Just for contrast, my line is 520m, the attenuation 18.5dB and SNR 6dB. My errors are low and my connection stays up for weeks on end but I've never synced above 46M. Still 3M above the 'handback threshold' though. Just unlucky I guess.

It is just bad luck. It may have tried dropping to 5dB and found instability or maybe it sees a medium length line (at 18dB attenuation) with a fairly low sync for the attenuation and knows it will not be worth trying!. At 16.5dB over 450m I have sync speeds up at mid 70s but know of one specific interferer that takes 6-8 Mbps.

Yes, it is true that 3dB will sometimes drop to 2.x margins.

In my case when I was on EO Line 3dB was impossible to maintain stable sync. The internet connection wouldn't even last for half a day before dropping out. With 6dB it will also drop out within 2-4 days regardless of what ISP or router quality I buy.

In my case only a manual ISP cap of 9dB on ADSL was my connection stable and even then 9dB would drop to 3dB almost every day in the mornings due to some interference that would occur for years. After 10 minutes to 2 hours it recovers back to 9dB. But because the SNR is high enough the drop didn't go to 0dB.

I personally have observed in my router stats over the years with a 6dB or 3dB default SNR under ADSL formerly EO Line the SNR would never be stable and would very frequently drop to 1dB even witnessed 0.8dB and that gave me clues why my connection was dropping out. Like it would disconnect 5-6 times a day with a 3dB profile. No Openreach engineer or Sky engineer was able to fix this issue.

FTTC finally came and that is when miracles happened. Now the same 3dB SNR no longer poses problems and from my observation last 12 months since having FTTC. Noise margins are very stable. Extremely rarely have I observed even minor fluctuations and that's an indicator of a stable line thanks to shorter copper lengths I believe.

I think the longer the copper line to the cabinet the higher the risk there is to picking up noise and that can reduce speeds. Now my cabinet is around 320 meters and I have measured it with google maps. But my line attenuation shows as 17.90-18dB. Some have said that, my cabinet must be longer than that. Even if it is quite a bit longer than that my connection has almost always synced at 80Mbps.

There was only one occasion a few months ago following TalkTalk router firmware update that DLM dropped my speed to 79.6Mbps for 1-2 months and then it went back up to 80Mbps.

In reply to a post by BLaZiNgSPEED:

Yes, not quite normal. At 118 meters from cabinet you should get the full 80/20 speed test. Now you can wait of-course for the line to stabilize as DLM can sometimes take a while. Make sure you don't power off or reboot your router as that can disrupt the DLM process.

Crosstalk can also be a factor in reduced speeds. However, seeing that you have Virgin Media as well, I'd expect crosstalk to be less of a factor with a chunk of customers still using Virgin Media.

I sync at 80/20 Mbps in my TalkTalk router settings and I also get 80/20 in speed tests. Not absolutely every single time but most of the time I get speed tests of 80 Mbps download and 20 Mbps upload with various speed test sites.

Here's proof. https://i.imgur.com/6XC1WI7.jpg
https://i.speedof.me/201205000112-233
https://i.imgur.com/h01UK8f.jpg
https://i.imgur.com/k8Lmgoh.jpg

https://librespeed.org/results/?id=0vmhvmn Latest speed test just done now with Libre 80Mbps download, 20 Mbps upload.

And btw, I am 320 meters away from cabinet and that's with TalkTalk! I pay only £21.95 a month.

If you are 118 meters away, you should be guaranteed to receive 80 Mbps considering that my copper length is 202 meters longer than yours! In fact there are people here who claim to get 80Mbps with a 450 meter cabinet distance.

What does the Broadband Checker show? https://www.broadbandchecker.btwholesale.com/#/ADSL/...
I'm pretty certain that if your cabinet is 118 meters away from your property, barring any crosstalk you should get VDSL Range A (Clean) 80 79 20 19 VDSL Range B (Impacted) 80 79 20 19

78Mbps would be the lowest I can expect you to receive. I have been with TalkTalk FTTC last 1 year and my router has always synced at 80/20 along with my speed tests even though my lowest estimates are between 64.8 and 68.2 Mbps in the BT Wholesale checker.

You should complain to Zen! I am assuming that you are paying £34.99 a month and only getting 73/16 Mbps? That's certainly not acceptable.

Thanks for the additional guidance. Thought I’d provide an update.

It’s all a bit strange to be honest, and I’m not sure what to think.

I spoke to Zen tech support, and in a nutshell the advisor said that around 70 Mb/s down and 16Mb/s up is the fastest t I could realistically expect. This was disappointing, but I accepted it at face value because that was what I had read on here early on.

He didn’t say much about my occasional dropouts in service, but he suggested that I should do a factory reset of the router and then try plugging it directly into the Master test socket.

He also said that the staples through the cable won’t affect my connection, and that only if it had been an outdoor cable would there have been a concern due to rain getting in. Don’t know how true that is.

I hate doing factory resets (PTSD from Virgin Media), but I did it anyway. I removed the faceplate on the Master Socket, connected the separate filter, and then connected the router directly to the test socket.

It’s been a week and there has been no difference in speed. Also, there has been one dropout of service for around 5 minutes.

The weird thing is that after I saw your post I decided to test my speed using the sites you had. Previously, to check my speeds I had always used:

1) broadbandspeedchecker.co.uk: 73 Mbps down, 14 Mbps up

2) speedtest.net: 73 Mbps down, 16 Mbps up

3) broadbandtest.which.co.uk: 73 Mbps down, 14 Mbps up

4) checker.ofcom.org.uk/broadband-test: 73 Mbps down, 17 Mbps up

5) Google’s own speed test: 72 Mbps down, 16 Mbps up


Since these were all so similar, I assumed they were accurate.

However, using your test sites, I consistently achieved:

1) AAISP: 79 Mbps down, 20 Mbps up

2) Speedofme: 82 Mbps down, 21 Mbps up!!

3) Internetfrog: 78 Mbps down, 20 Mbps up

I’m surprised that there was such a significant difference, given that the tests were repeated around the same times with minimal load on the connection. I knew different test sites could give different speed results depending on the server you connected to, but I didn’t think it could be so substantial.

Is it simply a case of choose your test, choose your result?

But then why did the Zen advisor tell me that around 70 Mbps downstream was the best I could expect?

For my address, BT’s Broadband Checker shows: VDSL Range A (Clean) 80 79 20 19 VDSL Range B (Impacted) 80 79 20 19.

If I choose the second set of results as the “true” reflections of my speed, then BT’s data is correct and the Zen advisor should have told me that something was wrong.

However, he never asked what site I used to check the speed, and simply defaulted to an answer that there was nothing to be done.

So do protocol overheads always mean that you get a lower speed than your sync speed?

Are the other speed test sites compensating for this perhaps and giving an inflated figure (or burst speeds)?

I’m still getting the occasional dropout in service. Is this common with FTTC in London? Is one dropout a week the best I can hope for?

NO you should not expect it to be normal to get drop dropouts on a regular basis. My current uptime is 19 days and that is since the last power cut here.
Have you posted anywhere what your router stats are reporting particularly the maximum achievable rate, actual sync speed and snr level? As this will tell you what you can expect your sync speed to be and subsequently the download speed achievable.

For example I am on a 50mbps package. My sync is 55000kbps max speed is 83000 noise level 17.3 and speed tests (TBB) invariably show 50000. I would expect you to get similar treatment.

But when all said and done the difference between your 73 and the max of 80 would be imperceptible.

Good point, I should probably include that info somewhere.

This is what my Fritz Box is telling me:

Max. DSLAM throughput: kbit/s Receive: 80000 Send: 20000
Min. DSLAM throughput: kbit/s Receive: 128 Send: 128
Attainable throughput: kbit/s Receive: 87900 Send: 29604
Current throughput: kbit/s Receive: 79998 Send: 19999

Latency: Receive: fast Send: fast
Impulse Noise Protection (INP): Receive: 43 Send: 0
G.INP: Receive: on Send: off
Signal-to-noise ratio (dB): Receive: 8 Send: 15
Bitswap: Receive: on Send: on
Line attenuation (dB): Receive: 7 Send: 7


I think I’m coming to terms with the fact that it is unlikely I’ll be able to tell the difference between 70 Mb/s and 80 Mb/s so I should just stop thinking about it.

It’s just a shame because even my shockingly poor Virgin Media would give me 220/20 on a 200/20 package on all speed tests.

At last you have come to the correct conclusion.....

As you said. ...." So do protocol overheads always mean that you get a lower speed than your sync speed?"

Yes throughput speed is always lower than sync speed.

Also ... "Are the other speed test sites compensating for this perhaps and giving an inflated figure (or burst speeds)?"

Speed test sites can be optimistic, fairly accurate or totally wrong.

I think Blazingspeed has not helped by claiming he can get 80Mbps throughput with an 80Mbps sync and providing "proof" with a number of obviously impossible speedtest results

Not too sure what your issue is. Are you uploading and downloading a lot? Another 2 - 3 mbps are not going to be noticed on such a high speed.

I'm on a 40 / 10 line and certain speed test show a burst of 60 mbps. Good grief, I get for a split second an extra 20 mbps. Don't notice that at all when using FaceAche or looking at emails.

Just wondered if you have viewed the info on the Zen website? i.e. view line data.

You'll probably find that your line speed will vary over the course of a few months anyway. DLM isn't as such a static tool.

I use to be with Plusnet and always got around 72Mb/s. Moved to Zen (same cable etc) and it seems 78Mb/s sync is more my normal. But I certainly don't lose any sleep about the few Mb of extra speed.

Over the past 15 months I feel I've had a good service from Zen. Sadly their days are numbered for me as we have a local ISP installing FTTP (500Mb/s up and down). But if Zen were doing the biz, I'd still have no hesitation using them.

As for their tech support, used them once when I had a line fault and I'd find it very difficult to fault their handling of it. Just did the biz.

The AAISP and Speedofme tests always give me speeds that are above what my package allows.
Their results can't be trusted.

77Mb is about as high as an 80Mb FTTC line can go, after taking off overheads.

The DLM profile can also change the maximum throughput seen.

I use the thinkbroadband speed tester and the speedtedt.net tester.
Try a few different servers on the speedtest.net site/app.