Pros & cons of networking design & machine comps.

[dubliner]

Since I got my machine I only transferred files via usb stick, well the other day I transferred them from design to SB on the same computer. I thought I hadnt hit the button it was so fast. Now I'm thinking of ditching the thumb drive method. I've run CAT6 to the machines in preparation but is there a lot involved in setting up a network from a design computer to 2 Shopbots?

[lstovall]

Hi Neville,

I have mine set up that way. You just need a router and it's pretty painless to set up. Since you already have the cable run, you wont need any of those wireless adapters. You can do it via the Network Setup Wizard in the Control Panel. I run three computers and a printer from mine. Half wireless and half wired.

[dubliner]

Hey Leon ( hope youre well ) Thanks thats what I was hoping to hear. I have a spare Netgear wireless router, I'll look into hardwiring that and setting up a network. Stay close to your phone :-)

[garyb]

Ah Neville welcome to the 21st century

I run both my machines across the network from my server (170') works just fine

[lstovall]

Neville,
Thanks, I'm doing fine. You will like it. Get's you away from using a thumb drive to transfer files.

Hi Gary, how's it going, have you unloaded the Ark yet or is it still raining?

[knight_toolworks]

it's easy if they both are xp and your on the same router or hub. open your my networks click on setup home or small network. then make sure you do this on both computers making the workgroup name the same.
then you go to the folder you want the files on and choose sharing and security. choose network sharing and make sure both boxes are checked. name it what you want. after that's done when you refresh network places on your non shopbot computer you should see that folder. open it up and drop files into it. I just use vcarve to save files in that folder. one thing that annoys me you have to refresh the view to see new files sometimes.

[garyb]

Hey Leon, hope things a good donwn in sunny FL.
finally stopped raining so put the ark project on hold, now jumped into the frying pan 93° 94° for the last few days. My kind of weather

[harryball]

The theoretical limit for wired eithernet on a UTP (unshiedled twisted pair, the standard stuff) is 333 feet.

The only issue I had was a network card that would cause an interrupt in the system that in turn caused a stall or hickup in the bot. I would disable the card (right click, disable on the network icon, systems may vary) to solve the problem. With the newest network card I'm using this is not a problem.

/RB

[dana_swift]

R Ball -- Just curious where does the 333 foot number come from?

D

[garyb]

Comes from the cat6 specs

"The maximum allowed length of a Cat-6 cable is 100 meters (330 ft)."

[dana_swift]

I looked up the 802.2 spec and indeed is specifies 100 meters. I am always curious to know what happens when a cable is 101 meters long? Or 102? Does the system slowly degrade, of fail to work suddenly?

What fails? When does the CD part of CDMA stop working? Is it the LVDS? Does accumulated impedance in the differential pairs cause the eye diagrams to collapse? Can a hub be placed at that point and then another X distance achieved? I have seen some huge 100base T networks in some of the local businesses, and some of them make their cables from bulk, I suspect they have violated the magic "100m" number at some point. The systems seem to work just fine.

I have been told it has to do with the PHY/MAC layer timeouts, but if that were true why would optical cable have a different limit? The speed of light is almost identical for both media. Both optical and UTP use the same PHY/MAC protocols. (Except optical uses 8b/10b encoding while 100base-T uses two 4b/5b groups for the same thing)

I remember the days of early RS-232 when much voodoo was bandied about. Cables had to be "just so" or horrible things were rumored to happen. Baud rates were sacred cows. Then I discovered systems that were FAR outside the specs and they worked flawlessly, I had to find out why. It turned out the specs were arbitrary. Now people pay little attention to the things that once were thought immutable.

Ethernet smells the same way, especially given the "cable tap specifications" of the first versions of coaxial Ethernet. At a whopping 1 megabit per second! That turned out to be ignorable.

IEEE 802.x was approved by a committee of folks who just needed a number they could agree on, all too often just because they needed to come up with "something". Perhaps the real limits imposed by physics is some number such as 87376.23221 meters (I just made that one up) then 100m would have a "safety factor" of 873.76, presumably causing very few packets to get lost bits and require re-transmission. (ALL IEEE 802.x networks detect garbled packets on reception)

Bringing it back to the Shopbot.. if you need to go 120m, my suggestion is run the cable. I will bet you a six pack of your favorite diet soda it will run just fine.

If not, I will happily spring for the soda in exchange for the education!

D

--- afterward

From the site:
http://www.yale.edu/pclt/COMM/ETHER.HTM

I excerpt.. its a good read (not just for techno junkies like me)..

"In simpler days, when Ethernet was dominated by heavy duty coax cable, it was possible to translate the 50 millisecond limit and other electrical restrictions into rules about cable length, number of stations, and number of repeaters. However, by adding new media (such as Fiber Optic cable) and smarter electronics, it becomes difficult to state physical distance limits with precision. ...

Note the last sentence.

D

[harryball]

Actually... at 333' and 2" (they are nice enough to give a 1" leeway) the cabling police will show up and arrest you. They are closely related to the matress tag enforcement division. :-)

You found one reason. Basically, the time required to transmit the signal on the copper would be greater than the time the ethernet protocol decides there is an error and restransmits. I have actually tried to run a cable about 450' and the network traffic indicator light was almost solid... but no traffic was going through. With some adjustments I was able to get that distance down to about 350' and it was functional.

The other reason... if you go over 100m (333ft) your cabling with not certify. If you are working in a datacenter that makes it useless.

In the end, no 100m is not the magic "go no further" number. The actualy limit will depend on the quality of the cable, the temperature, the number and degree of bends in the cable etc... It's actually very interesting when you realize that even staples along the cable will affect trasnmission speed and quality.

/RB