Handmade vs. Commercially Produced Ethernet Cables 837
An anonymous reader writes "We have a T1 line coming into our satellite office and we rely fairly heavily on it to transfer large amounts of data over a VPN to the head office across the country. Recently, we decided to upgrade to a 20 Mbit line. Being the lone IT guy here, it fell on me to run cable from the ISP's box to our server room so I went out and bought a spool of Cat6. I mentioned the purchase and the plan to run the cable myself to my boss in head office and in an emailed response he stated that it's next to impossible to create quality cable (ie: cable that will pass a Time Domain Reflectometer test) by hand without expensive dies, special Ethernet jacks and special cable. He even went so far as to say that handmade cable couldn't compare to even the cheapest Belkin cables. I've never once ran into a problem with handmade patch cables. Do you create your own cable or do you bite the bullet and buy it from some place?"
Always buy them (Score:5, Interesting)
Whatever saves time (Score:5, Interesting)
I've spent many hours debugging things that ended up being poor quality TP connectors, but I've also saved countless more hours producing them myself compared to running to the store everytime.
For any permanent installation, go for the molded cables. For anything thats temporary, just pick whatever cable is closest.
And you're not guaranteed to be free of problems just because you buy expensive stuff, I've had problems with Dell PowerEdge switches and factory-made, properly molded STP cables, the RJ45 plug was simply too small and the copper pins didnt connect every time. Really odd, we had to throw away a whole box of STP patch cables for that reason.
Hand-made is time consuming (Score:3, Interesting)
but makes perfectly fine cables from what I saw. I generally don't do it anymore unless I have a very custom length as pre-made are really inexpensive and over 10 cables I usually have to re-crimp at least one end. Does your boss have any proof that hand-made cables are inferior?
Don't cross the data streams! (Score:5, Interesting)
Your boss is a dumbass.. (Score:5, Interesting)
Ask him how the premise wiring in every commercial building in the world is installed. They order patch cables from some commercial patch cable vendor for every run, riiiiiiiight.
Also, CAT5e is fine for what you are doing. I agree with the previous poster that you could practically use tin cans and a string for this.
These special dies, jacks, and connectors are called "CAT5" parts and you can buy them at Home Depot I think. Does that make them "special" ?
Comment removed (Score:5, Interesting)
Yes and No (Score:2, Interesting)
We have a roll of bulk cable for when location X needs a network run Right Now. I route it, cut to length, and terminate it. I'm pretty good.
I don't have a TDR, so I run 200M of data at the target link speed. If it isn't good enough (i.e. more than 10% away from my target throughput rate), I reterminate the cable. If it still isn't good enough, I pull new cable.
This is for those projects where waiting a week for a shipment of manufactured cable won't do. For anything else, you are wasting time and money by making your own cable. Tested chinese patch cables are cheaper than buying bulk cable, and they have a higher chance of working right the first time, and they're probably the right kind of cable for what you're doing.
Your boss is being paranoid - I'm sure you can install cable to handle the 20M link without problems... but he's right to say that you should look to save money elsewhere. I'm guessing you make more than $3/hr - your time can be put to better use than making a $20 cable.
Now, on the other hand - if you're doing a run that's more than 100ft long, yes. Make it yourself (or hire a professional installer). Long cables are stupid expensive - but that's horizontal cabling, not patch cabling. Still have to pull, route, and terminate it properly. Getting good connectors on it is the tricky part - none of the local places carry the kind of jacks we use (Panduit MiniCom - all the locals carry some crappy cheap variety of a keystone jack).
TLDR: You had a T1, probably at the same demarcation point. Why aren't you reusing that cabling to move the data from the new channel bank that's sitting 3 feet away from the old T1 interface over to the network closet?
Re:Not sure if I'm stating the obvious here but (Score:3, Interesting)
If you get the rated speed and it's reliable, need we delve further?
But how can you be sure that a cable is as reliable as you think it is without a $1,200 device [slashdot.org] that comprehensively tests the reliability?
Re:Create your own but TEST the cables... (Score:4, Interesting)
I guess you may need a special NIC, but even still, its gotta be cheaper than $1200.
At NASA I regularaly good cables, (Score:3, Interesting)
with equipment that's not much different than stock equipment. I test these cables with a DTX-1800, they do great.
They're sticklers for BlackBox brand cable, I don't know if it's because the cables good, or the more likely scenario that instead of specifying TIA-568B compliant cable they have have to give a part number to make a "Typical". A "Typical" is a blue print for a cable. Remember, it's government, loads of red tape.
We also use Black Box brand connectors, again, for part number reasons I'm almost certain. For the Cat-5 stuff there is something a bit different than your run of the mill cables, it's the inclusion of black load bars that get crimped into the connection. A bit different than most connectors I've used.
The only Cat-6 I've made was a specialized connector with additional grounding added, so I wont get into that.
Beyond what's mentioned the only difference between NASA and the rest of the world is the use of really expensive test equipment, and the insistence that calibrated ratcheting crimpers are used. For test reasons I've made cables using my own stuff and put it on the Fluke, I hate to say it, but my uncalibrated out of the box $20 crimpers from Ideal do just as well as there $150 at minimum crimpers that are custom pieced together. At least according to the Fluke.
Cheap Cables: Buyer Beware (Score:2, Interesting)
Running to the store is great, but I've learned the hard way to trust only one manufacturer. Their cables are guaranteed for life!
Buy a few hundred 'Brand X' cables and a percentage of them could be useless.
Once that happens, you'll have a box of cables you'll never want to use. Just can't trust them.
Beside, who has a spool of beige, black, blue, gray, green, orange, pink, purple, red (crossover only), white, and yellow laying around?
Re:Create your own but TEST the cables... (Score:5, Interesting)
If you're testing to certify cat5, cat5e, or cat6 you need a cable tester. If you cannot certify the cable to a category you cannot guarantee the cable will work. So the cable is always suspect when you have connectivity issues.
Keep the OSI model in mind, errors at the physical layer cause the whole stack to collapse.
The advantage of cabling over wireless is that you can guarantee that the cable will work where there's no such promise with unlicensed RF spectrum.
CAT 6 Throughout the house (Score:2, Interesting)
That being said, if your boss says do it his way and he'll give you the budget to do it that way, salute and say "Yes sir." Doing it yourself means that, at best, you proved your boss wrong. Not a good career move. If anything goes wrong (even if it's not your fault like a squirrel chews through a cable) it will be your fault for not doing what your boss said.
Cheers,
Dave
Re:Rewiring our building (Score:1, Interesting)
"Your boss is unclear on the tools needed and the difficulty...just simple hand crimpers were all they needed."
What you didn't see is that they likely took care to avoid the many things which can cause interference (light ballasts, electrical cables, fans, etc), and correctly observing cable bend radius'. In the end, several of the connections may have required being re-terminated because tests failed, and even the possibility of running a whole new line.
When I was working as a cabling tech and Cat 6 started to become the norm great lengths had to be taken to ensure everything tested at it's optimal specifications. We would even test every box of cable *before* putting it in use, as well as any patch cables the client required. Not everything passed, but it was a lot easier to send it back to the manufacturer before it was in a wall, ceiling, or floor.
Re:How much is your time worth (Score:5, Interesting)
1. Amateur cable. These are done just any old way as long as the colors match at both ends. The pairs don't even have to be twisted for it to work over very short distances (2 to 6 feet) at 1GB.
2. Professional Cable. All the pinouts done properly according to whichever standard you are working with, by someone who knows what he is doing.
3. Factory cables. Here is the dirty secret. Some of these are done by robots and some are just professional cables. There is no way for you to tell which is which.
Now to your specific problem. If your boss insists on paying $300 for $20 worth of cable just to satisfy his own misguided notions of quality, you as the highered help just have to accept his decision and go cry into your beer.
Or better yet. Smile. they had no intention of using the money you would have saved to enhance your salary.
TDR is the modern ball-joint tester (Score:1, Interesting)
Long ago most every auto repair shop had a "ball-joint tester". This was a $40 device that would wiggle your car's ball-joint and show PASS or FAIL on a big analog dial. These were extremely popular with the repair folks as it let them keep up with their yacht payments.
The modern TDR is a very similar tho higher-tech gadget. It can measure and display fraction of an ohm and split-picosecond discontinuities. In other words, a yacht payment maker.
In case you don't get it-- the ball-joint tester was a bogus test in two ways- (1) It applied forces not ever seen in normal use, and (2) It displayed teensy bits of jiggle that did not matter at all in normal use.
Same thing with a TDR-- it applies waveforms never seen in normal use, and it can display aberrations that are many many times smaller than can ever cause any data loss.
In other words, a yacht payment guaranteeor.
Re:How much is your time worth (Score:3, Interesting)
I second that.
Drive to Best Buy: 15 Min
Time to purchase: 10 Min
Drive back home: 15 Min
Cost of cable: $45 (ball park)
vs
Pull cable to desired length: 30 seconds
Crimp end 1: 1 min
Crimp end 2: 1 min
Cost of raw material: $5 (ball park)
So Unless you make about 40$ in 2 min i.e. 1200$ an hour, its better to crimp your own.
Stranded (Score:2, Interesting)
You really want to use stranded wire for patch cables. Solid will end up cracking with the repeated bending that most patches are subjected to. I've made patches by hand with stranded and found it much harder to work with than the solid most people are used to.
It's definitely not worth my time unless it's an emergency with no alternative (i.e. poor planning).
-db
Re:How much is your time worth (Score:1, Interesting)
Set up a dummy company, and get them to sell you the $300 cables for $250. Then crimp the cables and sell them to yourself. You pocket $250 less materials, the boss gets his "professionally made" cables, and everybody is happy.
Unfortunately that's called embezzlement and it's sort-of highly illegal. Now for a privately owned company, you can create your own side business of making cables and bid the job for the cable creation and charge $250.00 when your competitor charges $300.00. You win the bid for being the lowest price and you then create the cable for $20 and make $230.00 in profit. That is NOT illegal. However for a publicly traded company, that too is unfortunately illegal.
Unfortunatly no... (Score:3, Interesting)
Something like the Fluke tester is a very sophisticated ANALOG device. Its measuring reflectivity and a whole host of analog properties, in order to determine that the cable meets the specification.
EG, it will tell you where there is an actual break in the cable.
Personally, I don't consider build-my-own cables saving money. Rather, it is some other reason (the necessity to be neat, an inability to pull pre-made jacks through the wall...) that is the reason to build your own.
Re:How much is your time worth (Score:3, Interesting)
Strictly speaking, 100m is a fuzzy limit. More of a rule of thumb than anything, and largely due to signal degradation and reduction causing you to leech bandwidth. It's true that high bandwidth connections are better over short distances, and it's true that you're better off running switches every so often to repeat/rebroadcast the signal, but when modern computers are coming with gigabit ethernet cards soldered onto the motherboard (and it's nearly impossible to find a mainstream mobo that doesn't have one), you can quite easily get away with installing long runs of > 100m without the users noticing a significant loss of quality. It'll still be there, but they're getting so much bandwidth that they probably won't care if they lose 15% of it.
Back when the fastest you could get was 10mbit, and even before then when you were looking at 1mbit connections, it was a little more important to keep the runs short.
All that said, you're still better off with a switch to repeat the signal every so often. But you might not have the budget for it.
Re:How much is your time worth (Score:2, Interesting)
Re:How much is your time worth (Score:5, Interesting)
My opinion is a little different: Don't build them one at a time. And don't buy them one at a time, either.
Just pick up a bunch of different lengths of pre-terminated cable from the good folks at deep-surplus.com. Buy a bunch of 1-foot cables, along with some 3-foot cables. 5-foot cables. 7-foot cables. 12-foot cables. So on, so forth. Then, when you need a cable of a given length, you've (gasp!) already got one!
They're easy to use, too! Just reach up on the shelf, and get one! Way faster than finding the strippers, the cutters, the crimpers, the box of ends, and the box of wire... And then you've still got to cut, strip, sort, cut, insert, and crimp the shit together, before doing the same thing on the other end.
Feh.
They cables from deep-surplus cheap, they're Chinese, they're durable, consistent[1], and I have never had a bad cable after years of doing this whenever possible. Plus, every order comes with a bag of Skittles.
The trick to making this economical and time-efficient is to put it all on one PO.
[1]: Speaking of consistency: I do have the occasional cable that I make myself go wonky, in applications where prefab cabling doesn't apply, like UV-rated Cat5 up a radio tower. This, despite using a good crimper with a good die, and high-quality ends which are made specifically for the wire in question, and a lot of practice to develop decent workmanship. The Chinese cables are consistently more consistent, and always work.
My simple rule (Score:3, Interesting)
Patch cables are bought pre-made in standard lengths. These are for use in racks, from desktops to wall jack, patch panel to switch, etc. In these cases the time taken to properly crimp the ends and re-doing the occasional bad crimp is just not worth it.
Cable RUNS, such as what you're talking about are done custom in-house.
One big advantage is that the cable pull can be done much more easily without the ends in the way. You can abuse the end of the cable as much as necessary (including tieing it in a knot) to get it pulled then just snip the abused part off.
Another advantage is getting the length right. It's a shame (and a mess) to have to coil up 30 meters of cable just because you could buy 50m or 100m and you needed 70. The alternative would be to measure the length you need and have someone else custom the cable 'professionally' for you. The problem is that you're now doing the pull twice (once with string to measure and then again with the custom cable) and the 'professional' cable will come from some guy doing exactly what you were going to do. It's not like it's rocket surgery and he may have no more experience than you do. Further, he doesn't have the sure knowledge that he'll be at the office till 3 A.M. if it goes bad to encourage him to do it well.
Further, if you buy pre-made of have it "professionally" made, you have to be extra careful not to abuse the end of the cable when you pull it. Most likely, you'll tie the pull string near the end as tightly as you can, and then when it comes out the other end, you'll find that it slipped and all of the pulling force has been supported by the connector pulling through the crimp. If the end didn't actually come off, it's probably worse than even an amateurish crimp by now. Either that, or the connector has caught on absolutely every obstruction along the path and so is similarly damaged. Covering it over with tape will help but not eliminate the problem.
I've never had a problem with a hand made cable that didn't show up right away with a tester and couldn't be solved by snipping the end off and trying again.
Big tip for getting a clean crimp: Strip off a little too much outer insulation, get all your conductors lined up in the connector, then pull it back out holding the conductors in alignment between your thumb and index finger. Now snip the excess off squarely and re-insert into the plug. If the outer sheath doesn't go neatly into the connector, nip off a bit more. You now have a nice professionally made cable (you are, after all, a networking professional and you made the cable).
The only actual difference is who does the testing.
Re:meh, easy... (Score:3, Interesting)
The best part of that cable is the fact that it is actually pretty poorly made. Look at the plugs, there are no guards, you can easily snap off the click thingy if you are careless.