All,

I have a question re: transmission speeds. We have a house that we went to where the computer-savvy customer claimed (everything was unhooked so we weren't able to see or test this):

- With his computer plugged directly into the modem he was getting speedtests of 600 Mbps down.

- With his computer plugged directly into a Netgear gigabit switch located in the panel (CAT cable goes from modem to switch) his speed went down to 300 Mbps down.

- With his computer plugged directly into a CAT5 wall jack in the Family Room (didn't test or check to see if it was CAT5 or CAT5e behind jack) he was getting 100 Mbps down. Path goes modem>switch>jack.

He's a big speed guy so he has CAT6 all over the floor in the interim to get the highest speeds possible. Rewiring will be messy due to the layout and multiple stories.

Thoughts? I can't find concrete numbers on transmission speed standards between CAT5 and CAT5e. I didn't check if the house was wired with CAT5 vs. CAT5e. The Netgear piece's manual states that it passes 100 Mbps down with CAT5 and 1000 Mbps down with CAT5e.

House isn't that big so the runs aren't ridiculously long (maybe 75'-100' runs?).

Any idea why the disparity in speeds? Has anyone seen CAT5 pre-wires that might cause this drop?

Thanks in advance

This kind of discussion always makes me think of Bob Pease's comment that he uses his computer for data entry, record keeping, and writing articles and technical papers, so he required a computer that would be able to handle his work speed, which was on the order of 0.00000001 GHz. People who create large and highly detailed artworks, on the other hand, really need that speed.

Your guy is a speed guy. What does he use the computer for?

Is he looking at actual usage requirements, or only how fast can it be?

Does he want a really fast LAN, faster than his internet?

What is he using to measure the speed?

Is he getting internet speed in any way comparable to his network speed?

Did he connect to a hub instead of a switch? (probably probably not, but just covering bases)

Forgetting that the prices are Australian, does this help at all?




You say

This is a bit suspect. To make it blatantly obvious, if one were to connect "the Netgear piece" with a foot long piece of CAT5 and a foot long piece of CAT5e, would these differences appear?

And of course..I was a bout to quote two hundred words from [Link: lifewire.com], but let me leave it at this:

There are two important things here. One is the bare facts, the other is the reminder that 1 GBps = 125 MBps.

What model is the Netgear? Is it a managed switch?

That’s where I’d be looking. I can easily get 1gb over cat5e under 100’. Longer distances will degrade to 100mb.

Also, is he using the modem as the home router?

Here’s an idea, and perhaps a sales pitch too.

Plug the switch side of the jack directly into the modem. Basically modem to jack. Does the speed increase?

If it does, but not up to the modem speed. Then have a CAT6 or even CAT7 handy that is roughly 100’ long. Run it out, don’t keep it coiled up at least, and plug it into the modem. What is the speed? Then plug it into the switch, what is the speed?

If the 100’ CAT 6 or 7 maintains the speed. Then use that knowledge as a sales pitch... but advice them on the mess that is involved in rerunning. So hopefully he’ll just deal with it. Or choose the CAT 7 for future proofing.

KOT

A cable certifier would be handy. Regardless of the type of wire you find on the other side of a jack, you don't really know what happened inside the walls. Unless you have a hint, as I have had, where one end of the cable is a different color.

If you end up pulling wire, pull a fiber too.

Electricians around here butcher everything, they are idiots. I consistently see them have cat5 in the wall and leave 3-4 inches of wire untwisted before terminating to the cheapest 'self-punch' keystones you can imagine. Same at other end.
When I test, I still get 980+Mbps from these trash connections.
You have to really f^ up a wire run and termination bad to cause a speed drop like you are talking - at least at those short distances.
When you get into 10Gbit networking, now you are talking needing perfect connections.

I would be looking at the equipment.
P.S. = I hate Netgear.

By Modem, do you mean Router?

Every time some piece of equipment is added, the performance can change. Who terminated the cables? Are they kinked, were they pulled with too much force, are they wrapped too tightly, were they pinched?

What model is the NetGear piece? Read the specs.

Why is Cat5 even in the conversation? Can't find specs for cabling? How are you searching? Be very literal in your search- write your questions as you would ask, verbally.

[Link: planetechusa.com]

Thanks all for the help. This was a walk-through where this portion came up from the client as an afterthought. We hadn't contracted to do this troubleshooting so we weren't going to do anything re: testing, troubleshooting, discovery until we received $$$. I wanted to reach out to all to get your thoughts.

Depending what you read, there are lots of given max. speed #s for CAT5 and CAT5e. I was curious what everyone's take on this was.

No...modem. He was plugging directly into the modem to see if the router was the limiting factor.

The worst thing I see is guys not knowing about the bend ratio on any of this data wire.

Take a piece of Cat 7 and have it make a really tight turn coming out of a top, or bottom, plate and you can reduce it to a nice low speed wire run since you have completely changed the relationship between the pairs. May as well tie a few knots in the wire.....

One might respond, well, technically, I suppose we don't want to consider using CAT5 at all, but we'd like to compare the technology we're going to use with the previous technology, to see how much better present technology performs.

For years I'd try to follow search parameter advice, meaning using as few terms as possible that relate EXACTLY to what I'm looking for. Meanwhile, google got better and better and now you can write a question just as you'd ask it. Of course, that doesn't mean our question-asking skills have advanced while google has gotten better....

He didn't say he wanted to see if the router was the limiting factor. He said Managed Switch.

I did a house that has wire mesh in the walls to hold the plaster in place and, of course, I was told that everyone wants to do things via WiFi. I may have said something like "Good for you" or "Well, life is full of little disappointments", but that was a few years ago. I explained why that would be difficult, especially since everyone uses Apple computers and hand-held devices, then looked for ways to hard wire access points. I found a phone jack in the Master Bedroom that was surface-mounted and that annoyed me to the point of wanting to flush-mount it just because that was a giant sign of the ATT installer's laziness and since it's a fine old home, it has no place. Sure enough, I removed the surface mount box and found a hole in the base molding that was more than large enough for a keystone insert to fit into. When I looked at the cable, I found that it had four pairs, but none were twisted the same as the Cat5e I have seen, but I needed to find a cable, so I punched it down, toned it and tested the speed with my computer via ethernet cable- it did over 40Mbps, which was the high end of what their plan specified, so I used it. I have heard of phone cable (Cat3) that worked for ethernet, so it seemed like a reasonable thing to attempt.

Cat5e was originally for 10/100 networks, but when I took a networking class in 2006, tests had shown it would do gigabit up to about 100 feet, as long as the cable wasn't abused.

I never thought I should need to learn to ask questions in a way that the search engine needs, especially after working with someone who was extremely good at programming and explaining that search engines are just 'sort programs' ('sort' used as a verb). Include what you need to find, or you won't find it.

Cat5e can do gigabit up to 328’. That is the spec.
If your not getting that at the switch or after it, i would suspect there is a bad cable.
You should use a tester for this.
I saw something like this last year. After looking at all of the obvious things first i tested the line and it came back with broken pairs in the cable and no easy way to fix it.