One of the more popular experiments that users with multiple PC’s will try is the installation of a home network or LAN. LAN is geekspeak for Local Area Network – the correct term applied to any network installed in a limited physical area. In the not-too-distant past, it was common practice to install home LAN’s using coaxial cable, as it was simple to install, readily available prefabricated in standard lengths, and it required no additional components besides the Network Interface Card (NIC) and some hardware – tees and terminators. The alternative was to use unshielded twisted pair (UTP) cable, usually called 10Base-T cable. This cable resembles standard telephone cable found in most homes today, except that it is heavier, having four pairs of wires instead of two pairs. Because of the additional four wires, the modular plug used with 10Base-T is an eight-conductor device, which is known as an RJ-45 connector.
Using 10Base-T with more than two PC’s requires the installation of an additional connectivity device. This device can be a hub, a switch, or a router, depending upon the particular requirements of the LAN. All of this means that the user will be installing multiple connecting cables of varying lengths, and possibly of varying circuit schemes. In general, there will be one cable connecting each PC to the hub. Additional cables may be used to daisy-chain multiple hubs together, or to connect the hub to a broadband modem, or to connect a second NIC in certain PC’s to another LAN. 10Base-T cables are also available prefabricated in standard lengths.
The choice between using prefabricated cables and “rolling your own” in a matter of personal preference. Some users will choose to simply purchase prefabricated cables, plug them in, and hopefully have an operable LAN. Others will decide to assemble the LAN cables as they are needed, which provides the option of having cables that are exactly the right length. An added benefit is found when actually pulling the cables into place, as the cable sans connector will not require large holes to be drilled in walls, ceilings, or floors. Depending upon the number of cables required, overall costs may well even out, as the cost savings realized with bulk cable will be reduced by the cost of the necessary cable assembly tools. Radio Shack offers an RJ45 crimper for around $30.00 (p/n 279-405), a BNC (coax) crimper for around $16.00 (p/n 278-238), and a coax stripper for about $12.00 (p/n 278-248). The UTP jacket stripper is usually built into the crimper tool assembly. Of course, prices for the necessary tools will vary based upon supplier and area. I mention the Radio Shack items only because they are available in retail stores across the country, and this gives the reader a basis for comparison both in terms of price and tool design.
OK – I said earlier “…hopefully have an operable LAN”. By now, you are probably wondering what I meant by that remark, right? Well, what happens when the LAN doesn’t perform as expected – or maybe doesn’t even seem to connect at all? Obviously the time is at hand for some basic LAN troubleshooting… and troubleshooting will sometimes require test equipment.
The purpose of this article is to help you create a simple but reliable 10Base-T cable tester that will work with any continuity tester or ohmmeter.
A Word on Ohmmeters
It is my humble opinion that a decent multimeter be a permanent resident in the toolkit of anyone who is going to be doing any kind PC testing and repair. There are many good meters available at very reasonable prices today. In keeping with the Radio Shack product suggestions, I would recommend their model 22-163 as an excellent general-purpose multimeter for around $50.00. Those who are looking for more bells and whistles should consider model 22-805, which offers a PC interface and sells for about $60.00, or model 22-174, which offers temperature sensing, a frequency counter, and true RMS read-out for about $90.00. For what it’s worth, the $50.00 model is in my service call toolkit. Make sure that whatever multimeter you use is of at least that quality, as a poorly designed or built multimeter will provide erroneous readings to the user. There’s no real need to go into the theory behind this – just accept my word on this, as it is fact.
A Word on Cable Types and Nomenclature
There are a few somewhat ambiguous terms that are frequently heard when discussing network cables. These include CAT 3, CAT 5 , 10Base-T, and 100Base-T. In order to take some of the confusion out, let’s take a quick look at what these terms mean.
I have already used the term 10Base-T. This term is used to indicate the character of the LAN, with the “10” referring to the maximum transmission speed of 10Mbps. The “Base” portion of the designation refers to “baseband signalling”, and is used to indicate that the medium is carrying Ethernet signals only. The “T” suffix indicates that the transmission medium is a twisted-pair wire.
The term 100Base-T is a similar term. In this case however, the “100” indicates a maximum transmission speed of 100Mbps, and the “T” suffix will sometimes be seen with an additional digit tagged on. The additional digit is used to further describe the medium type, with a “4” indicating telephone-grade twisted pair wire, and an “X” indicating data-grade twisted pair wire.
The twisted pair wire and associated connectors and junctions used in the transmission of network signals are rated as regards the maximum data rates that they can “sustain”, or carry continually. The American National Standards Institute/Electronic Industries Association (ANSI/EIA) Standard 568 specifies “categories” of these devices, based upon their performance capability. A component that is labeled as a CAT 3 component is one that meets the Category 3 minimum performance requirements. The pertinent categories for Ethernet cabling are CAT 3 and CAT 5. CAT 3 components must be able to sustain a data rate of 16 Mbps, while CAT 5 components must be able to sustain a data rate of 100Mbps. In order for a fabricated cable to meet the CAT 3 standard, it must be performance tested (after assembly) at a data rate of 16Mbps. Similarly, in order for a fabricated cable to meet the CAT 5 standard, it must be performance tested (after assembly) at a data rate of 100Mbps.
Category 5 cable can always be used in a 10Base-T LAN, as it far exceeds the data rate capability involved there. On the other hand, the fact that CAT 5 rated cable, connectors, and patch panels are used does not automatically mean that an installation will be 100Base-T capable, and I have seen numerous 100Base-T LAN’s operating over cables assembled from CAT 3 rated bulk cable.
The primary difference between the two isn’t the cable itself, and the CAT 5 designation as applied to the bulk cable can be misleading to the uninformed. CAT 5 approved cable assembled to CAT 5 approved connectors may not even meet CAT 3 near-end crosstalk (NEXT) minimums if poorly assembled, whereas CAT 3 approved cable, when properly assembled to high-quality connectors can and often will perform at CAT 5 levels. Many homemade “CAT 5” cables, when tested for performance certification, will fail the testing. These cables will therefore not perform at CAT 5 levels at all. The fact that a cable assembly is constructed using CAT 5 materials doesn’t make it a CAT 5 (100Base-T) cable… the only thing that makes it a 100Base-T cable is passing the required performance tests at that level.
The tester described in this article is a basic continuity tester rather than a performance level test device. It is intended for use as a troubleshooting tool for basic connectivity testing rather than for performance issue resolution. However, it will work as a basic continuity tester for any standard UTP/RJ45 cable as used in an Ethernet installation. This applies to cable assemblies made with both CAT 3 and CAT 5 bulk cable.
A Word on Cable Fabrication
It is not the intent of this article to go into any detail on the steps involved in the fabrication of LAN cables. There are several references available to the user on that topic, including a PC911 How-To that explains the process of assembling 10Base-T cables. Rather, I will deal with the steps in making a tester for these cables that you will use in conjunction with your multimeter to quickly verify the proper assembly of 10Base-T cables.
On With the Show
If you have gotten this far – if I haven’t lost you yet – you are probably ready to make your cable tester. This tester is a very simple two-piece unit that allows one person to check the operation of 10Base-T cables with a minimum of fuss. One unit is a loopback unit, while the other is a breakout unit to which the multimeter is connected. It’s really pretty simple. One end of the assembled cable is plugged into each of the tester units. The total cost of the tester unit (less the multimeter as discussed earlier) is under $15.00! If you have some scrap bell wire around, your cost will be right around $10.00.
Here’s what you will need: