ControlNet T Taps

D

Thread Starter

Doug

I am trying to better understand the connections for ControlNet and I can not seem to find detailed information on the wire connections.
Thank you,
Doug
 
K

Kelvin Proctor

Doug,

I can't give you the detals you want be it might be helpfull to know the ControlNet is based on the standard IEEE802.4 Token Bus protocol. This might help you in searching for the information you want.

Cheers,

Kelvin
 
All the ControlNet taps are made by a company called Wirepro:
http://www.wpi-interconnect.com/cables2.htm
Allen-Bradley and ControlNet International are pretty much their only customers for this particular cable. Wirepro has put a lot of time and money into the large production runs of tuned components and their minimum order is immense.

The ControlNet International specification, Part 2, Chapter 5.8 gives all the technical specifications for the frequency response and various impedances of the T- and Y- taps. It makes me dizzy.

Inside the injection-molded housing between the two BNC connectors are three components in parallel: a 56 pf capacitor, a .56 uH inductor, and a special center-tapped autotransformer. The center tap of that autotransformer goes to the 1 meter drop cable.

Could you make your own ControlNet taps? Maybe. I recommend against it because finicky components like custom transformers aren't my (or my
Company's) cup of tea.

The single most common problem I've found with ControlNet installations (other than missing terminating resistors) is a center wire in the co-ax that's too short and didn't get crimped properly to the center pin. This will let the pin vibrate inside the BNC barrel and give an intermittent connection. Usually the outer crimp is cruddy too and the connection will come off in my hands. A properly crimped RG-6 connector will withstand about 75 lbs of pull.

The ControlNet Cable System Design Guide from Allen-Bradley is a pretty darn good publication for the more mundane crimping and cable length
matters of ControlNet, and it's free from www.ab.com/manuals or www.theautomationbookstore.com.

Cheers,

Ken Roach
A-B Seattle
[email protected]
 
>Ken,
to better understand the ControlNet I was hoping to find information on the actual wiring connection of the t & y taps.

Basically does the connection to the trunk line actually make a T or does the trunk line actualy go through the node (do the nodes after the T loose comms if a node is unplugged)?

Thank you,
doug
 
C

Curt Wuollet

That's because you're not supposed to. What good would a proprietary lock-in scheme be if you could easily avoid high priced components
and one of a kind cable? And if you really understood what was there, it'd only make you unhappy. That's why I'm holding out for tunneled
protos through TCP/IP and commodity Ethernet. I sure wish Modicon would let us GPL Modbus/TCP code. That's the only thing holding us back from fieldbus development.

Regards

cww
 
M

Mike Johnson

I think ControlNet Taps and cable have the same characteristics as Ethernet taps and cables.
I have found that Ethernet taps and RG-6 cables from Radio Shack work quite well.

Mike
 
Although Curt will call me a hegemonist for saying it, DO NOT use Ethernet taps with a ControlNet system. It might work in the short run, but it will cause you problems down the road.

"Thin Ethernet", aka 10Base-2, uses RG-58 A/U cables and connectors and terminators with 50 ohm resistance. ControlNet uses RG-6U quad-shield
cable, special taps, and 75-ohm resistors. The BNC connectors are the only things which are the same between the two kinds of wiring. I use
Radio-Shack components (and my local cable house, Radar Electric) but they're RG-6, made for closed-circuit TV.

Say what you want about ControlNet taps, but having the resistor, capacitor, and transformer inside those allows a ControlNet segment to be
up to 1000 meters long, instead of the 185 meter limit for 10Base2 Ethernet. That's important in a big factory, and A-B is a big-factory
controls supplier.

Ken Roach
A-B Seattle
[email protected]



 
I am mistaken. I retrack my comment on the Ethernet tap. I thought Ethernet taps where 75 ohms. I am familar with the importance of
impedance matching on a transmission line in order to minimize reflection. It was also brought to my attension that the tap we used
was a A-B "Y" style tap with RG-6 cable from Radio Shack. It just seemed to me that ControlNet uses Manchester biphase
encoding/decoding like Ethernet and besides the fact that ControlNet I think uses 8.2 V peak to peak and Ethernet I think uses 12 V peak to peak, the signals are identical.
I would have thought that as long as the cabling was uniform with terminators equalling that of the cable and coupled to the transceivers via transformers with characteristics that there is high impedance as "seen" by the cable with low impedance coupling at the line drivers of the
transceivers, signal transmission should be fine.
Am I wrong in this assumption?
Or just a rambling dumb ass who has no clue about digital communication over coaxial cable?

Mike
 
C

Curt Wuollet

Hi Ken

Ken Roach wrote:
>
> Although Curt will call me a hegemonist for saying it, DO NOT use Ethernet
> taps with a ControlNet system. It might work in the short run, but it
> will cause you problems down the road.

I would do no such thing. Correct is correct and I am very attuned to transmission line theory. I might find a collegue with a network
analyzer to verify it or failing that do my own analysis with a pulse generator. In fact, I would add that said BNC's are not shared, the 75 Ohm has a different shape and diameter pin and interchanging them not only causes reflections due to the difference in characteristic impedance
but mechanically deforms the female contacts leading to intermittant problems.

The correct solution to an overpriced, single sourced, inadequately documented, proprietary fieldbus system is to find out before you are locked-in and simply say no. If that means changing PLCs, the PLC functionality is far more universal and standardized than the closed environment built around it. Shop for Open and vote with your wallet. Good Luck. Hint, Modicon seems the most enlightened.

>
> "Thin Ethernet", aka 10Base-2, uses RG-58 A/U cables and connectors and
> terminators with 50 ohm resistance. ControlNet uses RG-6U quad-shield
> cable, special taps, and 75-ohm resistors. The BNC connectors are the
> only things which are the same between the two kinds of wiring. I use
> Radio-Shack components (and my local cable house, Radar Electric) but
> they're RG-6, made for closed-circuit TV.
>
> Say what you want about ControlNet taps, but having the resistor,
> capacitor, and transformer inside those allows a ControlNet segment to be
> up to 1000 meters long, instead of the 185 meter limit for 10Base2
> Ethernet. That's important in a big factory, and A-B is a big-factory
> controls supplier.

Yes but I can do 2km with fiber and Ethernet for probably less cost and copper with repeaters for shorter hauls. And the bigger your physical
plant, the more money you save. In addition to having hundreds of competing sources and commodity pricing. All I would say about the
taps is that that's a lot of money for $3.00 worth of passive components. That's not hegemony.

Regards

cww

--
Free Tools!
Machine Automation Tools (LinuxPLC) Free, Truly Open & Publicly Owned
Industrial Automation Software For Linux. mat.sourceforge.net.
Day Job: Heartland Engineering, Automation & ATE for Automotive
Rebuilders.
Consultancy: Wide Open Technologies: Moving Business & Automation to
Linux.

 
C
Hi Mike

I was making a political statement not casting aspersions.

> I am mistaken. I retrack my comment on the Ethernet tap. I thought
> Ethernet taps where 75 ohms. I am familar with the importance of
> impedance matching on a transmission line in order to minimize
> reflection. It was also brought to my attension that the tap we used
> was a A-B "Y" style tap with RG-6 cable from Radio Shack.
> It just seemed to me that ControlNet uses Manchester biphase
> encoding/decoding like Ethernet and besides the fact that
> ControlNet I think uses 8.2 V peak to peak and Ethernet I think
> uses 12 V peak to peak, the signals are identical.
> I would have thought that as long as the cabling was uniform with terminators
> equalling that of the cable and coupled to the transceivers via
> transformers with characteristics that there is high impedance as "seen"
> by the cable with low impedance coupling at the line drivers of the
> transceivers, signal transmission should be fine.
> Am I wrong in this assumption?

Not necessarily, If everything gets matched, how you do it is immaterial.

> Or just a rambling dumb ass who has no clue about digital communication
> over coaxial cable?

At the very least, you want to know and you have have the value equation figured out, that puts you ahead of the game.

Regards

cww
 
Top