SiddMartin
Senior Member
- Location
- PA
I was always told that you do not splice data (cat5e / cat6 ) . I never have and never will, but I was wondering why it wouldn't work if you did. Anyone know?
Data Splicing?
- Thread starter SiddMartin
- Start date
- Status
brantmacga
Señor Member
- Location
- Georgia
- Occupation
- Former Child
it would most likely work, but at crippled speeds. if you can keep the twists as close to original as possible, i think you could limp by without much problem. you won't reach gigabit speed, but i think 100Mb would be no problem. I guess it depends on the severity of damage and quality of the splice.
i'm not an expert on it though.
where's hal?
i'm not an expert on it though.
where's hal?
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
090512-2138 EST
Terminate both ends with connectors at the point requiring the splice, and connect with a coupler. This will probably work quite well.
.
Terminate both ends with connectors at the point requiring the splice, and connect with a coupler. This will probably work quite well.
.
brantmacga
Señor Member
- Location
- Georgia
- Occupation
- Former Child
TOOL_5150
Senior Member
- Location
- bay area, ca
CarlAshcraft
Member
- Location
- Orlando, FL
I terminate one cable with a keystone and terminate the other cable with a RJ45 end. Plug the two together giving me a good splice with no data loss.
SiddMartin
Senior Member
- Location
- PA
http://en.wikipedia.org/wiki/Category_6_cable
cat 6 is a specification. splicing of the cables at any point can provide serious degradation to the signal for any number of reasons, so just like nec wiring methods provide for reasonably safe installations, specs like cat 6 provide for an expectation that the performance of the installation will fall within the expected parameters (bandwidth, throughput, attenuation). Adhering to the spec - wire type, terminals, splicing methods, etc - is vital to insure the integrity of the installation.
you can splice it all you want, but it will not be cat 6.
cat 6 is a specification. splicing of the cables at any point can provide serious degradation to the signal for any number of reasons, so just like nec wiring methods provide for reasonably safe installations, specs like cat 6 provide for an expectation that the performance of the installation will fall within the expected parameters (bandwidth, throughput, attenuation). Adhering to the spec - wire type, terminals, splicing methods, etc - is vital to insure the integrity of the installation.
you can splice it all you want, but it will not be cat 6.
gar
Senior Member
- Location
- Ann Arbor, Michigan
- Occupation
- EE
090513-2106 EST
SiddMartin:
The theory of transmission lines says that you should have a load at the destination of the line equal to the characteristic impedance of the line. You also want to drive the line with a source resistance equal to the characteristic impedance. With an ideal transmission line when this is done all energy input to the line is absorbed at the load with no reflection. When there is a mismatch, then energy is reflected causing signal distortion and data errors.
When you create a splice it will probably spread the wires apart and this will change the characteristic impedance at this point. This will cause some reflection of energy. In turn it may cause data errors. In a TCP/IP transfer over Ethernet data errors cause data retransmission and this slows down or inhibits data transfer.
My suggestion is possibly not as good or convenient as the one by TOOL-5150. However, there are couplings made for CAT-6 which implies my technique may be fairly good.
At the following page on my web site are some low frequency waveforms showing the result of very bad line termination resistance:
http://www.beta-a2.com/cat-5e_photo.html see the first photo.
The pulse width I used is long compared to the transit time down and back on the cable. Ethernet data pulses will be very short compared to some typical transit times. In photo P1 the transit time from source to destination was about 0.2 microseconds, and the reflected wave shows up at the source at 0.4 microseconds and is of polarity to add to the original pulse.
If you had shorter pulses in the source data, then the signals would get all scrambled.
In reality your splice might not produce a large amount of reflected energy and maybe not much effect on the signal.
.
SiddMartin:
The theory of transmission lines says that you should have a load at the destination of the line equal to the characteristic impedance of the line. You also want to drive the line with a source resistance equal to the characteristic impedance. With an ideal transmission line when this is done all energy input to the line is absorbed at the load with no reflection. When there is a mismatch, then energy is reflected causing signal distortion and data errors.
When you create a splice it will probably spread the wires apart and this will change the characteristic impedance at this point. This will cause some reflection of energy. In turn it may cause data errors. In a TCP/IP transfer over Ethernet data errors cause data retransmission and this slows down or inhibits data transfer.
My suggestion is possibly not as good or convenient as the one by TOOL-5150. However, there are couplings made for CAT-6 which implies my technique may be fairly good.
At the following page on my web site are some low frequency waveforms showing the result of very bad line termination resistance:
http://www.beta-a2.com/cat-5e_photo.html see the first photo.
The pulse width I used is long compared to the transit time down and back on the cable. Ethernet data pulses will be very short compared to some typical transit times. In photo P1 the transit time from source to destination was about 0.2 microseconds, and the reflected wave shows up at the source at 0.4 microseconds and is of polarity to add to the original pulse.
If you had shorter pulses in the source data, then the signals would get all scrambled.
In reality your splice might not produce a large amount of reflected energy and maybe not much effect on the signal.
.
SiddMartin
Senior Member
- Location
- PA
I now understand.
....maybe I should read it a couple more times though.
Thank you all!:smile:
....maybe I should read it a couple more times though.
Thank you all!:smile:
- Location
- Massachusetts
No doubt it can be spliced, I have done so many times during major remodeling of offices and stores and it has gotten the systems back on line. In those cases all lines where eventually replaced with new complete runs.
We even tied to sales registers into one 'home run' when the demo guys demoed the runs so well we lost one of them entirely. It seemed to work, I don't know how much slower it was.
More recently I have spliced data lines for items like supermarket scales and employee time clocks. These items do not move a lot of data and the splices do not cause any obvious problems.
Before Hal yells at me ... I know they should be complete point to point .... but that is not what the customer wants to pay for.
We even tied to sales registers into one 'home run' when the demo guys demoed the runs so well we lost one of them entirely. It seemed to work, I don't know how much slower it was.
More recently I have spliced data lines for items like supermarket scales and employee time clocks. These items do not move a lot of data and the splices do not cause any obvious problems.
Before Hal yells at me ... I know they should be complete point to point .... but that is not what the customer wants to pay for.
- Status