Kitz Forum

Broadband Related => Broadband Hardware => Topic started by: Alucidnation on December 30, 2018, 08:38:07 AM

Title: Draytek Vs Billion
Post by: Alucidnation on December 30, 2018, 08:38:07 AM
Hi guys,

I am sure this has been asked a million times before however...

I am on an ECI cab and i have recently replaced the ageing ECI OR modem & router with a Draytek 2762 (non wifi).

However, after reading a few threads etc, it seems a Billion would have been the better choice due to its superior chip?

Thinking of the Billion BIPAC 8900X R3 as i don't need wifi as i have access points.

Also, i'm sure i read somewhere that Billion were moving out of the market?

Anyway, hints, tips and abuse appreciated.

 ;D

Title: Re: Draytek Vs Billion
Post by: Weaver on December 30, 2018, 08:55:13 AM
I’m not sure about this, and I’m no expert in VDSL2 / FTTC, just the opposite, but from what I recall things are not so simple. I have a faint recollection that some people have said that the Sraytek works well with ECI. But anything with a Broadcom top-end chipset is supposed to be a good bet, and Billion has a good reputation. I use ZyXEL VMG 1312-B10A modems, but on ADSL2 on ultra slow lines. A number of kitizens, including Burrakucat and Kitz herself, use ZyXEL modems or modem-routers. Note that the newer ZyXEL VMG 1312-B10D unit is supposed to be inferior.
Title: Re: Draytek Vs Billion
Post by: atkinsong on December 30, 2018, 12:35:33 PM
As I've said before on here, I've tried all the recommended Broadcom options on my ECI cab, and none can match the combination of speed and stability offered by the Lantiq based Draytek 2760. At 6db snr it syncs around 2 Mb less than the Broadcom options but with far far fewer error seconds. With the Draytek snr tweaked to 4db it has around a 3Mb advantage over the Broadcoms, with only a slight increase in error seconds, typically around 50 per day.
Title: Re: Draytek Vs Billion
Post by: j0hn on December 30, 2018, 12:54:42 PM
Every line is different.

Some prefer Broadcom on their ECI cabinet (I definitely did h and others prefer Lantiq chipsets.

Unfortunately my line has far too many daily ES with Lantiq chipsets without lowering the SNR target.
My priority was maintaining fastpath on the line, which only the ZyXEL devices with Broadcom chipsets and the additional noise filters could manage.

I tried other Broadcom modems and a couple Lantiq devices and they would all but DLM ES limits and take the line interleaved.

It depends what the OP wants to achieve.
If it's aggressively go for higher sync and your ES are fairly low then the Lantiq modems can lower the SNR and squeeze more speed out the line.

If you believe OpenReach will fix G.INP and rollout lower SNR targets to ECI lines then having that as a feature wouldn't be needed.

The Broadcoms make stats monitoring easier.

Pro's and Con's to each chipset.
Title: Re: Draytek Vs Billion
Post by: Alucidnation on December 30, 2018, 01:11:32 PM
Thanks for the replies.

These is my stats so far on the Draytek

        
Downstream               Upstream               
Actual Rate   48447   Kbps   15023   Kbps
Attainable Rate   54535   Kbps   14654   Kbps
Path Mode   Interleave   Fast
Interleave Depth   706   1
Actual PSD   6. 7   dB   6. 9   dB
Near End                    Far End                   
Trellis   ON   ON
Bitswap   ON   ON
ReTx   0       0   
SNR Margin   5   dB   6   dB
Attenuation   18   dB   26   dB
CRC   0   85
FECS   25983154   s   2453   s
ES   0   s   36   s
SES   0   s   0   s
LOSS   0   s   131   s
UAS   34   s   8079   s
HEC Errors   0   0
RS Corrections   0   0
LOS Failure   0   0
LOF Failure   0   0
LPR Failure   0   0
NCD Failure   0   0
LCD Failure   0   0
NFEC   255   85
RFEC   16   16
LYSMB   4029   14992

I have had OR out a couple of times to rectify the speed issues as i was getting a healthy 65 DL a few weeks ago, and according to them, the speed plummeted almost overnight, so this was my thinking with changing the modem.

One engineer apparently tested the line back to the cab and said it was all ok and was going to reset it at the cabinet.

The second one swapped pairs all the way back to the cabinet, and so far nothing has improved.

I have had the ISP post up their test results and they are all saying my line length is 450 -650 meters from the cabinet so i dont know what to believe anymore. The results also showed no sign of crosstalk etc.



Title: Re: Draytek Vs Billion
Post by: j0hn on December 30, 2018, 06:49:10 PM
Quote
The results also showed no sign of crosstalk etc.

What results? I've never heard such a phrase.

If you have other lines on the cabinet with FTTC then you have crosstalk.

A single bad crosstalker can easily knock 10-15Mb off your line the instant their service goes live and the modem syncs.

I'm not aware of any test that shows crosstalk.
Title: Re: Draytek Vs Billion
Post by: ejs on December 30, 2018, 07:10:26 PM
The GEA FTTC service test results contain a line labelled "Crosstalk", with a result usually of "Not detected", but very rarely saying "Detected". This can lead to all sorts of chaos on the Plusnet forums, when people insist it's not crosstalk because the test says it's not, so there must be some fault that's caused them to lose a few Mbps, so lets have another 5 futile Openreach visits trying to get a non-problem fixed.

I think the test saying crosstalk detected is supposed to indicate an excessive level of crosstalk or something like that.
Title: Re: Draytek Vs Billion
Post by: Alucidnation on December 30, 2018, 07:38:21 PM
That's fair enough.

I knew there had to be crosstalk of some form as it has been an issue for a long time.

If i had lost a few Mbps i wouldn't care, but when i have lost around 20 overnight and i'm paying for it, i DO care.
Title: Re: Draytek Vs Billion
Post by: navzptc on January 01, 2019, 02:49:47 AM
I am on an ECI cabinet running a Draytek 2862N with vdsl set to -3db on sync which has been 100% for the past 9 months - shame my cap hasn't been removed yet by BT :(

Can't recommend it enough as very pleased with it - never gives me a problem and very stable.

Stats below - (8 days uptime) - Just wish BT would release cap - been like this for nearly 1.5 yrs + :(
Title: Re: Draytek Vs Billion
Post by: Alucidnation on January 02, 2019, 01:54:49 PM
Have been doing a bit more digging, and i was struggling to find any reviews at all about the Billion.

I have discovered that they are not (as of November 2018) not on the Openreach MCT approved list.

Having said that, i thought sod it and took the plunge and bought an 8900x R3 from Amazon so if it isn't any good i'll send it back.


 ;D
Title: Re: Draytek Vs Billion
Post by: j0hn on January 02, 2019, 02:19:00 PM
I am on an ECI cabinet running a Draytek 2862N with vdsl set to -3db on sync which has been 100% for the past 9 months - shame my cap hasn't been removed yet by BT :(

Can't recommend it enough as very pleased with it - never gives me a problem and very stable.

Stats below - (8 days uptime) - Just wish BT would release cap - been like this for nearly 1.5 yrs + :(

Can I ask (What seems to me anyway) an obvious question?

Why would you set the snr offset to -3dB if the line is banded?
If it's running at a current SNRM of 6dB (because of the cap) with a target of 3dB, you gain nothing.

If anything would it not make the line slightly more unstable as it causes bits/tones to be used that wouldn't be with a 6dB target SNRM?

DLM would be more likely to remove banding with increased stability.
Title: Re: Draytek Vs Billion
Post by: navzptc on January 03, 2019, 12:45:41 AM

Why would you set the snr offset to -3dB if the line is banded?
If it's running at a current SNRM of 6dB (because of the cap) with a target of 3dB, you gain nothing.


Well I had my system running for about 10 months at 6db and nothing changed, so hoping if it saw a higher 'max rate' that it might, but still nothing changed!!
Title: Re: Draytek Vs Billion
Post by: j0hn on January 03, 2019, 03:46:54 AM
Setting it to +3dB (so 9dB target) is more likely to remove banding.

Who is the ISP? Many will request DLM resets now.
Title: Re: Draytek Vs Billion
Post by: navzptc on January 04, 2019, 12:56:32 AM
Setting it to +3dB (so 9dB target) is more likely to remove banding.

Who is the ISP? Many will request DLM resets now.

BT - Tried a couple of times on their forum to see about getting DLM reset and also rang them up with no joy (said they couldn't request a DLM reset!).

Might just change to another supplier in June when contract is due  >:(
Title: Re: Draytek Vs Billion
Post by: j0hn on January 04, 2019, 12:37:51 PM
They couldn't before but now they can.
Title: Re: Draytek Vs Billion
Post by: underzone on January 04, 2019, 07:25:51 PM
"BT - Tried a couple of times on their forum"

That forum is full of complete kn0bs - post whores quoting nonsense as fast as they can in the hope of being 'thanked'.
Title: Re: Draytek Vs Billion
Post by: Alucidnation on January 07, 2019, 08:25:41 PM
Well, i have had openreach out 5 times now to sort my issues.

They have swapped the pairs numerous times, remade joints and nothing has made any difference whatsoever.

The ISP put in a request for a 'lift & shift', however, the engineer said he was reluctant to do this as he doesn't know what each port is rated at and i could end up on a slower one, as I am on a 105.

But how did he know mine was a 105?

He explained that although the contract is capped at 80, the ports are rated from 90-130 Mbps to aid pushing the speed further up the line.

He also said that if he puts in a request to do this anyway, if i did end up on a lower card, it could not be changed back for some reason.



Title: Re: Draytek Vs Billion
Post by: j0hn on January 07, 2019, 08:43:01 PM
He can sync to your port at the cabinet.
The attainable at the cabinet can be up near 140Mb.

Crosstalk is also caused by the copper tie cables that link the PCP to the FTTC cabinet so each port can show different rates when tested at the cabinet.

However I don't think he can just test the free ports and find you the best.

What he said is technically correct but if the ISP ordered it and you wanted it then he really should have.
It's absolutely true to say that if the line was worse but within the estimates you would be stuck with it.

I would have asked him to proceed with the lift and shift.
At 18dB attenuation and the quoted distance your line could do better.
It could also do worse though.
Title: Re: Draytek Vs Billion
Post by: Alucidnation on January 07, 2019, 08:54:43 PM
Thanks John.

Must admit i was on the verge of asking him to do it anyway, but there is always that chance (and knowing my luck) i would end up slower.

Having said that, i am only just inside the estimates so if it made things worse, surely they (ISP?) would be obliged to put in another request for me to be moved back?

To be fair, my ISP have been brilliant with trying to sort it out and have admitted there is a problem, as one agent sent me a graph of huge error spikes.

This is where i am at the moment.

My estimated  is 69 ( between 51 & 73).

They also have my current line speed at 53


ATU-R Information
           Type:   VDSL2
           Hardware:   Annex A
           Firmware:   07-07-09-05-01-07
           Power Mngt Mode:   DSL_G997_PMS_L0
           Line State:   SHOWTIME
           Running Mode:   17A
           Vendor ID:   00000000 00000000
ATU-C Information
          Vendor ID:   b5004946 544eb206 [IFTN]
Line Statistics
        
Downstream               Upstream               
Actual Rate   52145   Kbps   15549   Kbps
Attainable Rate   61007   Kbps   15646   Kbps
Path Mode   Interleave   Fast
Interleave Depth   759   1
Actual PSD   6. 7   dB   6. 9   dB
Near End                    Far End                   
Trellis   ON   ON
Bitswap   ON   ON
ReTx   0       0   
SNR Margin   6   dB   6   dB
Attenuation   19   dB   24   dB
CRC   0   171
FECS   50308848   s   10550   s
ES   0   s   47   s
SES   0   s   0   s
LOSS   0   s   284   s
UAS   87   s   24022   s
HEC Errors   0   0
RS Corrections   0   0
LOS Failure   0   0
LOF Failure   2   0
LPR Failure   0   0
NCD Failure   0   0
LCD Failure   0   0
NFEC   255   85
RFEC   16   16
LYSMB   4170   16136