I've been a BT Broadband user for a year or two and have been
reasonably happy with the speed. Since installing my new (much faster)
PC I haven't done any tests until today. But with symptoms of sluggish
downloads I just went to http://www.testmy.net/d_load.php and got a
consistent rating over several tests of only 0.8 Mbps.

Am I right that I should expect 3-4 Mbps? Any advice on practical
steps I should take to isolate the cause and get it fixed would be
appreciated please.

(Just checked speed of the Pipex Broadband connection my wife uses,
with same tests, and that was 2.1 Mbps.)

--
Terry, East Grinstead, UK

Terry Pinnell wrote:

Well, if you insist on using SHIT ISPs (both of them) what do you expect ?

I get > 6Mbps day in day out.

Graham

Terry Pinnell wrote:

http://www.speedtest.bbmax.co.uk/res...6221&v=4426374

100% predictable and probably only costs a few more folding per month than
yours does.

ISP ? Idnet. One of the very best in the UK. Maybe even THE best. NO
Indian call centres, totally British staff who know what they're talking
about you can call on an free 0800 number. Faults ? What faults ? If you
do get one it's rare it takes more than 2 hours before it's sorted wheras
Tiscali (Pipex) will hapilly take 2 MONTHS like they did with my
neighbour.

Fancy trying it ? There's no migration cost and just a one month contract.
They work on the basis that the best way to keep customers is to provide a
top-notch service instead of tying them into nasty contracts.

Email me on the address in my headers, move to Idnet mentioning my details
and we both get a £10 pressie.

Graham

As you are with BT use BT's speedtester.
Look in your routers control panel to determine your sync
speed, attenuation and SNR (Signal Noise Ratio).
Try the same speedtest with all phones stuff and
ADSL filter removed, plugging straight into BT's
master incoming line jack. Which will give the
fastest sync speed for the line and modem
combination.
Take the same measurements with and without
the phones and compare.
I trust that there is no long extension wiring
or other fancy wiring of phones, extensions,
fax machines, caller ID units, extesion ringers
involved, and that if you have extension wiring
it isn't running parallel to mains wiring, and that
your above quoted speed doesn't involve a
wirelees link to your router.
Report back, so we can comment on the
results of your tests.

The speed you get depends on a variety of factors such as the line length
and how noisy the line is. All sorts of other factors can mak a significant
difference. The first thing to do is to check the wiring and equipment.

This is a good place to start and explains the basic steps.

http://www.dslzoneuk.net/socket.php

Follow the steps carefully and if need be make the changes. Any suggestion
that just changing ISP is a cure for all broadband ills is ludicrous and
should be ignored.

Peter Crosland

Peter Crosland wrote:

Naturally. The likes of BT don't score badly at all - it's just a conspiracy.

Graham

On Thursday 05 June 2008 8:15 pm, in MID <48483B56.FEEA74E9@hotmail.com>,
Eeyore (rabbitsfriendsandrelations@hotmail.com) wrote:
That test site consistently give me results of about half (or worse) the
speed that I get from other speedtest sites, and from real-world downloads.

Who knows ?

What's your sync speed ?

Me too.

http://www.speedtest.bbmax.co.uk/res...1329&v=4427220

--
Ian...

Eeyore <rabbitsfriendsandrelations@hotmail.com> wrote:

Graham: Thanks. That test reported a very great improvement, to 3.4
Mbps download and 381 Kbps upload. I'm not averse to changing ISP as a
last resort, but I'd like to avoid the hassle if at all possible. So
I'll first do some more research and experimenting.

Peter: Thanks. Will study that as suggested.

---------

Just had a long conversation with BT Support, a pleasant and patient
chap in India. But although he is convinced from tests we conducted
together that I am "getting 4.4 Mbps; and that would be 6 Mbps if you
moved the Prestige router to your main phone socket downstairs", I am
so far unable to square that with *my* tests. The only test he appears
to regard as authentic/reliable is the one at
http://www.speedtester.bt.com/
I found that myself before phoning BT, but it fThat failed to work for
me. And during my phone session, on repeating the attempt I get a
message that I must wait 3 hours! Not a very useful test IMO.

BTW, in case it has any bearing, although I have successfully had a
Remote Assistance session with BT two months ago from this PC (over a
username/password issue) unfortunately it failed to work tonight.

Any further help would be welcomed, as I'm rather out of my depth on
this stuff.

--
Terry, East Grinstead, UK

The BT test is very specific and is intended to establish if the speed
problem is in the BT part of the network. For most other purposes it is, as
you have found out, useless. This is a much more useful tester

http://www.thinkbroadband.com/speedtest.html

Remember that the test results are only a snapshot at the time of the test.
A test at a busy time may not give a true picture of the line speed.

In order to eliminate any wiring faults it is normal to conduct the tests
using the test socket behind the reoveable faceplate but that does NOT
guarantee an increase in speed.

Be aware that changing ISP is a very last resort and quite pointless unless,
and until, you have eliminated any hardware/wiring issues. Also remember
that unless you move to a cable ISP the line to the exchange will be the
same regardless of which ISP you use though some ISPs have their own
equipment in the exchange. However, you do need to do the basic things
first!

Peter Crosland

ato_zee@hotmail.com wrote:

Thanks a lot, very helpful. I'll experiment along those lines and
report back.

The Prestige 600 Series router is connected by cable to my PC and the
phone extension socket here upstairs in my office. The master incoming
socket is downstairs. The connecting cable doesn't run closely to the
mains as far as I know.

I take it from your suggestion that the broadband connection will work
if I remove the ADSL filter, Is it just the voice line that will
suffer?

See also my later reply when I reported my discussion with BT, and the
failure of the BT speedtester. I'm unable to try it again yet, because
of the inflexible 3 hour wait ;-(

--
Terry, East Grinstead, UK

"Peter Crosland" <g6jns@yahoo.co.uk> wrote:

Thanks, appreciate the follow-up. That gave me a result of '3284.86
Kbps ( 3.2 Mbps )' which closely matches what I got earlier from
http://www.speedtest.bbmax.co.uk/res...6221&v=4426374

I just repeated the bbmax test and again got about 3.3 Mbps.
So, conclusions so far:

1) The test at http://www.testmy.net/d_load.php seems to read low. Now
showing 1.0. And Pipex on wife's PC, on different phone line (direct
connection) = 2.3. (So it does have merit as a *comparative* test. My
wife is getting more than twice my speed. Which also squares with the
fact that when you hover over her internet tray icon it reports
7.something. (My icon doesn't perform that function at all.) Is Pipex
generally regarded as faster than BT?

2) I'm not too unhappy with 3.2 if that's a realistic measure, but my
potential with current setup via extension is apparently 6.something
according to the BT guy. And he reckons I should get 4.4.

Understood. As per my earlier reply, it would definitely be a last
resort.

BTW, no one's mentioned settings on my PC. Can I assume that the
'default' settings that came with the PC, with XP Pro installed, are
unlikely to benefit from 'tweaking', either manually or via one of the
several tools I see mentioned on various sites?

--
Terry, East Grinstead, UK

ian <news.spam@glitton.org.uk> wrote:

Can you explain what that is, and how I measure it please?


Thanks. Yes, now using that, following Graham's recommendation.

--
Terry, East Grinstead, UK

You disconnect all the phone stuff, and plug your router directly
into the BT incoming line master socket.
This will give you your best speed.
If your filter/filtering is sub-standard then stuff on the phone side
can shunt off some of the ADSL frequencies. By removing
the phone stuff you eliminate the phone side affecting the ADSL.
Wiring on the phone side, including extensions, can act as
an aerial and pick up noise, which degrades the ADSL signal,
lowering the speed, so for the purpose of testing disconnect by
unplugging the phone side.
Running an extension cable to your router, particularly if it
runs alongside a mains cable can also pick up noise from the
mains cable, since extension cables are not normally
twisted pair to reduce noise pickup.
Putting the router next to the incoming BT socket will eliminate
the effect of an extension cable.
Once you have established your fastest speed, no phones
connected, no extension cable, router plugged into the
BT master socket, you can record the speed and router
control panel readings for sync speed, attenuation, SNR.
Now when you put things back as you normally have
them you can see from taking the same measurements if
you normal way of connecting is degrading the ADSL
and reducing your speed.
I found I got a significant increase in speed by moving my
router next to the phone socket and connecting my PC's
with CAT5 ethernet instead of an extension cable, I also
improved my filtering, which may have helped.
I suspect from experiments that the original extension
cable running next to the mains, was picking up noise
from low energy lamps, which I find contain two
switching power transistors, just what is needed to
generate noise, though anything in the
house with a switch mode PSU would probably
generate noise.

Beware trying to change more than one variable at a time. The standard
settings in XP should be fine. At this point concentrate on the
hardware/wiring. The sync speed is the speed at which you are connected to
the BT network. The actual throughput achieved will be less than that
depending on the amount of traffic on the network at the time of the test.
Comparisons of throughput are potentially misleading so just concentrate on
you sync speed for the time being. It would be helpful to know what
make/model of modem/router you are using.

Peter Crosland

Eeyore wrote:
Id say Peters statement is correct and incorrect at the same time. Changing
ISP will not fix any physical problems you have with your hardware or the BT
line - it won't increase your sync or IP profile. I've assumed here that
you will remain on BT kit in the exchange. Going LLU may increase sync but
if the BT line has problems it won't fix that. Changing ISP will fix issues
with traffic shaping and general throughput depending on which ISP you go
to.

--

WCZ

Terry Pinnell wrote:

Pretty much. I've never needed to. Some people talk of tweating MTUs or whatever
but I think maybe that's more relevant to cable modems.

Graham

Peter Crosland wrote:

If the network has enough capacity (unlikely for BT's retail ISP offering) and
there's no other congestion in the way (never noticed any worth talking of for
the bbmax speed test) the data speed should be about 85% of the sync speed.

Graham

WCZ wrote:

Whuch is why Idnet is so good.

And it may regularly fix issues relating to incompetent customer support. which
is the problem 90 + % of the time IME.

Graham

You are both correct and incorrect. The problem is that before gettingn into
the hypothetical advantages of LLU the OP needs to concentrate on the
basics. Given that the OP is not au fait with ADSL introducing more detail
before he has even sorted the basics is likely to be counterproductive.


Peter Crosland

"Peter Crosland" <g6jns@yahoo.co.uk> wrote:

I'm reporting back after following the steps recommended. Bottom line
is that direct connection to BT master socket with all phones
disconnected made no difference to speed. But I have discovered that
it seems 'capped' or 'configured' to 3.5 Mbps, whatever that means.
Here's the detail:

1) 08:00 Using current configuration, bbmax reported 3395 Kbps
2) 08:10 With both phones removed and a long extension to my
shed/workshop (with no phone currently at its other end) also removed:
bbmax = 3386
3) 08:12 Comparative test on wife's PC/Pipex/separate phone line:
bbmax = 6303
4) 08:50-08:53 No phones connected, and router direct to main socket
using long extension cable: bbmax = 3393, 3399, 3389
[Note: I say 'direct', but the router cable plug is not the type that
can fit into a phone socket. So I kept it connected via the filter,
but I assume that's a straight-through connection.]
5) 08:57 As #4, but with BT cover removed and extension connected
directly to socket inside (although I don't see why that *should* make
any difference?): bbmax = 3395
6) 09:25 As #5: BT Speedtest http://www.speedtester.bt.com/ now
worked, as it has been more than 3 hours since I last used it. First
gave this message:
"Your configured download throughput speed for this service is 3500
Kbps"
Then it ran the test and gave this result:
"Test1 comprises of Best Effort Test: -provides background
information.
IP profile for your line is - 3500 kbps
DSL connection rate: 448 kbps(UP-STREAM) 7616 kbps(DOWN-STREAM)
Actual IP throughput achieved during the test was - 3263 kbps"

Can someone explain that please, and advise what I do next?

7) 09:40 With everything restored to as it was: bbmax = 3394, 3385,
3368
8) 10:49 Repeated the above to see if the later hour made any
difference: bbmax = 2923

Note: I haven't reported 'Sync Speed' as I don't know how.

To remind you, my router is ZyXEL Prestige Series 600. When I open its
'Configurator' from the shortcut on my desktop I see:
http://img212.imageshack.us/img212/5...urator1wc4.jpg

Happy to provide any further information to help diagnose the problem
and improve this mediocre speed.

--
Terry, East Grinstead, UK