I have to install an Internet connection at a location somewhat removed from the origination point. The building is very large, spread out, and difficult to run wire through. Presently there already is a connection terminated in an Rj45 at a point that is only used for one week out of every year. All other times there is nothing plugged into this jack but the connection remains active. In order to make the connection to the new location about 100 feet away can I tap off the existing jack and "daisy chain" the wire to the new location and then terminate that wire in a second R45? I now this is not common practice but if only one computer is using the line at any particular time would this work? There will never be a time that two computers will ever be operating on this line at the same time although for that week in question they both may be plugged into it though. would this be an issue? I can arrange to have the new location unplugged during this time if need be. Thanks for any advice. Lenny
yes, but only one can be plugged in and it will be a nightmare for the next guy who tries to figger out why it doesn't work like it should. Put a plug on the extension wire so you can have only one connected.
Stick an ethernet switch at the location and you can have as many plugged in as you have ports on the switch. They're $10 on ebay, free at garage sales. If you walk outside in a populated area and yell, "I need an ethernet swtich," people will come out of the woodwork glad to get rid of them. You only have to have the switch powered for a week.
Mike I really like the idea of a plug on the extension wire. The existing jack i s on a wall in a public building in a single gang lock box fed through 1/2 inch EMT. I can figure a way to arrange this so that the extension runs thr ough another length of pipe into the same box and is then plugged into the existing jack and the box locked for the 51 weeks of the year that it isn't used. That way the remote location will be live for those 51 weeks. When t hey need to use the present location for that one week, they'll simply unlo ck the box, unplug the extension wire, (similar to what you do to test a CO line at a telephone interface Demarc point), thereby freeing up the line d ownstream. They can then connect their equipment without an issue. Great id ea, Thanks, Lenny
Yes, it will work with only one computah connected.
Yes, it's an issue. Twisted pair ethernet is *NOT* a bus arrangement as in POTS phone lines. It's a star. Just add a cheap 5 port ethernet switch where everything comes together and you'll be fine.
No need with an ethernet switch.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
Ethernet switches can also be found on the back of (unwanted) DSL modems found on the shelves of second hand stores. Just put some black tape over the blinking "DSL Fail" LED and you've got yourself an inexpensive ethernet switch. Many DSL Modems come with 4-way switches. I did that with an ActionTec DSL Modem for a few years, before I found a 10/100
5-Port LinkSys for $5 on a shelf at a second hand store.
HTH Jonesy
--
Marvin L Jones | jonz | W3DHJ | linux
38.24N 104.55W | @ config.com | Jonesy | OS/2
* Killfiling google & XXXXbanter.com: jonz.net/ng.htm
I just ripped out almost exactly the same great idea and replaced it with a proper ethernet switch.
No DSL modem comes with more than one ethernet port. The modem is a bridge between DSL and ethernet and only requires one port each. However, a DSL *ROUTER* will have a built in ethernet switch as you describe. However, it will also have a DHCP server, NAT, RIP, and a mess of other acronyms. If you don't turn these off on the configuration page, you're going to have problems on the network. For example, having a second DHCP server running on your LAN will create a problem when a workstation decides to get its DHCP assigned IP address from your DSL router, and finds the gateway address pointing to a device that goes nowhere.
In my case, the DSL router was actually setup correctly with all the extra services disabled. It also was working well for about a year. Then, someone punched the reset button on the back of the DSL router, and the problems started. Since nobody bothered to tell me that they had added this device to the network, it took some sniffing to find it.
Maybe some more black tape over the reset switch hole? Or, just use the correct piece of hardware in the first place.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
It would probably work OK when you connect something to the jack at the end of the new run of wire.
It may very well *not* work, for devices plugged into the existing jack, during that one week per year.
The problem is that your new run of wire will create an "un-terminated stub" in the Ethernet transmission line. This is a no-no - signals travelling down the wire (from the middle jack to the unused new end jack) will arrive at the end jack, and will "see" an abrupt change in the characteristic impedance of the line (jumping up from 100 ohms or so in the twisted pair, to a very high impedance at the jack). The signals will be reflected by this impedance discontinuity, and travel back along the wires towards the existing (middle) jack. When they arrive there, they will mix with (and interfere with) the signals from the other end of the line, and it's very likely that the computed plugged into this jack will see a completely garbled packet as a result and will discard it.
This is essentially the same problem which can create "ghosts" in an analog TV system, if you have an extra length of coax daisy-chained from a coax port.
There are several ways to deal with this problem effectively.
The one I'd recommend (inexpensive, easy, and it should comply with all of the standards and work fine): convert the existing outlet plate from one RJ-45, to two RJ-45 jacks. Run the "tap" line to the new location to the second RJ-45 on the new jack... do *not* hard-wire it into the existing line or jack!
When you want to use the original location, plug the computer into RJ-45 #1 (the one which goes to your central location). When you're not using this location, and want to use the new location, simply plug a short Ethernet jumper cable into the two RJ-45 jacks.
This approach can be extended to allow you to use both locations, at low expense. Simply buy an inexpensive 10/100 switch (4- or 5-port type) and place it at the existing location. Connect its upstream port to jack #1. Connect jack #2 to one of its downstream ports. When you want to use a computer at this location, simply connect it to one of the other downstream ports on the switch... no need to disconnect your new location.
--
Dave Platt AE6EO
Friends of Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!
On Monday, October 1, 2012 2:44:30 PM UTC-4, Dave Platt wrote:
from the origination point. The building is very large, spread >out, and d ifficult to run wire through. Presently there already is a >connection term inated in an Rj45 at a point that is only used for one >week out of every y ear. All other times there is nothing plugged into >this jack but the conne ction remains active. In order to make the >connection to the new location about 100 feet away can I tap off the >existing jack and "daisy chain" the wire to the new location and then >terminate that wire in a second R45? I n ow this is not common practice >but if only one computer is using the line at any particular time >would this work? It would probably work OK when you connect something to the jack at the end of the new run of wire. It may ve ry well *not* work, for devices plugged into the existing jack, during that one week per year. The problem is that your new run of wire will create an "un-terminated stub" in the Ethernet transmission line. This is a no-no - signals travelling down the wire (from the middle jack to the unused new en d jack) will arrive at the end jack, and will "see" an abrupt change in the characteristic impedance of the line (jumping up from 100 ohms or so in th e twisted pair, to a very high impedance at the jack). The signals will be reflected by this impedance discontinuity, and travel back along the wires towards the existing (middle) jack. When they arrive there, they will mix w ith (and interfere with) the signals from the other end of the line, and it 's very likely that the computed plugged into this jack will see a complete ly garbled packet as a result and will discard it. This is essentially the same problem which can create "ghosts" in an analog TV system, if you have an extra length of coax daisy-chained from a coax port. There are several w ays to deal with this problem effectively. The one I'd recommend (inexpensi ve, easy, and it should comply with all of the standards and work fine): co nvert the existing outlet plate from one RJ-45, to two RJ-45 jacks. Run the "tap" line to the new location to the second RJ-45 on the new jack... do * not* hard-wire it into the existing line or jack! When you want to use the original location, plug the computer into RJ-45 #1 (the one which goes to y our central location). When you're not using this location, and want to use the new location, simply plug a short Ethernet jumper cable into the two R J-45 jacks. This approach can be extended to allow you to use both location s, at low expense. Simply buy an inexpensive 10/100 switch (4- or 5-port ty pe) and place it at the existing location. Connect its upstream port to jac k #1. Connect jack #2 to one of its downstream ports. When you want to use a computer at this location, simply connect it to one of the other downstre am ports on the switch... no need to disconnect your new location. -- Dave Platt AE6EO Friends of Jade Warrior home page:
formatting link
I do _not_ wish to receive unsolicited comme rcial email, and I will boycott any company which has the gall to send me s uch ads!
Thanks for all the great ideas guys. I really appreciate them. Lenny
Your posts indicate intelligence, are logical and all that, but this one time you erred. Claimed a negative. You know better.
Two 2WIRE modems I know (and actually did the setup) have four RJ45s as well as four channel 802.11. (g I think)
Of course you could argue that these are routers or whatever, but really fastforwarding through the argument which would be fruitless anyway, how different is it ? Isn't the DSL MODEM just a router or switch that connects to a "slightly" bigger network ?
Two ways of thinking about it but I think neither is wrong.
Give me a 2wire model number and my guess(tm) is that it will be a router inside. All (and I do mean all) 2wire devices that have built in wireless also have a router inside. Marketing may call is a "modem", but if it has a router inside, it's really a "DSL router".
fastforwarding through the argument which would be fruitless anyway, how different is it ? Isn't the DSL MODEM just a router or switch that connects to a "slightly" bigger network ?
Not bigger network, but different network. The definition of a router is a device that connects two DIFFERENT networks at the IP layer (ISO layer 3). One port is connected to the greater internet network. The other port is connected to a local area network, that uses non-routeable IP addresses.
My definitions, which might be different than yours:
DSL Modem: DSL (actually ATM) to ethernet bridge. Everything done at the MAC layer (ISO Layer 2) with no involvement with IP layer (ISO layer 3) except for configuration management.
DSL Router: Contains a DSL modem where the ethernet port is connected to an internal router. Always includes NAT/PAT to allow a single IP address to serve multiple non-routeable IP addresses on the LAN. The modem works on the MAC layer, while the router works on the IP layer. A 5 port (yes 5 ports, not 4) ethernet switch on the router output allows connections with additional devices. If an optional internal wireless access point is included, it connects to the 5th ethernet port allowing additional connections.
Not the greatest definition, but I think it will suffice.
Only two ways? In the distant past, I tried to reconcile the various names for wireless ethernet bridges and gave up. Too much naming creativity in an industry that can't even get NAT and PAT correct. With that much creative naming, I don't want to even think about cleaning up the descriptions of the various wireless access point, routers, bridges, gateways, switches, media servers, etc.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
The real issue is line length to the next device. 100 meters or 320 feet depending typical local units. Beyond that function is NOT guaranteed by the standards. Non-compliant cables, connectors or other connections may interfere with reliable as well.
I mostly agree with your distinction on the difference between modem and router.
I do not think that what is appearing at the DSL end user terminals is ATM but closer to VT45 or VT135 with highly compressed data. Moreover you are looking at data likely being a shared service with a much more broadband (TV) type service on the same pair.
It's an oversimplified definition, which falls apart on devices like a brouters (bridge-router) and using a DSL modem to simultaneously connect to multiple services. Simple definitions never seem to stay simple.
Ok, you got me. What is V45 and V135? Google wasn't particularly helpful by suggesting that VT meant vacuum tube. Are you thinking of AT&T U-Verse IPTV service which uses VDSL? If so, I know nothing because AT&T doesn't offer it in my area.
The DSL modem with diagnostics include "ATM Ping" which should be a clue. Articles on how DSL works always mention ATM as the underlying virtual circuit mechanism to connect to the DSLAM (ATM switch) which does the IP packet reassembly from the tiny ATM pieces. See section "ADSL and ATM".
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
VT45 is the payload portion of an STS-1 (about equal to a DS-3), likewise VT135 is the payload portion of an STS-3.
Thanks for the link. It ties up a lot of loose ends for me. But the signal coming down the wire to my modem-router does not seem to include the ATM overhead.
"Give me a 2wire model number and my guess(tm) is that it will be a router inside. All (and I do mean all) 2wire devices that have built in wireless also have a router inside. Marketing may call is a "modem", but if it has a router inside, it's really a "DSL router". "
It's a 2WIRE RG2701HG-00. One link calls it a modem and switch, another a m odem switch/router and another a modem/router.
Talk about definitions.
In the old days I got the gist of electronics, and I mean to where I could do a little bitt of designing. That was all analog. Even then some of the d efinitions were not memorized by me, like even though I know the names Colp itts and Hartley I couldn't tell you the (dis)advantages of either right no w. I can look it up of course but I really only committ(ed) to memory what I need(ed).
I see no difference between a switch and a router, but of course there is.
What I see is that a router can have a hardware firewall, but I've never se en it on a switch. I may need to know these things soon if I go to get the network at work running again.
Note that the article was from 2001. That was intentional as improvements and additions to DSL have added some complexity.
ATM overhead is the "ATM Tax". With 5 extra bytes of header for every
53 bytes of payload, that's: 5/53 = 9.4% of overhead. The ATM Tax and the additional header overhead for various layers totals about 15% which is why a 1.5Mbit/sec DSL line actually delivers only 1.3Mbits/sec.
The header include the VCI (virtual circuit identifier) so that the packets can be disassembled into multiple VC's (virtual circuits), thus allowing a single DSL modem to support multiple IP data streams. DSL could theoretically connect to multiple ISP's at the same time, over the same phone line, but AT&T doesn't allow that in the US.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
I have a pile of 2701HG wireless routers in the office. The routers work fine, but the power supplies suck. Fortunately, they're easy to crack open and repair. When yours quits, reboots erratically, or acts weird, crack open the PS and fix it rather than tossing the router.
Incidentally, try: on your 2701HG.
The 2701HG routers have an additional problem in that it is impossible to update the firmware. I've identified about 12 different firmware versions with little clue as to the differences. Now that 2-wire is part of Pace, they won't even talk to me about firmware. Fortunately, the most disgusting problem (failure to reconnect after connecting to another access point with Apple iPads) isn't very common. Also, there's no way to save the restore the settings, so everyone needs to be setup from scratch. Still, I kinda like them (probably the white color).
Anyway, if you want to be technically correct, the description has to include all the major features. So the 2710HG would be a: "DSL modem wireless firewall IP router and switch". Somehow, I don't think marketing would like that.
The old daze are gone. I read sci.electronics.repair and must admit that I have never heard of some of the devices, acronyms, nomenclature, software, etc. So, I use Google to search for clues, read what I find, and sometimes actually learn something new. I suggest you do the same.
Easy. A switch works at the MAC address layer. It directs packets based on the destination MAC address. No IP addresses are involved.
A router connects two different IP networks and works on the IP layer. It directs packets based on the destination IP address. Although the MAC layer is involved, no MAC addresses are passed through a router.
All 802.11 wireless works on the MAC address layer. Packets are directed by the destination MAC address layer. With an encrypted wireless network, these MAC addresses are sent unencrypted and can be easily sniffed. However, the relevant IP addresses, encapsulated and encrypted inside 802.11 packets, are not sniffable.
Also, a bridge is a two port ethernet switch.
Firewalls work on the IP layer and therefore require a router.
Groan. There are better newsgroups and forums to obtain information on networking.
--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.