Communications controller?

Jun 23, 2005 30 Replies

It might not be a problem for end-to-end transmission. However, for stations tapping in the middle of the line. The main line now become branches, which are quite significant (between 1/8 to 1/16 wavelength).

I am sure both ends are open (unterminated). If the OP idea is to make use of unconnected phone lines.

Not a simple signal drive by I2C (open wired-or) for instance.

Higher voltage is needed because of lower impendance at higher frequency.

Different bus configurations, not cable differences.

Exactly where did he say that they should be unc>>I'd like to use simple 4 wire telephone cables (ground, Vcc, 2 communication),

The OP might be willing to consider using CAT5 cabling, but would prefer to use cheaper and thinner, easier to bend telephone cable. I don't see any references to existing phone cables.

The characteristic impedance for coaxial cables and twisted pairs remains constant for a quite large frequency range. At very low (audio) frequencies it is higher and at several GHz some strange waveguide mode will emerge, causing impedance fluctuations. For practical frequencies used in data communication, it is quite safe to assume that the twisted pair characteristic impedance is in the

100-120 ohm range.

Paul

A display, unless a graphic one, requires very little resources, provided it is done with a statemachine. And of course no floating point.

Rene

Ing.Buero R.Tschaggelar - http://www.ibrtses.com & commercial newsgroups - http://www.talkto.net

If he is not dealing with existing wires, forget about phone wires and jacks. Just use cat5 and ethernet (cheaper than CAN).

communication),

He is also asking for multiple taps on the line, similar to the phone wires are connected.

Significantly higher. It is at least several hundred kilo-ohms at DC. For example, to use it for I2C, both signals (SDA, SCK) need to be pulled up by 1.5K resistor. If you terminate both ends with 100 ohms, there is no hope for a high logic state.

beyond a minimum frequency, perhaps 10K Hz.

The problem with CAT5/6 is that it is quite thick and in order to operate at the specified speeds, the bending radius rules must also be obeyed. Thus, doing any pretty surface mounting will become quite hard.

However, using a telephone cable CAN or RS-485 system at a reasonable speeds allows a thin cable with a small bending radius to be used, which makes it possible to do a pretty surface installation.

I guess there might be some difference if there is a 10 cm-2 m branch from the main bus, compared to some 100-1000 m unterminated stub :-).

Paul

ArcNet.

Didn't some guy named Bell already do this?

Yes, but I don't think it ever took off. :-)

-Mike

What are yoy guys talking about ? I am using a Agent news reader, but without any context this message makes no sense.

Anton Erasmus

Sounds like a design deficiency in your newsreader.

The OP wanted to come up with a design for a voice communications device with various over-kill features. It boiled down to a telephone. One requirement was that it would recognize devices being added or removed from the "net" and reconfigure. ArcNet does that. (there are versions of ArcNet that are fast enough to do what he wanted to do)

Why do you think so? This thread only has 4 messages. I presume that it came from another group but there is no hint as to where.

The way Usenet distribution works means that delivery of messages cannot be guaranteed: each post needs to stand on its own. Google is not the answer because there is no guarantee they have the messages either.

Paul

David K. Bryant wrote: [snip]

The problem is that the thread you responded to was nine months old. You shouldn't assume that everyone is on a newsserver that stores threads that old, nor that people still have it in their local news files.

OK, using Google to catch up is always an option.

/Henrik

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required