephone lines use to take one line and have it have go out to 2,4,6, or 8 lines being split from the main phone jack using those phone line splitter jacks?

The difference is that telephone is audio frequencies (less than 3kHz). The LOR network uses typically 56kbps. At that speed the distributed nature of the transmission line becomes more pronounced.

I'd really like to know. LAN cabling has always been a bit of a puzzler with me as to why some things will work, but others won't.

This is one of those few cases where I can actually identify a class I took in school that specifically answers a real question. The class, if I recall, was called "Transmission Lines." A transmission line is anything where you put a signal into one end at it comes out the other. In this case, the transmission line is one pair of a Cat5 cable.

As the frequency gets higher, transmission lines behave less as just connections and have distributed effects, such as reflections. A reflection happens when the impedance of the transmission line changes, such as the end of the unterminated cable, or at a 'T'. If the impedance change is large enough, the reflected signal interferes with the data signal and causes the receivers to decode the signal incorrectly.

A 'T' will also reduce the signal strength by half (which means the voltage will drop by about 30%). This reduction does not happen on a straight-line network because the receivers don't load the network very much.

[Guest guest]

Thank you very much Steven. That does make some sense to me about the operating characteristics of these "adapter" type connectors, but got to say I'm still somewhat confused.

Why?

Because why would LOR have a dual cat5 jack (parallel) on their USB485B units, if running the cable as one straight series based line is better, why wouldn't this only have a single jack?

Or does this reflection effect only happen if attempting to use more than 2 jacks in parallel?

I'm just trying to really understand what the difference would be in a T that would have 5 parallel cat5 connections, as opposed to a LOR product that has 2 cat5 connections in parallel, especially since parallel delivers the exact same criteria in all jacks, multiple cat5 jacks in series I could see this issue cropping up where say 5 controllers would all be plugged in and this would cause a power loss along the cat5 cable, but parallel jacks, no matter how many should not do this (or so I would think).

So I'm still somewhat puzzled as to why a multi-parallel cat5 connection system would create any type of issues, when LOR uses a dual cat5 system on the USB485B and the USB485 repeater product lines.

Maybe I'm over-thinking this, but it sure seems like there shouldn't be any issues using one, not as long as all the cat5 jacks were wired in parallel to each other.

[Steven]

Orville wrote:

Because why would LOR have a dual cat5 jack (parallel) on their USB485B units, if running the cable as one straight series based line is better, why wouldn't this only have a single jack?

A straight line network has 2 ends.

The computer (USB485B) is not unique in the network, it is just another node. (Every node can receive and transmit on the network. Since it's a straight line network, the signal propagates both directions from the transmitting node until it reaches the ends, which should each have a terminator.) The computer, being a node on the network, can be any distance from either end. Since it can be in the middle of the network, it has two sockets, just like the controllers.

In a typical display, you probably could add another jack in parallel at the USB485B, and because the computer is doing most of the transmitting there probably wouldn't be a problem with reflections. However, the other controllers do transmit sometimes, such as when they are polled for the state of the inputs. Also, adding a 3rd segment onto the network would reduce the signal power, because it would be divided by 3 instead of 2, which would make it slightly less reliable.

[Guest guest]

Steven wrote:

Orville wrote:

Because why would LOR have a dual cat5 jack (parallel) on their USB485B units, if running the cable as one straight series based line is better, why wouldn't this only have a single jack?

A straight line network has 2 ends.

The computer (USB485B) is not unique in the network, it is just another node. (Every node can receive and transmit on the network. Since it's a straight line network, the signal propagates both directions from the transmitting node until it reaches the ends, which should each have a terminator.) The computer, being a node on the network, can be any distance from either end. Since it can be in the middle of the network, it has two sockets, just like the controllers.

In a typical display, you probably could add another jack in parallel at the USB485B, and because the computer is doing most of the transmitting there probably wouldn't be a problem with reflections. However, the other controllers do transmit sometimes, such as when they are polled for the state of the inputs. Also, adding a 3rd segment onto the network would reduce the signal power, because it would be divided by 3 instead of 2, which would make it slightly less reliable.

Thanks Steven! Now that makes perfect sense! That I understood completely.

Thank you for explaining that, as this is one thing I would have possibly tried to do myself, not realizing it could have repurcussions on my display as the show is running!

Even though I don't use a dedicated computer, I use the LOR DC-MP3 Showtime Director, but can see that no matter what runs the show, you want to keep the data/power (cat5) lines as dedicated as possible to prevent any issues that could really mess up your show!

[Max-Paul]

Couple of my thoughts about Orville's questions.

The repeater with dual outputs. Each of those outputs are treated as two individual circuits that merge going back to the computer. Meaning each of those outputs more than likely have their own RS-485 chips with the inputs tied together in parallel.

Ok, why you want to avoid stubs (T) circuits. As Steven pointed out. In a normal RS-485 daisy chained topology you have two points that can cause reflections. With stubs each stub becomes a reflection point. In the normal circuit you put termination resistors of 120 ohms at each end to kill any reflected signal. But if you did this at the end of each stub. Then you would load down the circuit to the point you would now kill the wanted signal.

[Guest guest]

Max-Paul wrote:

Couple of my thoughts about Orville's questions.

The repeater with dual outputs. Each of those outputs are treated as two individual circuits that merge going back to the computer. Meaning each of those outputs more than likely have their own RS-485 chips with the inputs tied together in parallel.

Ok, why you want to avoid stubs (T) circuits. As Steven pointed out. In a normal RS-485 daisy chained topology you have two points that can cause reflections. With stubs each stub becomes a reflection point. In the normal circuit you put termination resistors of 120 ohms at each end to kill any reflected signal. But if you did this at the end of each stub. Then you would load down the circuit to the point you would now kill the wanted signal.

Just cracked open my USB485B with power booster, nope, only has one 485 chip in there. Both cat5 sockets wired together in parallel to the single 485 chip.

Just so you know.

[Ron Amedee]

Orville wrote

Just cracked open my USB485B with power booster, nope, only has one 485 chip in there. Both cat5 sockets wired together in parallel to the single 485 chip.

Just so you know.

You are abrave man my friend!!!

[Guest guest]

Ron Amedee wrote:

Orville wrote

Just cracked open my USB485B with power booster, nope, only has one 485 chip in there. Both cat5 sockets wired together in parallel to the single 485 chip.

Just so you know.

You are abrave man my friend!!!

Two screws and it slides right apart. Two more inside and you can remove the circuit board.

Took less than 2 minutes to open it up, look at the chips and see how it's wired up.

Actually took me longer to type this, than it did to crack 'er open and look at 'er guts! LOL

[Steven]

Orville wrote:

Just cracked open my USB485B with power booster...

I opened mine when I got it a couple of years ago because I wanted to wire jumpers on the sockets to pass the unused pairs from one jack to the other.

[Guest guest]

Steven wrote:

Orville wrote:

Just cracked open my USB485B with power booster...

I opened mine when I got it a couple of years ago because I wanted to wire jumpers on the sockets to pass the unused pairs from one jack to the other.

Steven what was the purpose in doing that?

One thing I had thought about doing was adding a POWER L.E.D. on my USB485B so I know the USB port is definitely supplying power to it.

Would have been nice if LOR would have done this like they did with the repeater.

At least that way at a glance you'd know the unit was getting power from the computer, especially if the controllers weren't being found, or the SE not sending commands to the controller(s), you could at least know the USB485 unit wasn't getting power and either try another USB port, and if the LED lit you'd know the USB port it was on has a problem. If LED on and controller(s) not being found(LOR Hardware Utility) and/or not working, then look to possibly reset controller(s). If no error indication or indication of a connection on the controller LED, then you'd know to suspect the 485 board may have an issue.

So I like little lights to tell me something definitely has power going to it from a computer port.

[Steven]

Orville wrote:

Steven what was the purpose in doing that?

It's a long story, but the short version is that I use both sockets because I feed controllers located both east and west of my garage, and I use the same cable to pass audio to outside speakers.

One thing I had thought about doing was adding a POWER L.E.D. on my USB485B so I know the USB port is definitely supplying power to it. Would have been nice if LOR would have done this like they did with the repeater.

The USB-RS485 Isolated Adapter has a power light.

[Max-Paul]

Orville,

Thanks for the info. Who would have thunk that they would do it that way? Well so much for presuming that someone would have made it so that it would have been a true two data circuits.

Again thanks for the info Orville.

[Guest guest]

Max-Paul wrote:

Orville,

Thanks for the info. Who would have thunk that they would do it that way? Well so much for presuming that someone would have made it so that it would have been a true two data circuits.

Again thanks for the info Orville.

That's why I cracked her open. Sometimes I just have to know what's inside!:shock: (used to drive my parents nuts because I'd be one of those kids that took everything apart, just to see what made it tick, they were even more surprised when I'd put it all back together again, sometimes making additions!) LOL

But I had thought the same as you, two seperate 485 chips and two true data circuits, but turns out it's only one chip connected to both jacks!

I was actually quite surprised that it was only a single chip set up.

I honestly did expect to see two data circuits in there myself!

[terrypowerz]

Ok, I am confused....

Orville, (Clay, Wilbur...whatever) cracked open his USB adaptor.... NOT a repeater.

It is a totally different animal.... it is at one end of the network, not midstream like the repeater is intended for...

I don't think that anything useful has been learned here.