How accurate are speeds on dslchecker.bt.com?

How accurate are the speeds on dslchecker.bt.com?

My cabinet was just activated for FTTC: https://puu.sh/urRpt/1f94e31489.png

Are these speed estimates for FTTC conservative? Do you think I could actually get a higher downstream speed than the 53.5Mbps top figure it claims I can get?

The distance from my home to my local cabinet is about 470 meters, according to this chart: http://www.increasebroadbandspeed.co.uk/2013/chart-b...

this means I should be able to get around a 60Mbps downstream speed, which is 7Mbps higher than the dslchecker says I should be able to get?

So, what do you guys think?

They're usually pretty accurate and I believe they take into consideration an element of crosstalk.

Just because you're 470m from the cab doesn't mean your line is that long. The routing often doesn't take the path you'd think.

If it means anything the checker shows max downstream 70.3 on clean for my address. I get 76 sync though with dated wiring. So it is possible that you could perform better than this estimate but it is not a guarantee.

They aren't very accurate for new builds. For my line the impacted range in the checker is 36.3 - 22.3 Mbps when the reality is the line can only support c. 16 Mbps. Somewhat less than the 28 Mbps minimum BT suggested I would get.

Normally they are fairly accurate.

How accurate is your 470 metres? Is that the actual path length? Does it take into account the FTTC cab to PCP distance? And then distance from PCP to house? What about the extra 5 metres to go up the pole? Or maybe 2 metres for the drop underground? Or a loop inside a footway chamber? Or the run from eaves to socket? All though little bits can soon add an extra 10 to 20 metres and thus a further drop in potential speed.

Simon, I know you've been in these forums a long time, but have you picked up that new builds, which you imply you are in, can have disastrous (for broadband) internal phone wiring? Including the physical cabling used, quite apart from the way it is wired up.

Plus of course the usual lack of optimisation by many users. Several megabytes difference can ensue.

If you haven't already made sure none of this applies to your line I suggest a new thread with an appropriate Subject might help you.

(And your Profile says ADSL2+).

Post deleted by kasg

In reply to a post by simon194:

They aren't very accurate for new builds. For my line the impacted range in the checker is 36.3 - 22.3 Mbps when the reality is the line can only support c. 16 Mbps. Somewhat less than the 28 Mbps minimum BT suggested I would get.

Whilst I agree somewhat, it's mostly down to the poor cabling builders use.

6m of internal wiring from the entry point to the NTE (installed by Bellway) caused my ADSL signal to drop from 2.7meg, to 0.6. I dread to think of the impact it would have om VDSL.

In reply to a post by mattocaster6:

Are these speed estimates for FTTC conservative? Do you think I could actually get a higher downstream speed than the 53.5Mbps top figure it claims I can get?

The ranges quoted are meant to come from similar lines - which means they have a similar amount of loss (attenuation) measured between the cabinet and the DP at the top of the pole.

The figures at the top and bottom of the range are meant o match the 20th and 80th percentiles - while means 20% of "similar lines" get speeds better than the top, and 20% get worse than the bottom.

Those "similar lines" will all have some amount of crosstalk (interference from other FTTC subscribers). On a new cabinet, the first people to sign up are likely to get significantly better speeds at first ... but as more people sign up in your neighbourhood, the speed will fall - usually to within the range somewhere.

Openreach will investigate slow lines, but the range they choose (A or B) will depend on the installation method used:
- ISPs that use self-install tend to quote the B range speed, so will only investigate when your speed is well below 26Mbps. For example, TalkTalk
- ISPs that use engineer installations tend to quote the A range speed, so will investigate when your speed is well below 39Mbps. For example, Plusnet.

Don't go into this thinking you can get 60Mbps, never mind keep 60Mbps even if you start with it.

I've read plenty about the potential wiring issues in new-builds. In mine the cabling comes into the house through the wall and runs behind the plasterboard and has the bedroom extension connected to this somewhere before it reaches the master socket.

Sky arranged a visit from an Openreach engineer last Monday because the speed was significantly below their 33 MBps minimum they quoted. I politely asked him if he could run a new cable from the BT66 to the master socket to cut out the original cabling which he did. The sync only went from 14 to 16 MBps which was probably because he did a DLM reset.

Still not sure what relevance that had to being a new build.

if youre 470 metres then i dont think you wlill get 60, and you have to also have to take into account crosstalk, which means the more people who join the quicker the drop will be on the max attainable.

im 300-350 metres away and get 70, i used to get 80 but as i say crosstalk.

if you do want fibre then i would go with bt, considering it says you can get a max of 53, their lowest package is 52/10, every other isps is 40/10, get that extra 12Mbps without having to pay a ton more.

In reply to a post by arronlowley:

if you do want fibre then i would go with bt, considering it says you can get a max of 53, their lowest package is 52/10, every other isps is 40/10 ....

Uh uh!!!

TalkTalk and Plusnet are 40/2. Probably some others as well.

In reply to a post by arronlowley:

if youre 470 metres then i dont think you wlill get 60, and you have to also have to take into account crosstalk, which means the more people who join the quicker the drop will be on the max attainable.

im 300-350 metres away and get 70, i used to get 80 but as i say crosstalk.

I'm 600 metres from the cabinet, which I was the fourth connection to. That was in 2011 so I'm sure there will be many more now.

Max: Upstream rate = 14201 Kbps, Downstream rate = 65850 Kbps
Bearer: 0, Upstream rate = 14193 Kbps, Downstream rate = 65258 Kbps
............Down Up
SNR (dB): 6.1 6.2
Attn(dB): 19.0 0.0
Pwr(dBm): 13.7 7.4

I pray I'm as lucky. I'm about 550m from ours

If you have ADSL at present, I would recommend you to tidy up etc including Quiet Line Tests, to first get the maximum from that - as it is good preparation for any later upgrade to VDSL.

What do you mean by "Quiet Line tests"?

Dial 17070 if you're on BT.

There is a facility available on most BT lines and on SKY lines, to test the lines for excess noise in the AUDIO Spectrum; but which if present, of ten affects ADSL and VDSL.

With the simplest, preferably corded phone available to you, plug it in to the Master Socket, also often described as an NTE, Network Terminating Equipment, preferably directly without the usual "dangly" xDSL/Modem Filter.

With recent NTEs, this may involve taking off the top which generally has a built-in equivalent, giving access to the otherwise hidden Test Socket.

Plug the old, simple, corded phone in.

On hearing the Dial Tone, Dial 17070

Assuming you get an answer, select Option 2 on BT or Option 4 on SKY, both described as "Quiet Line Test".

Basically, the line should go very Quiet, with possibly an extremely faint constant hum at most.

Listen for at least 5 repetitions of the QLT announcement, about 1 minute total.

(The SKY one is slightly different but achieves similar results.)

Hopefully your line will be "quiet"/clean.

If you hear "SNAP - CRACKLE - POP" or Engineering Tones etc, report is a PHONE LINE Fault to your PHONE Supplier - ie the Company that you pay PHONE Line Rental and any Call Charges to.

AVOID mentioning Broadband in any shape or form.

----------------------

If the Line is quiet, reconnect it back to normal, then try the QLT again from any sockets within your house.

If quiet - good.

If noisy, you need to try to locate the source, as generally it will be either faulty equipment somewhere in your house, or faulty phone or electric wiring or possibly induced from TV sets etc.

This may involve working progressively through any such equipment, switching off and preferably unplugging individual items, until hopefully you find the cause.

As the "noise" is not coming in on the phone line, it is for you to sort it out, although your local computer shop may offer the service, at a cost.


It may seem a nuisance and lengthy procedure - it generally is actually very short, from a few minutes to possibly an hour.

And it is highly advisable to sort such aspects out BEFORE upgrading to VDSL - the much higher frequencies involved are much more affected by such faults.

If you check the BT and SKY Web sites, the QLT is mentioned very early in their fault-finding procedures for Customers.

At 440m I am currently at 72Mbps sync ... and if one noisy neighbour turns their connection off I can get nearer 75Mbps. The 72 sync gives around 65 or 66 actual

So at 470 metres over 60Mbps can be achieved.


edit to add:

Have just run up DSL stats and it is showing max achievable of 79272 kbps and an SNR of 4.2 ... looks as though the low SNR trial has been expanded or is rolling out.

Edited by MHC (Sat 04-Mar-17 11:51:23)

In reply to a post by MHC:

Have just run up DSL stats and it is showing max achievable of 79272 kbps and an SNR of 4.2 ... looks as though the low SNR trial has been expanded or is rolling out.

I'm not sure 4.2 proves anything, I have often had it that low for several years. Currently on 5 but it was 4.8 when I looked a few minutes ago.

Edited by kasg (Sat 04-Mar-17 12:01:05)

Mine has been sitting at 6.1 for a long time and rarely moved more than 0.1 ... there was a 10 minute loss of sync at around 1:00 AM on Thursday. The line is now synced at 79.3Mbps and speedtests are up at 74Mbps.

Trial hasn't been expanded.

If that is the case, then how can my line stats be explained?

Power outage for the whole cab and you synced as one of the first with a 6db. More subscribers come online and your snr drops?

Though, to be fair, when that happens, the reduction in SNRM goes hand-in-hand with a reduction in the max attainable figure too.

Usually, anyway.

Best way to figure out the current target - do a manual power cycle, and see what the outcome is.

I'm still on ADSL2 and about a mile from my fibre cabinet by road.

A few months ago, the checker was showing a clean high downstream of 16.5 Mbps but since then, it's dropped to 5 (the speed I'm on at the moment).

Does anyone know what's caused this large decrease?

Is it likely to go back up to 16.5 or stay on 5?

Thanks.

Try re-syncing to the DSLAM and see what happens.

Was this a new cabinet when you checked the first time?

Edited by deleted (Sat 04-Mar-17 21:00:20)

It would have been.

Is the decrease most likely down to the cabinet filling up? There have been some new houses built at the end of my road in the last year.

It's more likely due to Openreach getting more data on people on your cabinet ordering the service and then more accurate data is available. It's likely that someone around you, on the same DP, ordered FTTC so Openreach have their connection data to provide a more accurate estimate for you.

Have you tried Openreach to see if they plan FTTP for you?

Back to your original question, I find that DSLchecker is too variable from test to test in the same time slot, to be relied on. I find Speedtest.net to be more consistent.

Test One

Test Two

Test Three



TBB admins swear by their own test, although I have my doubts. I get tests equally bad during on and off-peak times, suggesting that their whole test is internally dodgy, whether they are multi-thread or not.

If you really want an accurate measure for FTTC I would use a file download method or speedtest.net multiple times and then average the results.

In the end though, speed tests are only worth getting into if you are finding that download speed issues are affecting the way you use your connection.

http://www.thinkbroadband.com/speedtest/button/14887...

Edited by deleted (Sun 05-Mar-17 09:36:09)

In reply to a post by roughbeast:

Back to your original question, I find that DSLchecker is too variable from test to test in the same time slot, to be relied on. I find Speedtest.net to be more consistent.

Test One

Test Two

Test Three



TBB admins swear by their own test, although I have my doubts. I get tests equally bad during on and off-peak times, suggesting that their whole test is internally dodgy, whether they are multi-thread or not.

If you really want an accurate measure for FTTC I would use a file download method or speedtest.net multiple times and then average the results.

In the end though, speed tests are only worth getting into if you are finding that download speed issues are affecting the way you use your connection.

http://www.thinkbroadband.com/speedtest/button/14887...

What have any speedtests got to do with the estimates from BTs DSLChecker?

As for our test being internally dodgy, care to elaborate on that?

The way speedtest.net drops the slowest 30% of samples and top 10% of samples and then calculates an average will mask the up/down speeds the graph you have from our test, our using a median will generally show a slower figure. So two methods and which is right depends on what you information you want to convey.

Nothing wrong the test itself that I have written e.g.
http://tbb.st/1488381736597421555 and http://tbb.st/1488381621221545555 from same user

And from someone in Scotland http://tbb.st/1488476575635064255

In reply to a post by MHC:

What have any speedtests got to do with the estimates from BTs DSLChecker?

Just what I was thinking. Particularly on a VM Media cable line.

Perhaps the poster's opinion isn't well-informed .

I am looking to upgrade to fibre, can only reach 4.5mb adsl speed because of the distance from exchange.
I have just done the BT dsl checker and these are the estimate download speeds etc,

VDSL A high 80 low 64.7


VDSL B high 78.9 low 54


FTTP 330

The cabinet is about 80 yards away, so should i be good to go to order a up to 38/40 mb fibre product.

With those estimates, you will probably get significant extra speed from an up to 77 Mbit/s compared to an up to 38 Mbit/s product. It is up to you to decide what is worthwhile, remembering that you can usually upgrade without penalty but often cannot downgrade without serving a minimum contract period.

I suspect the FTTP mention is FTTP On Demand which is extremely expensive. It is very unlikely that Openreach will provide both FTTP and high speed FTTC to the same location.

Yes. The "A" line is if you have a very clean line between the FTTC cabinet, through the PCP, to your master socket. The "B" line is if there is anything dodgy between the cabinet and reaching your house, and/or you have bad wiring in the house or set it up badly.

In your case it looks as if the only factor the estimator is covering with the B figures is what your internal wiring may be like, as BTW don't know about that.

Bearing in mind that faulty electrical equipment next door could also cause problems, but that applies to all of us.

You should have no problem at all with any 40Mbps product, but note both TalkTalk retail and Plusnet only provide 2Mbps upstream on that. Most ISPs supply 10Mbps. If you don't upload much it probably won't matter. But lots of photos, or file backups, and a 10Mbps one may be advisable.

Also, BT Retail these days is 55/10, not 40/10. To me that doesn't matter but I believe the 55 is to do with streaming the highest definition TV or something like that. 40/10 might struggle.

All ISPs that provide the 80Mbps product give 20Mbps upstream on it, but of course it costs more.

Whoops! I just realised I misread DSL checker as meaning DSLReports. Apologies.

I am happy for my post to be deleted.

However, I do stand by my comments regarding the utter inaccuracy of the TBB single thread and multi-thread tests. I do not apologise for that. But that is another story.

In reply to a post by RobertoS:

Yes. The "A" line is if you have a very clean line between the FTTC cabinet, through the PCP, to your master socket. The "B" line is if there is anything dodgy between the cabinet and reaching your house, and/or you have bad wiring in the house or set it up badly.

In your case it looks as if the only factor the estimator is covering with the B figures is what your internal wiring may be like, as BTW don't know about that.

Bearing in mind that faulty electrical equipment next door could also cause problems, but that applies to all of us.

You should have no problem at all with any 40Mbps product, but note both TalkTalk retail and Plusnet only provide 2Mbps upstream on that. Most ISPs supply 10Mbps. If you don't upload much it probably won't matter. But lots of photos, or file backups, and a 10Mbps one may be advisable.

Also, BT Retail these days is 55/10, not 40/10. To me that doesn't matter but I believe the 55 is to do with streaming the highest definition TV or something like that. 40/10 might struggle.

All ISPs that provide the 80Mbps product give 20Mbps upstream on it, but of course it costs more.

Ok thanks.

Also how does isp guaranteed minimum speed work, for example just checked with sky bb checker and it says i can get the up to 38/40mb product.
If i get much lower speeds, ruling out dodgy wiring, filters etc,would i be able to cancel this product without penalty?

In reply to a post by roughbeast:

I am happy for my post to be deleted.

You can edit or delete it yourself up until 21:34. (Twelve hours from original posting).

In your case, with those estimates, there would be a glaring fault and the job should not have been signed off by the engineer working at the cabinet.

Your line should run at 80Mbps if on the 80/20 product. The 40Mbps is simply an Openreach speed cap. The underlying technology is identical. Given a valid installation there is no way it will not achieve the 40Mbps you are intending to order .

The 38 figure by the way is simply the maximum they can advertise, and means at least 10% of people on the 40Mbps product get a 38Mbps connection. Note that connection speed and throughput speed are not the same thing at all, but I believe Sky are pretty good at throughput speeds as well.

And I've explained why the two will differ by sometimes large margins on services where the speed is fluctuating

In reply to a post by RobertoS:

In your case, with those estimates, there would be a glaring fault and the job should not have been signed off by the engineer working at the cabinet.

Your line should run at 80Mbps if on the 80/20 product. The 40Mbps is simply an Openreach speed cap. The underlying technology is identical. Given a valid installation there is no way it will not achieve the 40Mbps you are intending to order .

The 38 figure by the way is simply the maximum they can advertise, and means at least 10% of people on the 40Mbps product get a 38Mbps connection. Note that connection speed and throughput speed are not the same thing at all, but I believe Sky are pretty good at throughput speeds as well.

Thanks again thats very helpful.

In reply to a post by David_W:

With those estimates, you will probably get significant extra speed from an up to 77 Mbit/s compared to an up to 38 Mbit/s product. It is up to you to decide what is worthwhile, remembering that you can usually upgrade without penalty but often cannot downgrade without serving a minimum contract period.

I suspect the FTTP mention is FTTP On Demand which is extremely expensive. It is very unlikely that Openreach will provide both FTTP and high speed FTTC to the same location.

Thanks.