Originally posted by: hanoverphist
each pair is twisted a bit differently. as long as you split the inner pairs, i bet it would work fine. usually. back in win98SE days you could put any 8 pins in a row, match them on the other side and it would work fine. XP changed that, and made me have to actually remember o/s-o-g/s-bl-bl/s-g-br/s-br
i follow the standards because i need to guarantee it will work on the system i make for my customers. if it doesnt work and my order is different, then they will blame the nonstandard cables.
Originally posted by: pontifex
Originally posted by: hanoverphist
each pair is twisted a bit differently. as long as you split the inner pairs, i bet it would work fine. usually. back in win98SE days you could put any 8 pins in a row, match them on the other side and it would work fine. XP changed that, and made me have to actually remember o/s-o-g/s-bl-bl/s-g-br/s-br
i follow the standards because i need to guarantee it will work on the system i make for my customers. if it doesnt work and my order is different, then they will blame the nonstandard cables.
what does the s mean in your example? isn't it usally white and a color and a solid color?
Originally posted by: pontifex
Originally posted by: hanoverphist
each pair is twisted a bit differently. as long as you split the inner pairs, i bet it would work fine. usually. back in win98SE days you could put any 8 pins in a row, match them on the other side and it would work fine. XP changed that, and made me have to actually remember o/s-o-g/s-bl-bl/s-g-br/s-br
i follow the standards because i need to guarantee it will work on the system i make for my customers. if it doesnt work and my order is different, then they will blame the nonstandard cables.
what does the s mean in your example? isn't it usally white and a color and a solid color?
Originally posted by: spidey07
If you follow the standards the cable adheres to verifiable and reliable electrical characteristics. If you have access to a cable scanner you can actually certify the link (patch cable, jack, horizontal cable, jack, patch cable) and can guarantee maximum performance and reliability. If you don't have access to such a scanner you don't really know if you have any category rating...you've got category nothing. This is why making patch cables are highly discouraged because most all of the time they fail a category certification test.
Also as mentioned the twists are different per the standard for a reason so you need to keep the color either 568a or 568b, 568b is used in the US. Category 6 cable and gigabit/10 gigabit ethernet needs this variation in twist and counter twist down the length of the cable.
Also if you just match color on each end you wind up what is called a "split pair". That pair of wires needs to be in the correct pin position or else you get very unusual performance and just weird stuff.
So yes, it's REALLY important to properly follow the standards if you want a network instead of a net-kinda-works.
Originally posted by: pontifex
Originally posted by: hanoverphist
each pair is twisted a bit differently. as long as you split the inner pairs, i bet it would work fine. usually. back in win98SE days you could put any 8 pins in a row, match them on the other side and it would work fine. XP changed that, and made me have to actually remember o/s-o-g/s-bl-bl/s-g-br/s-br
i follow the standards because i need to guarantee it will work on the system i make for my customers. if it doesnt work and my order is different, then they will blame the nonstandard cables.
what does the s mean in your example? isn't it usally white and a color and a solid color?
Originally posted by: torpid
Wow, I'm old. The last time I made a cable was when it didn't matter as long as the colors matched up.
Originally posted by: yelo333
I think you need to keep the twisted pairs together or you may get interference with longer cables. It's hard enough to make a cable that performs to spec; why make it harder on yourself?
Originally posted by: Kabrinski
This I would recommend just to make sure interference stays at a minimum. But if you went Br, WBr, G, WG, O, WO, Bl, WBl (for example) on both ends, it will work fine.
Originally posted by: torpid
Wow, I'm old. The last time I made a cable was when it didn't matter as long as the colors matched up.
Originally posted by: Kabrinski
Originally posted by: torpid
Wow, I'm old. The last time I made a cable was when it didn't matter as long as the colors matched up.
It still doesn't, so long as you match the colors from one end to the other for straight through or switch the proper wires for cross-over. I wouldn't recommend it for professional use due to standards, but if you are making a few new cables for your home PC, it doesn't matter.
I made a couple cables not long ago like this and they worked fine.
Originally posted by: yelo333
I think you need to keep the twisted pairs together or you may get interference with longer cables. It's hard enough to make a cable that performs to spec; why make it harder on yourself?
This I would recommend just to make sure interference stays at a minimum. But if you went Br, WBr, G, WG, O, WO, Bl, WBl (for example) on both ends, it will work fine.
Originally posted by: ScottMac
The color pattern you describe will suck for any full duplex, and especially suck for 100 meg full duplex, by virtue of the split pair on pins 3&6 jacking up the crosstalk in the cable to seriously ugly levels.
Originally posted by: Raduque
Originally posted by: ScottMac
The color pattern you describe will suck for any full duplex, and especially suck for 100 meg full duplex, by virtue of the split pair on pins 3&6 jacking up the crosstalk in the cable to seriously ugly levels.
Can you explain how this works or at least link to something? I always thought it didn't matter what order the colors were in, as long as the twisted pairs stayed twisted and matched up correctly at the far end of the cable.
Is there something special about each wire that the colors dictate?
Originally posted by: ScottMac
Originally posted by: Kabrinski
Originally posted by: torpid
Wow, I'm old. The last time I made a cable was when it didn't matter as long as the colors matched up.
It still doesn't, so long as you match the colors from one end to the other for straight through or switch the proper wires for cross-over. I wouldn't recommend it for professional use due to standards, but if you are making a few new cables for your home PC, it doesn't matter.
I made a couple cables not long ago like this and they worked fine.
Originally posted by: yelo333
I think you need to keep the twisted pairs together or you may get interference with longer cables. It's hard enough to make a cable that performs to spec; why make it harder on yourself?
This I would recommend just to make sure interference stays at a minimum. But if you went Br, WBr, G, WG, O, WO, Bl, WBl (for example) on both ends, it will work fine.
Please stop posting on matters such as these, you are totally clueless.
"Working" doesn't mean you're getting optimum throughput. For perspective, would you like the mechanic that fixes your (you, wife's, parent's) brakes to do it right, or "probably" good enough?
The color pattern you describe will suck for any full duplex, and especially suck for 100 meg full duplex, by virtue of the split pair on pins 3&6 jacking up the crosstalk in the cable to seriously ugly levels.
So please, just chill and read; avoid the urge to respond to something you know absolutely nothing about.
Originally posted by: spidey07
Originally posted by: Kabrinski
This I would recommend just to make sure interference stays at a minimum. But if you went Br, WBr, G, WG, O, WO, Bl, WBl (for example) on both ends, it will work fine.
That's a split pair. It may work but performance will be degraded and weird stuff can happen (crosstalk). The pairs MUST stay together on pins:
1-2
3-6
4-5
7-8