Wifi better than cat 5 ????

[croak5]

Hello,

I have a home office located in my garage about 7 walls, and 70' from my wifi router (connected to a my DSL). Right now I have a 2 computers on my desk, One connects to the router through a USR wifi adaptor. THe other connects thru a homemade 70' cat5 cable with with 1 junction to a 8 ft cat6 (or 7 whatever) cable.

The wifi conection is about 5-10% faster.

Clues?

Thanx

[Odanez]

so the one connected via wifi is connected even with 7 walls between the pc and the router? serious?

I get signal problems when there are 2 walls between my router and PC

but it could also be that the cat5 cable has some signal degradation, as you said it is home made, maybe the length of it is the problem. Have you tried jsut taking the computer to the router and using another short cable?

[Wicker]

In theory, It should be faster because the maximum length for a cat5 is 100m (without a repeater) but you are plenty within that. I would look to buy a cable that is 70' rather than using a home made one.

[croak5]

Ya 7 or so walls, I even think it goes through my fridge to boot, I do have a little desktop antenae but thats about it. The little USR signal strength shows "POOR 20%"

You should be able to get through 2 walls easily unless theres somthing weird in your walls. Allso, somtimes if you mix up brands you can have problems.

I've wondered about my cable and the fact that its cat 5, but I kinda figured it either works or not, nothin in between. moveing my computer closer to the router is tuff because right now it's a mother board sitting on my desk with drives and the like scattered around it.

THanx

[Odanez]

well you could check a few things:

check whether the lan connection is really running at 100mbps, see if it's running full duplex (dunno how to check that myself),

and see if the connection logs any package loss. maybe even ping the router and see if that runs fine.

as for the wireless: I had the problem with 2 walls when using a usb wireless adaptor. my laptop has a good connection through 2 walls. but 7 walls seems somewhat much, unless your walls are not made of cement

20% signal strength and you still get a good transfer? if my signal strength is below say 40%, it can happen occasionally the the connection is dropped.

you must have a very good wifi setup

[croak5]

Originally Posted by Wicker

In theory, It should be faster because the maximum length for a cat5 is 100m (without a repeater) but you are plenty within that. I would look to buy a cable that is 70' rather than using a home made one.

Hmmm, I do happen to have a 100 ft CAT 6 (OR 7, whatever is the next notch up) store bought cable. Would the cat 6/7 be better? even if its longer (100 ft)?

As to " In theory, It should be faster ", my little strenght meter on the wifi computer says "20% poor". It would seem to be more than a theroy. I always thought wifi was a lazy mans (me) compromise...no?

Thanx

[drake85]

Cat 6 is better than Cat-5e and cat-5 due to more twists per meter. Cat-5e is better than cat-5 due to more twists per meter and a higher hertz rating. As Wicker said, the max length of Cat-5/e is 100m. I believe that will also apply to Cat-6. You should be able to get a descent connection even though it is home made, unless you have bad cable (actual cat-5 and not cat-5e) or you have poorly terminated ends. I use all home-made cables around my apartment and don't have a problem (except the occasional router issue... **** linksys). Look at the ends of your cable and make sure you're following EIA/TIA 568B specs on both ends (Wiki Linky). If it's not, fix it as that can cause serious cross-talk issues. Also, and this would apply no matter what the situation, make sure the cable isn't running close to any florescent lights or other things that create a large magnetic field. Magnetic fields destroy the packets at the physical layer, so a lot of retransmits will occur. Hope some of this helps.

[Taichan]

lots of answers

you only have a 10mbit network connection

you have a PreN mimo wireless network

i'd need to know all the hardware your running on the network end to tell you more

i know wiht my network my 100mbit connection is about 4x faster my 54g line (i think there an issue with my laptop)

[MrObvious]

Tai hit it on the nose probably. You are maybe using a newer wireless standard called 802.11n. It's supposed to run 2x the speed of .11a & g. Also it could only be running at 10Mbps 10BaseT Ethernet instead of 100. I'd check the settings on the router and your NIC. You can change the NIC settings through device manager.

Also, the cable could be bad. Check if you are losing packets. The best way would be to do some pings to your router continuously. Play with ping options and try sending bigger packets than the 32 bytes. I think I've used ping ip address -t 1000 -l 1000 or something like that.

[croak5]

Originally Posted by Taichan

lots of answers

you only have a 10mbit network connection

you have a PreN mimo wireless network

i'd need to know all the hardware your running on the network end to tell you more

i know wiht my network my 100mbit connection is about 4x faster my 54g line (i think there an issue with my laptop)

Umm, I guess the most I can say is that it's all 54 g & mostly U.S. robotics & all about 1.5 yrs old. The built in network on my new gigabyte ds3 is "marvel Yukon" which as I recall, had something about being 10/100 (something like that).

As you can tell, i'm pretty cluless on networking.

Croak

P.S. I know I've done it before but how do I "ping my network and what am I looking for...a pong?

[onioneater36]

The homemade cable is prolly your problem if you are pretty clueless on networking as you have stated. It is easy to make a CAT5 ethernet cable that works. It is also easy that makes one that will pass certification (IF YOU KNOW HOW). Certification will pass numerous frequencies thru the cable and measure crosstalk. My guess is your cable will no pass cert and is therefore bottlenecking your connection (higher error rate than normal). Search for the EIA-586A standard. Here is an OK link, but there are better ones out there.

http://www.vpi.us/standard.html

Most people make the 1-1,2-2,3-3,4-4,etc cable but are unaware of the special twist that must be done to separate the pairs. Your colors should be...

1-white w/green
2-green
3-white w/orange
4-blue
5-white w/blue
6-orange
7-white w/brown
8-brown

notice it goes light,dark,light,dark the whole way thru and that the orange wires are split apart on 3 and 6. Also when crimping the connector on, make sure that the wire leads themseleves are short enough that when the "spear" comes down on the cable, it is hitting the insulation jacket of the cable. There should be no wire outside of the connector.

You can buy equipment to cert the cable. It usually runs around $12,000 for a good one and I'd guess you don't want to do that.

But if you follow these instructions (if your cable is made this way) it will most likely cert anyway unless you got some bad wire which does happen too.

Good luck.

[drake85]

Originally Posted by onioneater36

The homemade cable is prolly your problem if you are pretty clueless on networking as you have stated. It is easy to make a CAT5 ethernet cable that works. It is also easy that makes one that will pass certification (IF YOU KNOW HOW). Certification will pass numerous frequencies thru the cable and measure crosstalk. My guess is your cable will no pass cert and is therefore bottlenecking your connection (higher error rate than normal). Search for the EIA-586A standard. Here is an OK link, but there are better ones out there.

http://www.vpi.us/standard.html

Most people make the 1-1,2-2,3-3,4-4,etc cable but are unaware of the special twist that must be done to separate the pairs. Your colors should be...

1-white w/green
2-green
3-white w/orange
4-blue
5-white w/blue
6-orange
7-white w/brown
8-brown

notice it goes light,dark,light,dark the whole way thru and that the orange wires are split apart on 3 and 6. Also when crimping the connector on, make sure that the wire leads themseleves are short enough that when the "spear" comes down on the cable, it is hitting the insulation jacket of the cable. There should be no wire outside of the connector.

You can buy equipment to cert the cable. It usually runs around $12,000 for a good one and I'd guess you don't want to do that.

But if you follow these instructions (if your cable is made this way) it will most likely cert anyway unless you got some bad wire which does happen too.

Good luck.

THAT COLOR PATTERN IS INCORRECT! That is the EIA/TIA 568A end. You only use this on one side while the other is EIA/TIA 568B. For straight through cables, both ends should be EIA/TIA 568B. That is the standard.

EIA/TIA 568B:
1 - White orange
2 - Orange
3 - White Green
4 - Blue
5 - White Blue
6 - Green
7 - White Brown
8 - Brown

[onioneater36]

Originally Posted by drake85

THAT COLOR PATTERN IS INCORRECT! That is the EIA/TIA 568A end. You only use this on one side while the other is EIA/TIA 568B. For straight through cables, both ends should be EIA/TIA 568B. That is the standard.

EIA/TIA 568B:
1 - White orange
2 - Orange
3 - White Green
4 - Blue
5 - White Blue
6 - Green
7 - White Brown
8 - Brown

I was under the impression that you could use either standard (as long as they were the SAME) to make a normal cable and if you needed to make a crossover, they have to be different.

[drake85]

You can do that... but it is highly recommended that you follow the standards. Standards say that you use 568B on both ends for straight through. Also, if you've ever ran a cable run to a punch down block or key-stone jack, you use the 568B on both ends.

[croak5]

It kinda looks like its my cable cat5 + homemade cable + little junction is a recipe for outright failure. I have to admit I was shocked when I plugged it in it worked at all. Since I have a store bought 100' cable (in a bag) I think I am going to just restring it to replace my own old cable.

As to the crimping of my existing cord, I followed the directions to a tee as I recall, however, when I splayed the wires in the correct order and pushed them into the connector (with my connectors any way) they disappeared, so I really couldn’t be sure they were still in the correct order.

For the record, it appears I do indeed have 100/10 mps router and onboard network card (in addition to the wifi 54g stuff on my net). Still have lingering questions.

1) What difference does it make the order of the colors wires iin the conectors as long as there the same on both ends (just curiosity)?

2) Why is a keystone better for a junction than a simple inline junction thingy?

3) Is the 100 meter max length a point where signal quality drops of precipitously or is does the signal quality degrade at a steady rate along the length of the cable (and 100 meters is just where it becomes untenable) ?

Anyway thanks for all your input.

Regards, Croak

[NokiaNokia]

wired is always better than wireless. end\

[croak5]

I agree, I just cant see how scattering packets though the muddy air could be as efficient as through nice clean isolated copper pipe. But some people say otherwise...

But here I sit, sided by side computers, wif & wire and the wifi's faster. It's gotta be in my cable.

Croak

Additional Comment:

That great just the type I'm glad to here from to confirm what I condsidered conventional wisdom.

If your still on line, can I pick your brain for a few?

Croak

[AntiNazi]

Originally Posted by croak5

1) What difference does it make the order of the colors wires iin the conectors as long as there the same on both ends (just curiosity)?

3) Is the 100 meter max length a point where signal quality drops of precipitously or is does the signal quality degrade at a steady rate along the length of the cable (and 100 meters is just where it becomes untenable) ?

1. its twisted pair for a reason... (this is probably half your problem as if you are asking this question you probably just color matched and didn't do it right)

3. the signal quality drops at your inline junction, and it would seem obvious that the signal drops gradually over the length of the cable. What would make it suddenly drop at 100m?

[croak5]

Originally Posted by AntiNazi

1. its twisted pair for a reason... (this is probably half your problem as if you are asking this question you probably just color matched and didn't do it right)

3. the signal quality drops at your inline junction, and it would seem obvious that the signal drops gradually over the length of the cable. What would make it suddenly drop at 100m?

Look, I'm not a dolt, I just haven't comited my life to networks, Thats why I'm ask questions from people who have, like yourself

1) ((its twisted pair for a reason...)) Reply:

asuming you're right I'm surprised it works at all In other words I haven'y a clue why pairs are twisted in the first place...sheild each other or somthing...


3) ((the signal quality drops at your inline junction)) Reply: why? How much % wise (significant)?


3) ((it would seem obvious that the signal drops gradually over the length of the cable)) Reply:

a)Sure it's obvious,conventional wisdom. But it's also conventional wisdom to me that a single little 1 inch long long juction (assuming good conection) along the lenght of a 100+8 ft cable could make spit of difference. so...Why would it?

3) What would make it suddenly drop at 100m? Again, defies conventional wisdom. But so does a little junction being such a problem.

So to does the theroy of the twisteded pairs. If anything, conventional wisdom would say twisting long lengthes of wire into ropes requires each strand to be significantly longer that the sum of the whole, thus longer wire (s), more signal drop of. How could that be obviously better?


Another defiance of conventional wisdom: Another seemly knowladgablel guy on another forum was down on junctions but said keystones were fine. Keystones have raw wires splayed all over whereas junctions don't. Your take?

I have zero doubt there are good reasons to all this stuff but, at face value, they do infact defi conventional wisdom. Thats why my question may seem lame... right now it's sort of voodoo.


...(I've known plenty of twisted pairs, they all ended in divorce)

Croak

[Steve]

In response to onioneater36 and croak 5,

586 A is prevalent in residential and 586 B is the standard for commercial applications for the most part.
Ether standard can be used but must not be mixed, I do all my work in B unless the customer or plan requests A witch I have found they almost never do.
The main objective is to not untwist the pairs more than 1/2", I try not to go more than 3/8". That includes keystones.

Steve