Helping people with computers... one answer at a time.

It's not uncommon to feel like the internet is slowing down. Determining if it's your ISP's fault is possible, but it needs to be done carefully.

I live in rural Arkansas and I have been paying for DSL at a premium. Most of the time, however, my system runs like dial-up rather than DSL. How can I check to see if I am actually getting what I am paying for?

There are many things that factor into the speed that you actually get from your internet connection. If you read the fine print on most ISP agreements, they're quoting you a maximum speed, not a minimum, and there is no guarantee of throughput.

Nonetheless, if you're paying for 1.5 megabits and you're only getting 33 kilobits on average, then it's time to talk to the provider.

I'll look at some of the reasons that speed might vary and, of course, show you a quick way to test.

Testing Your Speed

The simplest way that I've found to get a quick read on how my internet connection is fairing is to simply visit a site like Speakeasy's speed test. That'll test both your download and upload speed and present you with the results.

Here are the results that I just got:

Leo's speed test results

On the surface, it looks like I'm not getting what I'm paying for - my T-1 should be giving me 1.544 Mbps both down and up. This test looks like about half that.

But I'm not calling my provider just yet.

Your Network Can Affect Your Results

After getting those results, I did a quick little inventory and discovered that I have 12 devices connected to my network, all sharing that single internet connection (five desktops, four laptops, a WebTV, my DirecTV recover and my Android phone).

"If you need an absolute test, then you must disconnect everything from the network except for one PC ..."

It's not surprising that I'm not seeing full speed, as it only takes one of those devices performing a download to cut my effective speed in half.

Granted, my case is a little extreme, but it raises one of the most important points about speed testing: other devices sharing that connection will impact your results. It's also easy to forget about some of them, such as your phone or DVR.

If you need an absolute test, you must disconnect everything from the network except for one PC to perform the test.

Your Software Can Affect Your Results

When I first got my T-1 line, I definitely tested to make sure that I was getting what I paid for, right down to disconnecting everything except for one computer. While my download speed was great, the upload speed wasn't.

I ended up calling the provider. A technician showed up and ran the test herself only to find that the speed was in fact correct both up and down.

The difference?

I was running Firefox and she was running Internet Explorer. When I re-ran the test using IE, I got the same, correct results.

The lesson here isn't to use IE for your tests (recent releases of Firefox run that test more consistently), but rather to be aware of what else is running on the machine when you run the test, as well as trying different browsers, perhaps with add-ons disabled, if you run into unexpected results.

Your Home Might Affect Your Results

Particularly, if you're running the test over a wireless connection, there are many things that could be interfering with the test. When testing an internet connection for speed, or anything else for that matter, it's best to use a wired connection to eliminate these issues.

Another potential problem is the wiring in or to your house. Years ago, I had an intermittent internet connection that was eventually traced down to water corroding the phone line between the street and my house. At the time, that phone line was carrying my DSL connection. Similar issues can easily manifest as underperforming internet connections.

Your Neighbors Might Affect Your Results

This is particularly true of cable internet connections where bandwidth is often shared between several customers.

The same concept applies to all internet connections. At some point, your internet connection is combined with those of other customers. If they suddenly begin making heavy use of the internet, the speed of your connection could be impacted.

This is one of the reasons that many ISPs disallow peer-to-peer file sharing; it's notorious for using lots of bandwidth.

Your ISP Might Affect Your Results

Of course, this is the main reason for running the test in the first place - to confirm that the ISP is delivering on their implied promise.

ISPs face a tough problem because it's impossible to provide enough infrastructure at anything close to a reasonable cost where everyone could use the full capacity of their connection at the same time.

Consider the telephone system: if everyone picked up their landline handset at the same time, only a scant few would actually get dial tone. The phone system simply doesn't have the capacity to provide connections to all of their customers simultaneously. They rely on the fact that rarely, if ever, are a significant percentage of their customers on the phone at the same time.

ISPs do the same thing.

Similarly the quality of equipment they use, the size of their own upstream connection to other ISPs or the internet, and even the architecture of their own network infrastructure can all affect their ability to deliver consistent speed.

And I'm assuming that the quality of what they provide might well have an impact on your ability - or desire - to continuing paying for their service.

Article C4775 - March 24, 2011 « »

Share this article with your friends:

Share this article on Facebook Tweet this article Email a link to this article
Leo Leo A. Notenboom has been playing with computers since he was required to take a programming class in 1976. An 18 year career as a programmer at Microsoft soon followed. After "retiring" in 2001, Leo started Ask Leo! in 2003 as a place for answers to common computer and technical questions. More about Leo.

Not what you needed?

17 Comments
Mike
March 25, 2011 3:57 AM

It looks like you've got a typo in the link that's supposed to take us to Speakeasy's speed test. There are two t's at the end but when I take one away it still doesn't work.

Thanks for letting me know. Should be fixed.
Leo
25-Mar-2011

Rickvv
March 25, 2011 4:01 AM

Try this one, Mike-
http://www.speakeasy.net/speedtest/

Jeff Cohan
March 25, 2011 6:41 AM

Thanks, Leo. Very informative, as always.

I jumped to this article from Twitter because of a longer-than usual FTP session yesterday. I had downloaded 40 MB of JPEGS from one server and then uploaded them to another. I'd almost forgotten what a big difference there is in upload/download speeds.

My results at Speakeasy.net were very close in FireFox and IE. Thanks for that link.

(By the way, 1998 called, and they want their WebTV back.)

:-)

I retweeted your tweet so that my 84 followers can learn from it.

Appreciatively,
Jeff Cohan

vincent
March 25, 2011 6:50 AM

Is there a European version of this test?

Mark J
March 25, 2011 7:36 AM

For Europe and the rest of th e world

http://www.speedtest.net/

I considered this link for the main article, but when I looked at it that page had several "Scan Now" links that are really ads - I found that missleading for this type of service. It's fine if you know what NOT to click on. (And it's the same service/technology as the one at speakeasy).
Leo
25-Mar-2011

Ken B
March 25, 2011 10:09 AM

Yes, there is a double-T at the end of his link. The reason that taking the extra "t" away from the page that's shown is that the redirector also takes away the "go." in front of "ask-leo.com". In any case, the other comments have a direct link to it as wll as to speedtest.net.

And I must be spoiled. My test shows 20.73 down and 18.01 up. :-)

http://speedtest.net/result/1220052387.png

Mark J
March 25, 2011 2:47 PM

Well they clocked me at 41.3 down and 7.6 up (a bit slow but WTF). That's in the heart of Germany's silicon valley. And it costs less than $40 a month and it includes 3 telephone numbers and 4 cell phone sim cards. For $14 more I can call 25 countries unlimited for free. Let's see AT&T or Verizon top that one.

frank golden
March 29, 2011 10:58 AM

Hm, speakeasy doesn't seem to work at this time.
I actually prefer Speedtest.net.

Alex Dow
March 30, 2011 12:26 PM

Leo's comment regarding "water corroding the phone line" should also be looked at in the opposite context.

With the enquirer living in "rural Arkansas", could it be that a local Earthing Connection for his phone line is drying out?

One of the factories I worked in over here in the Kingdom of Fife was having phone problems, which the phone company could not trace the origins.

Eventually I took a look at it and found that the main Earthing Point had become partially covered by a slabbed footpath, very much reducing the amount of rain water percolating in to the earth/dirt in that vicinity.

A bucket of water cured the problem.

Also, where possible, re-make or "shoogle" the various contacts and plugs.

Phone systems use DC operationally, and over a period of time, the gold-plated contacts can build up "Dendrites", pillars of poor or non-conducting crystalline gold, leading to poor or dead contacts.

Traditionally, these were cured on radios and TV sets (mainly DC driven internally) by giving the cabinets a solid whack or two, shaking the contents about and breaking down the dendrites.

WD40 or a similar contact cleaning oil may help at plug/socket interfaces.

The phone side might be checked by using the "Quiet Line" facility. Over here in the UK, we can ring a Test Number, listening to the phone to get some idea of the amount of noise on the line.

It may be that joints in Distribution Cabinets on the route to the Exchange need re-making if the installation is old or badly done.

Is the line shared with another Customer - in which case the problem may be arising there. At one time such shared or multi-drop phone lines were common in the UK, both generally and particularly in rural areas.

Also ensure that the Exchange equipment is properly checked out.

Alex Dow
March 30, 2011 12:30 PM

I should have added - Being in rural Arkansas, the Earthing Point may also be effected by deep winter frosts.

Frozen water in the soil is crystalline; and most crystals are non-conducting, by definition not having free electrons to provide a conducting path.

Alex Dow
March 31, 2011 4:49 AM

You also need to take into consideration how far you are from the Exchange or DSL Equipment.

I am less than 1 (one) mile (about 1.25 Kilometers) from my exchange, as measured "on the cable" which is underground throughout.

Over the years, I have carried out more than 200 tests, the results showing an average of just under half the "Up to" speed, whilst the spread is from extremely slow, about 6% up to about 75% of that unattainable speed.

I suspect that distances there fore attenuation in rural Arkansas are likely to lead to significantly poorer performance.

dino
April 1, 2011 6:06 PM

I have many complaints about fl brighthouse cable but their 7mb road runner runs at 9-10 day and night for $32 a mo. I am miles from any central location. I like speedtest.net, great graphics. Same results as speakeasy.
I had their 15mb and it ran at about 27mb all the time.
Waiting for optics.

Ari
April 3, 2011 5:45 PM

It is a ver nice article. Most of the answers are already given except what I am going to share. I am in Japan I use ADSL. Once during downloading file I noticed slow speed. I contacted my provider and explain them. They asked me to run a test which took one day next day I was told it seems ADSL Modem is faulty.(I used it one year)I got it replaced and speed increased.
After replacing ADSL Modem, after few months again I noticed slow speed, I contacted provider after running test I was told, "Sorry some how your internet speed was dropped" they adjusted that and internet worked fine.
Note: before I contacted provider I did run online speed test.

Frank
April 6, 2011 8:44 PM

The truth about dsl speed is inconsistency. The major reason is AT&T are still using 1980 equipment and most ISP's use AT&T. There is too much traffic for the bandwidth. I am in Northern California, I have Highest Speed DSL, but around 9:00pm it really slows down.

Carlos R Coquet
April 8, 2011 6:16 PM

Another thing that can cause DSL to run at dial-up speed is to plug the DSL line in the "Phone" jack in the modem instead of plugging it in the "Line" jack.

Bob
June 28, 2012 8:36 AM

My download speeds were terrible, pages took forever to download. Broadband was acting like dialup. Finally checked my ISP website and found they no longer supported my cable modem. Purchased a new reccomended model and everything is like new again. $90 for a new cable modem was worth every penny. Sure solved my problem.

Kevin Manner
January 29, 2013 5:35 PM

This is just the article I was looking for. I have been suffering from slow Internet for month's.
Called COX many times. The adress for showing your modem satistics is
http://192.168.100.1
I had a reading of ----Transmit Power Level
28.2.2 dBmV
Now it's 30.2 and well??? what should I do. I was told by them it should be much higher. But aleast web pages kinda open now. Maybe they want me to subscribe to faster digital speed, but that won't work unless they get the higher then 30.2 dBmV Transmit Power Level I believe.

Comments on this entry are closed.

If you have a question, start by using the search box up at the top of the page - there's a very good chance that your question has already been answered on Ask Leo!.

If you don't find your answer, head out to http://askleo.com/ask to ask your question.