10mb bb more like 6mb! [Archive]

View Full Version : 10mb bb more like 6mb!

GENERALZODD

12-05-2006, 00:50

hi all can anyone give me some advice on the following problem? here goes....re-tested my 10mb bb in off peak time and i am only getting around 6mb, i know i have an ntl100 modem and know that this can be the cause, what do i do now, do i ask for my modem to be upgraded ie ntl 250 thus enabling them to fulfull their promise and give the 10mb bb. what are my options?

thanks

d

Paul K

12-05-2006, 07:22

Think there was a time when NTL would only upgrade the modem for 10Mb if you paid for it or if yours was faulty (think it's still the same now). It's UP TO 10Mb by the way so if you complain to them they will point this out to you that they haven't really broken a promise to provide 10Mb. Do you connect through a router at all as some routers cannot handle the speed either.

Adcuz

12-05-2006, 07:32

anyway i thought it was a sneaky "up to 10 meg" maybe its becuase they cant be bothered to fix things when they dont work, or maybe there network cant handle it in some places.

Paradox26

12-05-2006, 08:44

Zodd, if you aint getting the full MB Id personally downgrade to 2 and then get upgraded to 4 for free. But your descision, that way you could save a few extra ÃƒÆ’Ã¢â‚¬Å¡Ãƒâ€šÃ‚Â£ÃƒÆ ’Ã¢â‚¬Å¡Ãƒâ€šÃ‚Â£ too. say ÃƒÆ’Ã¢â‚¬Å¡Ãƒâ€šÃ‚Â£5/month which is ÃƒÆ’Ã¢â‚¬Å¡Ãƒâ€šÃ‚Â£60 a year!

GENERALZODD

12-05-2006, 16:51

was upgraded from 24.99 (2mb) to 17.50 for 10mb. but not getting anywhere near this. (and yes i am connected through a belkin router, could this make a difference?
d :)

Druchii

12-05-2006, 17:01

was upgraded from 24.99 (2mb) to 17.50 for 10mb. but not getting anywhere near this. (and yes i am connected through a belkin router, could this make a difference? d :) Take the router out of the equiation and do some tests? You'll see for yourself.

chrisjackson1985

12-05-2006, 17:53

i think its the ntl modem 100, request a 250.

they should swap it, as i dont think the ntl 100 modem can handle speeds of upto 10mb.

the speed your getting through it, i think is its max.

get a 250,
set it up through ethernet, your router should make no difference to the speed.

and then ...

go to,

http://www.speedtest.bbmax.co.uk/

and test your speed :)

good luck

alarmeng

12-05-2006, 18:06

Ur lucky fella

I was only getting around 2.3to 3.2 so i went back to 2mb and it is 2mb.

So i paid ÃƒÆ’Ã¢â‚¬Å¡Ãƒâ€šÃ‚Â£50.00 for nowt!

__spc__

12-05-2006, 20:55

My upgrade via retentions to 10MBps sees me get 5.5Mbps regularly under XP and 9Mbps under Linux. Consistently.

Postcode lottery until ntl can stabilise the 10Mbps...

GENERALZODD

12-05-2006, 23:00

Router is currently connected via ethernet port, makes no difference when unpluggged. Think its could be the modem, what do i say to them when i ring to get my modem replaced with 250?
Speed now hovers around 1mb :mad: very disapointed!
d

James Henry

12-05-2006, 23:05

My upgrade via retentions to 10MBps sees me get 5.5Mbps regularly under XP and 9Mbps under Linux. Consistently.

Postcode lottery until ntl can stabilise the 10Mbps...

Retentions upgrading people at a whim to 10Mbit probably won't help with the stabilisation :(

OP - the ntl100 is NOT capable of 10Mbit. You will need a new modem, which in theory they'd charge you ÃƒÆ’Ã¢â‚¬Å¡Ãƒâ€šÃ‚Â£50 for swapping out, however with retentions throwing a lot more than this about I'm sure you'll have no probs getting it done for free.

Phil Capron

12-05-2006, 23:29

Can people please sort out their bits and bytes and say how they're testing their speed.There's a lot of errors going on in some of these speed threads and not just from Ntl.

Nilrem

12-05-2006, 23:55

A quick note for those who are finding they are getting better speeds under one OS than another, if you're using different firewall/security applications that could be it.

I found Zone Alarm was limiting me to about 2.5mb on my systems, I had put it down to the modem (tailfin) or router, but noticed a freshly formatted (and much lower spec) system was reaching 6mb+.

__spc__

13-05-2006, 07:15

I posted elsewhere; Linux gives me ca. 9Mbps, XP gives me 5.5Mbps on the same dual-boot machine.

Disabling XP firewall makes no difference.

mmm

13-05-2006, 08:53

I posted elsewhere; Linux gives me ca. 9Mbps, XP gives me 5.5Mbps on the same dual-boot machine.

Disabling XP firewall makes no difference.

This can be due to TCP/IP settings typically used to optimize windows dial-up connections, but restricting broadband.

What does this link say about your connection in Windows and Linux?

http://www.speedguide.net/analyzer.php

I found that having CISCO VPN client software on my PC changed these settings for all adaptors.

__spc__

13-05-2006, 13:11

Cracked it - I used DrTCP to change my MTU and RWIN; I now get:

Sat, 13 May 2006 12:09:44 UTC

1st 512K took 441 ms = 1161 KB/sec, approx 9567 Kbps, 9.34 Mbps
2nd 512K took 431 ms = 1187.9 KB/sec, approx 9788 Kbps, 9.56 Mbps
3rd 512K took 430 ms = 1190.7 KB/sec, approx 9811 Kbps, 9.58 Mbps
4th 512K took 431 ms = 1187.9 KB/sec, approx 9788 Kbps, 9.56 Mbps

Overall Average Speed = approx 9739 Kbps, 9.51 Mbps

Thanks very much!

mmm

13-05-2006, 15:46

I'm glad you are sorted, pity I can't fix my own:-

Sat, 13 May 2006 14:44:49 GMT

1st 512K took 515 ms = 994.2 KB/sec, approx 8192 Kbps, 8 Mbps
2nd 512K took 500 ms = 1024 KB/sec, approx 8438 Kbps, 8.24 Mbps
3rd 512K took 500 ms = 1024 KB/sec, approx 8438 Kbps, 8.24 Mbps
4th 512K took 719 ms = 712.1 KB/sec, approx 5868 Kbps, 5.73 Mbps

Overall Average Speed = approx 7734 Kbps, 7.55 Mbps

But I'm sure that's due to my old Terayon cable modem and its 10Mb Half duplex WAN port!

Druchii

13-05-2006, 21:58

Cracked it - I used DrTCP to change my MTU and RWIN; I now get: Sat, 13 May 2006 12:09:44 UTC 1st 512K took 441 ms = 1161 KB/sec, approx 9567 Kbps, 9.34 Mbps 2nd 512K took 431 ms = 1187.9 KB/sec, approx 9788 Kbps, 9.56 Mbps 3rd 512K took 430 ms = 1190.7 KB/sec, approx 9811 Kbps, 9.58 Mbps 4th 512K took 431 ms = 1187.9 KB/sec, approx 9788 Kbps, 9.56 Mbps Overall Average Speed = approx 9739 Kbps, 9.51 Mbps Thanks very much! Lovely.

And mmm thats amazing speeds from the Tarayon. But shouldn't a Base10 only be capable of less than that?

James Henry

13-05-2006, 22:00

Lovely.

And mmm thats amazing speeds from the Tarayon. But shouldn't a Base10 only be capable of less than that?

Speed of 10Mbit half duplex connection depends a lot on how the machine is set up on the TCP settings, etc.

7.5Mbit isn't unrealistic.

Druchii

13-05-2006, 22:02

Speed of 10Mbit half duplex connection depends a lot on how the machine is set up on the TCP settings, etc. 7.5Mbit isn't unrealistic. Ah, of course, thanks for clearing it up :)

mmm

14-05-2006, 09:30

Speed of 10Mbit half duplex connection depends a lot on how the machine is set up on the TCP settings, etc.

7.5Mbit isn't unrealistic.

The speed test is a bit optimistic I think (but repeatable!), because each download packet size is only half a second. Sustained speed on big downloads is about 6Mbps, but I'm happy with this as it means the rate limiting point in any real use is generally the source not the modem.

Rik

14-05-2006, 09:41

I would like to add that all these folks relying on HTTP speed tests, they arent very accurate.

To see that all you have to do is change your proxy settings, through my normal proxy (no proxy settings) i was getting around 6Meg, changed to a Hersham Proxy and the difference was much more noticeable.

A good speed test and very accurate is to use NNTP (Usenet) I connect via Giganews Europe Server and my connection is MAXXED out on 10Meg :)

1208k/Bs :)

Chrysalis

14-05-2006, 20:05

mmm

14-05-2006, 23:53

MTU should be 1500 on NTL cable (My VPN software reduces this to 1300 by default)

RWIN (receive Window) probably most significant, mine is currently at 256960 using TCP 13323 windows scaling (=64240*4=1460*44*4). XP PRO auto adjusts this parameter but I don't think it uses the scaling so it doesn't get above 65535.

__spc__

15-05-2006, 06:28

Running the tweaks test here: http://www.dslreports.com/tweaks

I set the MTU to 1500 and the RWIN to the maximum recommended by the test analysis.

OLDo

15-05-2006, 11:28

If NTL are actually providing an "up to" 10Mbit service - they should advertise it as such. As per http://www.home.ntl.com/page/broadband3 it's "10Mbit broadband", not "upto" or "near to".

As far as i'm aware, it is mainly ADSL connections that provide "up to" services. This is due, in part, to the nature of the *dsl technology, in that it deteriorates heavily over long cable distances. ISPs are therefore required to state that they are "up to" connections, because each individial line will sync at a speed dependant on their distance from the exchange.

Cable is different. Due to the way it is brought into the users home - everyone should be able to connect at 10mbit speeds, assuming they have the correct hardware.

__spc__

15-05-2006, 11:33

CS made it clear to me that it was upto 10Mb, guaranteed 2MB until 1 June, then guaranteed 4Mb thereafter...

OLDo

15-05-2006, 11:38

strange.. i hate the way ISPs can mince their words so easily and get away with it. :td:

Chrysalis

16-05-2006, 00:35

some more fiddling done here, if I turn of mtu discovery transfers speed up almost instantly instead of a triangle on graph and slow increase but I know this will cause me problems on anything that wont work with mtu 1500.

My rwin is currently on 128280 which helps a lot outside of peak hours but is detrimental in peak hours, I find in peak hours lower rwin is better because the retransmits are less damaging to speed and outside peak hours when there is no retransmits the larger window is effective.

moon_t

18-05-2006, 13:54

To add fuel to the fire, I have an NTL home 100 modem from the days of 512k, I have been very loyal :) and have never been able to get above 5.1 meg on any speed tests or sustained downloads in Windows or Linux (Leicester Area). I spoke to the notorious Indian call centre believing my modem was the problem, they advised me all kinds of crap that I knew wouldn't work (delete your cache to increase your speed etc etc... plus they had no idea about my Linux box settings).

Anyway I digress, the upshot is they couldn't sort my speed (or lack of) so I downgraded to 2mb, knowing that the 4mb upgrade is just around the corner, and strangley enough I get a full 2mb service every time. Why pay for a service they can't provide to me?

Roll on ADSL2 when I'll do what I haven't for the past 6 years and desert ntl entirely, taking my phone, internet and TV with me.

__spc__

18-05-2006, 17:55

Did you investigaet MTU and RWIN settings? Tweaking these raised my speed test results from 5 to 9Mb...

moon_t

18-05-2006, 18:55

__spc__

18-05-2006, 20:16

user2006

18-05-2006, 22:13

I used a number of tweak utilities but didn't alter the MTU or RWin manually, I have to admit that I'm loathe to upgrade back to the 10mb service and the extra cost just to try manual alteration of the settings.

Having seen no difference on my Fedora Core box I'm not altogether convinced it would be the settings anyhow. From the number of comments about the Leicester area on the forums, I have assumed that ntl just provide an inferior service to customers in this location.

Just did a test as i'm in the leicester area:
Thu, 18 May 2006 21:09:17 UTC

1st 512K took 437 ms = 1171.6 KB/sec, approx 9654 Kbps, 9.43 Mbps
2nd 512K took 438 ms = 1169 KB/sec, approx 9633 Kbps, 9.41 Mbps
3rd 512K took 437 ms = 1171.6 KB/sec, approx 9654 Kbps, 9.43 Mbps
4th 512K took 766 ms = 668.4 KB/sec, approx 5508 Kbps, 5.38 Mbps

Overall Average Speed = approx 8612 Kbps, 8.41 Mbps

2'nd attempt:
Thu, 18 May 2006 21:10:47 UTC

1st 512K took 453 ms = 1130.2 KB/sec, approx 9313 Kbps, 9.09 Mbps
2nd 512K took 422 ms = 1213.3 KB/sec, approx 9998 Kbps, 9.76 Mbps
3rd 512K took 437 ms = 1171.6 KB/sec, approx 9654 Kbps, 9.43 Mbps
4th 512K took 438 ms = 1169 KB/sec, approx 9633 Kbps, 9.41 Mbps

Overall Average Speed = approx 9650 Kbps, 9.42 Mbps

Chrysalis

18-05-2006, 23:15

I notice you are leicestershire rather then leicester, which part of leicester you in?

I agree with the other poster that judging from the posts I have seen on here and another forum well over 80% of leics posters are complaining of poor speeds which seems high compared to other areas.

My ubr has just had a reseg but things havent improved to the extenct that my performance matches different areas.

Thu, 18 May 2006 22:14:00 UTC

1st 512K took 751 ms = 681.8 KB/sec, approx 5618 Kbps, 5.49 Mbps
2nd 512K took 711 ms = 720.1 KB/sec, approx 5934 Kbps, 5.79 Mbps
3rd 512K took 571 ms = 896.7 KB/sec, approx 7389 Kbps, 7.22 Mbps
4th 512K took 811 ms = 631.3 KB/sec, approx 5202 Kbps, 5.08 Mbps

Overall Average Speed = approx 6036 Kbps, 5.9 Mbps

my conclusion is that the routing over ntl's network for the leics area is unoptimal and perhaps using poorer peering links then other areas. Some sites show this eg. newshosting and gigenews providers allow reverse traceroutes and the routes to ntl are different depending on what area you in and go in over different peers/transit so I can see this is probably true for much of the internet.

Also I have been told by more then one person that some parts of leics network that are analogue based have been done badly in the days they were installed so this probably affects it and some parts are so bad thats why they cant have broadband. I dont know if this has any bearing on speeds but it would be something to consider.

James Henry

18-05-2006, 23:26

moon_t the ntl 100 can't get anywhere near 10Mbit so yeah your modem was certainly the problem.

The general concensus from here appears to be that the Indian call centre, and that's all it is a call centre doesn't deserve to be dignified with the title of technical support centre, is useless.

About the only service it appears to provide most of the time is to get customers off the phone, which of course makes ntl ecstatic as we all know they do not and have never cared about customer service beyond paying it lip service in lame internal and external propaganda.

---------- Post added at 23:26 ---------- Previous post was at 23:19 ----------

my conclusion is that the routing over ntl's network for the leics area is unoptimal and perhaps using poorer peering links then other areas. Some sites show this eg. newshosting and gigenews providers allow reverse traceroutes and the routes to ntl are different depending on what area you in and go in over different peers/transit so I can see this is probably true for much of the internet.

That doesn't explain why a speed test inside the ntl network wouldn't score very highly.

Also I have been told by more then one person that some parts of leics network that are analogue based have been done badly in the days they were installed so this probably affects it and some parts are so bad thats why they cant have broadband. I dont know if this has any bearing on speeds but it would be something to consider.

No every network everywhere is analogue based, there is no digital component to the bare bones physical layer, even the lasers powering the fibre optics are analogue modulated. Digital signals carried are infact digital running over analogue transport.

Leicester does have its' own unique issues though. Student population and the original network build being two of them, but to blame peering for an internal network speedtest being slow is clearly not appropriate.

moon_t

18-05-2006, 23:26

Well I decided to give the tweaks shown here and on other sites a go.
So upgraded online back to 10mb, spent 3 hours tweaking settings to
recommended or identified by tweaking programs as being optimal,
and here is what I got :td:

1st 512K took 1000 ms = 512 KB/sec, approx 4219 Kbps, 4.12 Mbps
2nd 512K took 1000 ms = 512 KB/sec, approx 4219 Kbps, 4.12 Mbps
3rd 512K took 1000 ms = 512 KB/sec, approx 4219 Kbps, 4.12 Mbps
4th 512K took 1016 ms = 503.9 KB/sec, approx 4152 Kbps, 4.05 Mbps

Overall Average Speed = approx 4202 Kbps, 4.1 Mbps

Thu, 18 May 2006 22:22:56 GMT

1st 512K took 953 ms = 537.3 KB/sec, approx 4427 Kbps, 4.32 Mbps
2nd 512K took 984 ms = 520.3 KB/sec, approx 4287 Kbps, 4.19 Mbps
3rd 512K took 1016 ms = 503.9 KB/sec, approx 4152 Kbps, 4.05 Mbps
4th 512K took 1015 ms = 504.4 KB/sec, approx 4156 Kbps, 4.06 Mbps

Overall Average Speed = approx 4256 Kbps, 4.16 Mbps

Unfortunately I'm now stuck with the 10mb for 30 days before I can
downgrade again, at least I have plenty of time to try tweaking :)

James Henry

18-05-2006, 23:27

Think 4Mbit has got to be the way to go for you sir :)

moon_t

18-05-2006, 23:41

Just did a test as i'm in the leicester area:
Thu, 18 May 2006 21:09:17 UTC

1st 512K took 437 ms = 1171.6 KB/sec, approx 9654 Kbps, 9.43 Mbps
2nd 512K took 438 ms = 1169 KB/sec, approx 9633 Kbps, 9.41 Mbps
3rd 512K took 437 ms = 1171.6 KB/sec, approx 9654 Kbps, 9.43 Mbps
4th 512K took 766 ms = 668.4 KB/sec, approx 5508 Kbps, 5.38 Mbps

Overall Average Speed = approx 8612 Kbps, 8.41 Mbps

2'nd attempt:
Thu, 18 May 2006 21:10:47 UTC

1st 512K took 453 ms = 1130.2 KB/sec, approx 9313 Kbps, 9.09 Mbps
2nd 512K took 422 ms = 1213.3 KB/sec, approx 9998 Kbps, 9.76 Mbps
3rd 512K took 437 ms = 1171.6 KB/sec, approx 9654 Kbps, 9.43 Mbps
4th 512K took 438 ms = 1169 KB/sec, approx 9633 Kbps, 9.41 Mbps

Overall Average Speed = approx 9650 Kbps, 9.42 Mbps

My brother lives in the county and has no problems with his connection, I am in central Leicester in a high density residential area, wherein lies the problem I believe. At least 4meg will be consistent, so I suppose it was worth the tenner to find out if it was my settings.

Of course thats in addition to my modem being crap and old :D.

etccarmageddon

19-05-2006, 00:17

mine is the 10meg tier but most often comes out at 6-7meg despite the occasional 9.5meg result.

Chrysalis

19-05-2006, 22:45

That doesn't explain why a speed test inside the ntl network wouldn't score very highly.

No every network everywhere is analogue based, there is no digital component to the bare bones physical layer, even the lasers powering the fibre optics are analogue modulated. Digital signals carried are infact digital running over analogue transport.

Leicester does have its' own unique issues though. Student population and the original network build being two of them, but to blame peering for an internal network speedtest being slow is clearly not appropriate.

To be honest I generally get low on that test anyway, and I still have to traverse ntl's network to ntlworld.com eg. if I do a test of my alternative speedtest site I can often beat ntlworld.com which suggest their is an issue with ntls core network between me and the site or a proxy problem.

Also although that test was not max speed it was still higher then I currently get from telia which also indicates their is some kind of saturation between me and telia also obviously it just isnt telia with problems which is where my peering comments were aimed at not the ntl speedtest.

Fri, 19 May 2006 21:35:09 UTC

1st 512K took 611 ms = 838 KB/sec, approx 6905 Kbps, 6.74 Mbps
2nd 512K took 681 ms = 751.8 KB/sec, approx 6195 Kbps, 6.05 Mbps
3rd 512K took 721 ms = 710.1 KB/sec, approx 5851 Kbps, 5.71 Mbps
4th 512K took 1262 ms = 405.7 KB/sec, approx 3343 Kbps, 3.26 Mbps

Overall Average Speed = approx 5574 Kbps, 5.44 Mbps

^^ ntlworld.com

Fri, 19 May 2006 21:35:55 UTC

1st 512K took 441 ms = 1161 KB/sec, approx 9567 Kbps, 9.34 Mbps
2nd 512K took 420 ms = 1219 KB/sec, approx 10045 Kbps, 9.81 Mbps
3rd 512K took 561 ms = 912.7 KB/sec, approx 7521 Kbps, 7.34 Mbps
4th 512K took 431 ms = 1187.9 KB/sec, approx 9788 Kbps, 9.56 Mbps

Overall Average Speed = approx 9230 Kbps, 9.01 Mbps

^^ http://www.chrysalisnet.org/ntl/hfspeedtest2.html

the reseg has done a world of good here actually finally :)

---------- Post added at 22:45 ---------- Previous post was at 22:34 ----------

1 7 ms 15 ms 10 ms 10.227.112.1
2 34 ms 14 ms 15 ms leic-t2cam1-a-ge914.inet.ntl.com [82.3.35.149]
3 12 ms 12 ms 11 ms leic-t2core-a-ge-210-0.inet.ntl.com [82.3.33.6]
4 21 ms 14 ms 11 ms lee-bb-a-so-220-0.inet.ntl.com [62.253.188.110]
5 34 ms 18 ms 24 ms pop-bb-b-so-100-0.inet.ntl.com [62.253.185.238]
6 23 ms 25 ms 33 ms gfd-bb-a-so-500-0.inet.ntl.com [213.105.172.137]
7 71 ms 16 ms 19 ms gfd-bb-b-ae0-0.inet.ntl.com [213.105.172.6]
8 16 ms 72 ms 39 ms redb-ic-1-as0-0.inet.ntl.com [213.105.174.138]
9 23 ms 16 ms 23 ms alice-rose.ukcore.as33970.net [195.66.226.247]
10 * * * Request timed out.

route to alternative speedtest hosted by me

1 38 ms 10 ms 27 ms 10.227.112.1
2 13 ms 10 ms 12 ms leic-t2cam1-a-ge914.inet.ntl.com [82.3.35.149]
3 13 ms 11 ms 16 ms leic-t2core-a-ge-220-0.inet.ntl.com [82.3.33.10]
4 29 ms 16 ms 37 ms lee-bb-a-so-220-0.inet.ntl.com [62.253.188.110]
5 25 ms 21 ms 18 ms pop-bb-b-so-100-0.inet.ntl.com [62.253.185.238]
6 25 ms 17 ms 14 ms pop-bb-a-ge-000-0.inet.ntl.com [213.105.174.229]
7 17 ms 21 ms 27 ms win-bb-b-so-500-0.inet.ntl.com [62.253.185.202]
8 28 ms 23 ms 17 ms win-dc-b-v902.inet.ntl.com [62.253.187.226]
9 21 ms 15 ms 24 ms homepage02-win.server.ntlworld.com [62.253.162.12]

route to ntl speedtest which is different from hop3 onwards, as the test data is over the proxies obviously the actual route could be different.

mmm

19-05-2006, 23:27

No real difference from here (still got the Terayon in the way)

Fri, 19 May 2006 22:23:35 GMT

1st 512K took 500 ms = 1024 KB/sec, approx 8438 Kbps, 8.24 Mbps
2nd 512K took 500 ms = 1024 KB/sec, approx 8438 Kbps, 8.24 Mbps
3rd 512K took 500 ms = 1024 KB/sec, approx 8438 Kbps, 8.24 Mbps
4th 512K took 547 ms = 936 KB/sec, approx 7713 Kbps, 7.53 Mbps

Overall Average Speed = approx 8257 Kbps, 8.06 Mbps

Bandwidth for test provided by hostingfreak.
To repeat this test from the source server click here.

nthellworld.co.uk test

Fri, 19 May 2006 22:24:53 GMT

1st 512K took 500 ms = 1024 KB/sec, approx 8438 Kbps, 8.24 Mbps
2nd 512K took 500 ms = 1024 KB/sec, approx 8438 Kbps, 8.24 Mbps
3rd 512K took 500 ms = 1024 KB/sec, approx 8438 Kbps, 8.24 Mbps
4th 512K took 563 ms = 909.4 KB/sec, approx 7493 Kbps, 7.32 Mbps

Overall Average Speed = approx 8202 Kbps, 8.01 Mbps

Chrysalis

19-05-2006, 23:42

unusual on both tests your first 3 values bang on same number.

etccarmageddon

20-05-2006, 01:05

unusual on both tests your first 3 values bang on same number."it's not unusual...."

Sat, 20 May 2006 00:05:23 UTC

1st 512K took 561 ms = 912.7 KB/sec, approx 7521 Kbps, 7.34 Mbps
2nd 512K took 561 ms = 912.7 KB/sec, approx 7521 Kbps, 7.34 Mbps
3rd 512K took 561 ms = 912.7 KB/sec, approx 7521 Kbps, 7.34 Mbps
4th 512K took 941 ms = 544.1 KB/sec, approx 4483 Kbps, 4.38 Mbps

Overall Average Speed = approx 6762 Kbps, 6.6 Mbps

Sat, 20 May 2006 00:06:21 UTC

1st 512K took 561 ms = 912.7 KB/sec, approx 7521 Kbps, 7.34 Mbps
2nd 512K took 561 ms = 912.7 KB/sec, approx 7521 Kbps, 7.34 Mbps
3rd 512K took 561 ms = 912.7 KB/sec, approx 7521 Kbps, 7.34 Mbps
4th 512K took 561 ms = 912.7 KB/sec, approx 7521 Kbps, 7.34 Mbps

Overall Average Speed = approx 7521 Kbps, 7.34 Mbps

Chrysalis

21-05-2006, 02:42

etccarmageddon

21-05-2006, 10:45

internet exploder

monkey2468

21-05-2006, 11:12

it is unusual, something not right there.

Their has to be some type of variation, even if you throttled it to 100kbit you would get 4 different set of results probably around 90-95kbit.

What browser are you using for the speedtest?

Wouldn't a router cause those type of results, ie that the max throughput has hit the limit? or is using USB?

mmm

21-05-2006, 11:29

I use Firefox, with a 'busy' connection:-

Sun, 21 May 2006 10:23:36 GMT

1st 512K took 671 ms = 763 KB/sec, approx 6287 Kbps, 6.14 Mbps
2nd 512K took 594 ms = 862 KB/sec, approx 7103 Kbps, 6.94 Mbps
3rd 512K took 500 ms = 1024 KB/sec, approx 8438 Kbps, 8.24 Mbps
4th 512K took 531 ms = 964.2 KB/sec, approx 7945 Kbps, 7.76 Mbps

Overall Average Speed = approx 7443 Kbps, 7.27 Mbps

Sun, 21 May 2006 10:27:46 GMT

1st 512K took 641 ms = 798.8 KB/sec, approx 6582 Kbps, 6.43 Mbps
2nd 512K took 796 ms = 643.2 KB/sec, approx 5300 Kbps, 5.18 Mbps
3rd 512K took 657 ms = 779.3 KB/sec, approx 6421 Kbps, 6.27 Mbps
4th 512K took 500 ms = 1024 KB/sec, approx 8438 Kbps, 8.24 Mbps

Overall Average Speed = approx 6685 Kbps, 6.53 Mbps

There's another couple of exactly 500ms timings. I can't see what the exact formula is but is seems to me that the resolution is only about 15-30ms in the timing, therefore if my speed is limited by the Terayon not too surprising gets exact repeats

etccarmageddon

21-05-2006, 12:05

I'm not on USB and my router is a linksys wrt54g plus I have had test results of 9.5meg up until a week ago and my results aren't always 3 or 4 identical ones.

---------- Post added at 12:05 ---------- Previous post was at 12:04 ----------

ps. in the last week I've also had to do the proxy dance a lot.