Kitz Forum
Broadband Related => Broadband Hardware => Topic started by: GigabitEthernet on September 21, 2020, 06:53:13 PM
-
I have the X6 firmware from here: https://github.com/johnson442/custom-zyxel-firmware/releases. I have this because in the past I had issues with the modem resyncing at half speed, due to an incompatibility with vectoring (which I have on my line) and the modem.
Unfortunately the modem is resyncing at half speed still.
====================================================================================
VDSL Training Status: Showtime
Mode: VDSL2 Annex B
VDSL Profile: Profile 17a
G.Vector: Enable
Traffic Type: PTM Mode
Link Uptime: 0 day: 9 hours: 6 minutes
====================================================================================
VDSL Port Details Upstream Downstream
Line Rate: 5.968 Mbps 18.645 Mbps
Actual Net Data Rate: 5.941 Mbps 18.646 Mbps
Trellis Coding: ON ON
SNR Margin: 9.9 dB 26.9 dB
Actual Delay: 0 ms 0 ms
Transmit Power: - 1.7 dBm 12.1 dBm
Receive Power: -22.2 dBm -16.9 dBm
Actual INP: 0.0 symbols 49.0 symbols
Total Attenuation: 0.0 dB 24.4 dB
Attainable Net Data Rate: 5.978 Mbps 56.707 Mbps
====================================================================================
VDSL Band Status U0 U1 U2 U3 U4 D1 D2 D3
Line Attenuation(dB): 9.6 40.5 62.0 N/A N/A 19.2 49.0 85.9
Signal Attenuation(dB): 9.6 39.9 N/A N/A N/A 27.6 48.6 N/A
SNR Margin(dB): 11.5 9.7 N/A N/A N/A 26.9 27.0 N/A
TX Power(dBm): -12.7 -2.1 N/A N/A N/A 10.6 6.6 N/A
====================================================================================
VDSL Counters
Downstream Upstream
Since Link time = 6 min 43 sec
FEC: 3727 356
CRC: 0 15
ES: 0 11
SES: 0 0
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
Latest 15 minutes time = 6 min 9 sec
FEC: 5 0
CRC: 0 0
ES: 0 0
SES: 0 0
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
Previous 15 minutes time = 15 min 0 sec
FEC: 0 0
CRC: 0 0
ES: 0 0
SES: 0 0
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
Latest 1 day time = 7 hours 6 min 9 sec
FEC: 120 34
CRC: 0 15
ES: 0 11
SES: 0 0
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
Previous 1 day time = 24 hours 0 sec
FEC: 11111541 322
CRC: 1384 68
ES: 30 36
SES: 24 4
UAS: 166 151
LOS: 1 0
LOF: 0 0
LOM: 10 0
Total time = 1 days 7 hours 6 min 9 sec
FEC: 160232548 356
CRC: 4823 427
ES: 116 258
SES: 83 4
UAS: 730 668
LOS: 4 0
LOF: 4 0
LOM: 45 0
====================================================================================
-
I wasn't aware that the ZyXEL VMG1312-B10A was vectoring capable. <Puzzled.> :-\
-
I wasn't aware that the ZyXEL VMG1312-B10A was vectoring capable. <Puzzled.> :-\
Well what modem would you advise?
I understood with the X6 firmware it worked fine. But if not, I am happy to buy something else, I only need the modem.
-
I'm not sure what to suggest. Sorry.
Seeing the line "G.Vector: Enable" in the data you included, above, resulted in some head-scratching. After a quick search on the ZyXEL website, the VMG1312-B10A (https://www.zyxel.com/uk/en/products_services/Wireless-N-VDSL2-4-port-Gateway-with-USB-VMG1312-B-Series/) page confirms that it is vectoring capable.
I've just learnt something.
-
I understood with the X6 firmware it worked fine.
Hard to say. I believe some people on this forum have had success with version x6, but the number that have tried it is tiny (1 or 2 tops) since vectoring is very rare. I remember reading a comment in another thread suggesting that a regression had occurred and that the older versions are better. There have certainly been a number of changes made regarding vectoring between various versions. These changes are listed as G.993.5 in this document https://www.cisco.com/web/software/282821780/140978/ReleaseNote_A2pv6F039x3.pdf (https://www.cisco.com/web/software/282821780/140978/ReleaseNote_A2pv6F039x3.pdf) for a subset of firmwares.
If you want to try an older version of the DSL firmware without having to rebuild the whole firmware image, the A2pv6F039o driver is in the official Zyxel AAJZ.6 build, and the A2pv6F039o1 driver is in the AAJZ.7 build. I would be inclined to test the A2pv6F039o version, since this is the version which passed Openreach MCT approval (perhaps they tested vectoring?). Other DSL firmware versions which have passed MCT testing are A2pv6F039f1, A2pv6F039r and A2pv6F039t but a custom firmware build would have to be made to test these. Also, Sky seem to like to stick with the A2pv6F039m1 version in their BCM63168 chipset equipped ER115 for some reason, so that could be another one to try.
Edit: Openreach do indeed test vectoring as part of the MCT tests. See p.27 of this https://www.openreach.co.uk/orpg/home/helpandsupport/sins/sins/downloads/SIN498v7p7.pdf (https://www.openreach.co.uk/orpg/home/helpandsupport/sins/sins/downloads/SIN498v7p7.pdf). I'd therefore suggest it is well worth trying one of the approved firmwares.
Edit 2: This thread is well worth a read https://forum.kitz.co.uk/index.php/topic,22306.0.html (https://forum.kitz.co.uk/index.php/topic,22306.0.html).
-
Sorry for the late reply.
I would really like to get the 1508 MTU working but I also need to use the earlier firmwares as reported in the link above. They seem to have no issues on my line.
Can anyone help with this?
-
Just to update this thread, I belatedly spotted GigabitEthernet's request for a custom firmware which has baby jumbo support together with an older DSL firmware, since neither the default A2pv6F039v or newer A2pv6F039x6 that johnson has builds of were working well with vectoring in this case, however an older official build with the A2pv6F039o1 was fine. I have johnson's brilliant build environment set up on my computer, so offered to build a version of the firmware that uses the A2pv6F039o1 firmware.
However, I thought it might be interesting to try a different firmware, since johnson remarked in this post https://forum.kitz.co.uk/index.php/topic,22306.msg381599.html#msg381599 (https://forum.kitz.co.uk/index.php/topic,22306.msg381599.html#msg381599) that while A2pv6F039o1 solved the sync speed issues for him, it was less stable. Since kitz pointed out (https://forum.kitz.co.uk/index.php/topic,22306.msg381566.html#msg381566 (https://forum.kitz.co.uk/index.php/topic,22306.msg381566.html#msg381566)) that the Openreach MCT approval tests the modem's behaviour when vectoring is used, I created a list of all the approved DSL firmwares for the BCM63168 chipset based on the MCT list here https://freeola.com/support/openreach-approved-vdsl-modems-router-mct.html (https://freeola.com/support/openreach-approved-vdsl-modems-router-mct.html) for various modems. This list is
- A2pv6F039o1.d26a
- A2pv6F039r.d26d
- A2pv6F039r.d26e
- A2pv6F039t.d26i
- A2pv6F039t.d26k
- A2pv6F039v2.d26m
The last one, A2pv6F039v2, caught my eye since it is the newest, is closest to the version in the official latest Zyxel release and also because it does not appear in the changelog seen here https://www.cisco.com/web/software/282821780/140978/ReleaseNote_A2pv6F039x3.pdf (https://www.cisco.com/web/software/282821780/140978/ReleaseNote_A2pv6F039x3.pdf) while A2pv6F039v and A2pv6F039w either side of it do. This led me to suspect that perhaps A2pv6F039v2 is a special release made for the Technicolor TG589vac V1/V2 and TG789vac V1/V2 in order for those modems to pass MCT (which they have done with this firmware), maybe having been modified to resolve the vectoring issues.
I therefore built a version of the VMG1312-B10A 1.00(AAJZ.17)C0 firmware which had johnson's baby jumbo frame patch using the A2pv6F039v2 firmware extracted from the image for a Technicolor TG789vac V2. GigabitEthernet has now tested it out and found that it is working well. If anyone else with vectoring is having difficulties with the halved sync speed issue, perhaps A2pv6F039v2 would be worth a go in your case too.
-
Something interesting to add with @siofjofj's firmware.
With every firmware I've used previously on ZyXEL devices, after a resync I've always seen this odd thing with the SNRM.
It will start at around 2dB lower and then slowly rise up over a few minute period. It happened every time without fail, which made it hard to work out if my target was actually say 4dB or actually 6dB.
I can only conclude this must be something to do with vectoring.
I didn't get this happening on the HG612 - and with the new firmware it doesn't happen either.
-
I wonder if it is a good alternative for ADSL, if it contains any other improvements?
-
I wonder if it is a good alternative for ADSL, if it contains any other improvements?
I can't find any publicly accessible changelogs for the A2pv6F039v2 firmware, so can't confirm definitively whether or not any changes related to ADSL were made between it and A2pv6F039v. The ADSL related items in the changelog we do have access to are: 'Fix US CRC's against Calix C7 ADSL2-24A in G.992.[35]' and 'Fix long training time in G.992.1 against AD72 DSLAM' for the A2pv6F039w firmware, and 'Improve kl0 estimate in G.992.[35]' for the A2pv6F039x1 firmware.
What I can say though, is that I have tried the A2pv6F039x6 and A2pv6F039v2 (mainly so that I could check that this firmware wasn't something weird that only worked on Technicolor modems before giving it to GigabitEthernet) firmwares on my VMG8924-B10A on my Sky ADSL2+ line, and noticed no difference for either of them.
I'm happy to build you a A2pv6F039v2 firmware with the stats server and baby jumbo frames (anything else?) Weaver if you want to try it out (subject to the proviso that I don't own a VMG1312-B10A so can't test it, so flashing it would be at your own risk), however I might not be able to upload it until my connection is fixed as thanks to the DLM I now only have 600kbps upload and it keeps disconnecting necessitating starting the upload again.
-
Yes please if you wouldn’t mind.
-
Yes please if you wouldn’t mind.
You are in the fortunate situation that one device could be updated with the newly built firmware image whilst leaving the other three on the older image, as control.
-
I wouldn’t do that. I have (literally) a dozen spares in stock, Janet tells me.
-
Yes please if you wouldn’t mind.
As requested, I've built a firmware based on the Zyxel VMG1312-B10A_1.00(AAJZ.17)C0 image but using the A2pv6F039v2 DSL firmware. I included all of johnson's patches (jumbo frames, multiple telnet sessions, stats server and custom commands) since I assume the firmware you currently use also includes all these features.
As it is completely untested since I don't have a VMG1312-B10A, I'd rather not post the download link publicly until you can verify that it at least doesn't brick your modem. Could I send it to you by PM instead (I'm currently blocked from doing this)?
-
siofjofj I have a VMG1312-B10A, just flashed the 1312-B10A-17-jumbo-v2.bin file however noticed there is no stats on it aka no mongoose server. Not sure if thats ment to be though. However will test the current firmware to see if any issues occur wih the 1312-B10A-17-jumbo-v2.bin firmware.
It boots and connects to the internet which is a start.
-
siofjofj I have a VMG1312-B10A, just flashed the 1312-B10A-17-jumbo-v2.bin file however noticed there is no stats on it aka no mongoose server. Not sure if thats ment to be though. However will test the current firmware to see if any issues occur wih the 1312-B10A-17-jumbo-v2.bin firmware.
GigabitEthernet didn't want the stats server so I didn't include it in that build, however I did a different build for Weaver which does include it. I haven't posted the link to the one I built for Weaver publicly (I just emailed it to him) as I can't test it myself and don't want large numbers of people bricking their routers if it turns out to be a dud (the other one has been tested by GigabitEthernet so I know it's OK). That said, if you're happy to take the risk I'm happy to PM you the link too.
-
Ah, good to know. I wanted all the side order of ‘fries’ and onion rings and everything.
-
@siofjofj , I understand the risk, nothing my SPI programmer can't sort if it goes south. Not only that might give me an excuse to get a new modem as well but I'll have to make up my mind as I have two in mind one being content 3030/3031u or the zyxtel VMG4005-B50B.
@weaver you a dam legend mate loved your reply.
-
. . . I'll have to make up my mind as I have two in mind one being content 3030/3031u or the zyxtel VMG4005-B50B.
Hmm . . . Comtrend VR-3030/VR-3031U or ZyXEL VMG4005-B50B, possibly? It looks as if the autocorrect has had a "funny-fit".
-
Hmm . . . Comtrend VR-3030/VR-3031U or ZyXEL VMG4005-B50B, possibly? It looks as if the autocorrect has had a "funny-fit".
Yes my phone's auto correct loves to troll me all the time yes I ment comtrend.
-
Yes my phone's auto correct loves to troll me all the time yes I ment comtrend.
But no harm has been done. :)
-
Just a quick update. The firmware I originally built for Weaver has now been tested to boot by hacktrix2006, so I'll now make the download link public: The VMG1312-B10A 1.00(AAJZ.17)C0 firmware with the A2pv6F039v2 DSL firmware and all of johnson's modifications (stats server, jumbo frames, multiple telnet sessions and custom startup commands) is available here https://www.dropbox.com/s/pdd7htdu16eudp8/1312-B10A-17-jumbo-tel-v2-stats1.1-cmd.bin?dl=0.
-
Thanks to hacktrix2006 as I still haven’t been well enough to flash a modem yet. Hitting the painkillers again.
-
Just to add to this, my new VDSL2 connection, for which the BT Wholesale broadband checker reports clean speed estimates 56.8 - 40 Mbps download and 11.0 - 7.7 Mbps upload, has the following stats:
Stats recorded 18 Dec 2020 12:09:53
DSLAM type / SW version: BDCM:0xc190 (193.144) / v0xc190
Modem/router firmware: AnnexA version - A2pv6F039v2.d26a
DSL mode: VDSL2 Profile 17a
Status: Showtime
Uptime: 15 hours 33 min 23 sec
Resyncs: 0 (since 18 Dec 2020 12:09:49)
Downstream Upstream
Line attenuation (dB): 21.3 0.0
Signal attenuation (dB): Not available on VDSL2
Connection speed (kbps): 40000 9997
SNR margin (dB): 13.4 6.7
Power (dBm): 11.9 7.4
Interleave depth: 16 1
INP: 45.00 0
G.INP: Enabled Not enabled
Vectoring status: 1 (VECT_FULL)
RSCorr/RS (%): 0.0001 0.0025
RSUnCorr/RS (%): 0.0000 0.0000
ES/hour: 0 0
As you can see, I have vectoring and am using the A2pv6F039v2 DSL firmware on my Zyxel VMG8924-B10A. I have none of the sync speed halving (if this was happening I'd expect to see perhaps 30 Mbps on my 40 Mbps capped connection (I only subscribe to the 40/10 product) as the maximum predicted speed is 56.8 Mbps and my reported max attainable speed is 60.5 Mbps) or other strange behaviour that is observed for the A2pv6F039v firmware. This means I now join GigabitEthernet in having success with the Openreach approved v2 DSL firmware on a vectoring enabled circuit.
Also of note is that I have INP = 45 and INPRein = 0, suggesting RetX low unlike GigabitEthernet here https://forum.kitz.co.uk/index.php/topic,22086.msg426585.html#msg426585 (https://forum.kitz.co.uk/index.php/topic,22086.msg426585.html#msg426585) who can't get rid of RetX high (though this may be because of my 40/10 cap).
-
Looks like you were provisioned on a 40/10 connection by the ISP.
-
Looks like you were provisioned on a 40/10 connection by the ISP.
Indeed, although this was by choice as I only ordered the 40/10 product since my previous ADSL2+ 19/1.2 connection was perfectly adequate for my needs in my old house. I chose VDSL in my new place as the telephone exchange is in the next village, so I'd only get 6/0.8 on ADSL2.
-
Just to add to this, my new VDSL2 connection, for which the BT Wholesale broadband checker reports clean speed estimates 56.8 - 40 Mbps download and 11.0 - 7.7 Mbps upload, has the following stats:
Stats recorded 18 Dec 2020 12:09:53
DSLAM type / SW version: BDCM:0xc190 (193.144) / v0xc190
Modem/router firmware: AnnexA version - A2pv6F039v2.d26a
DSL mode: VDSL2 Profile 17a
Status: Showtime
Uptime: 15 hours 33 min 23 sec
Resyncs: 0 (since 18 Dec 2020 12:09:49)
Downstream Upstream
Line attenuation (dB): 21.3 0.0
Signal attenuation (dB): Not available on VDSL2
Connection speed (kbps): 40000 9997
SNR margin (dB): 13.4 6.7
Power (dBm): 11.9 7.4
Interleave depth: 16 1
INP: 45.00 0
G.INP: Enabled Not enabled
Vectoring status: 1 (VECT_FULL)
RSCorr/RS (%): 0.0001 0.0025
RSUnCorr/RS (%): 0.0000 0.0000
ES/hour: 0 0
As you can see, I have vectoring and am using the A2pv6F039v2 DSL firmware on my Zyxel VMG8924-B10A. I have none of the sync speed halving (if this was happening I'd expect to see perhaps 30 Mbps on my 40 Mbps capped connection (I only subscribe to the 40/10 product) as the maximum predicted speed is 56.8 Mbps and my reported max attainable speed is 60.5 Mbps) or other strange behaviour that is observed for the A2pv6F039v firmware. This means I now join GigabitEthernet in having success with the Openreach approved v2 DSL firmware on a vectoring enabled circuit.
Also of note is that I have INP = 45 and INPRein = 0, suggesting RetX low unlike GigabitEthernet here https://forum.kitz.co.uk/index.php/topic,22086.msg426585.html#msg426585 (https://forum.kitz.co.uk/index.php/topic,22086.msg426585.html#msg426585) who can't get rid of RetX high (though this may be because of my 40/10 cap).
What are your FECs looking like? I am convinced that is the cause of why I can't shake it off here.
-
What are your FECs looking like? I am convinced that is the cause of why I can't shake it off here.
My line only got activated 20 hours ago, but in that time I have had 1853 downstream FECs and 120 upstream FECs.
-
My line only got activated 20 hours ago, but in that time I have had 1853 downstream FECs and 120 upstream FECs.
Yeah you'll be fine then, I get millions a day.
-
Yeah you'll be fine then, I get millions a day.
Interesting, I wonder why. Did your attempts at sync capping in the other thread noticeably reduce the FEC count?
Something I have noticed though is I get a slight drop (by 1.0 dB) in downstream SNRM and one single upstream CRC error occurs each time I pick up the phone, then the SNRM returns to normal and a second CRC occurs upon hanging up. This is not causing me any issues so I will not be doing anything about it, but I'm just wondering if this is normal on VDSL2? I'm using a known good (on my old ADSL2+ line anyway) VDSL Mk3 faceplate for filtering. Perhaps if I ever upgrade to the 80/20 product and end up running at 3 dB target SNRM it might cause a problem as I'd then be dropping from 3 dB to 2 dB instead of 13.4 dB to 12.4 dB?
-
You would probably get high 60's with a 3dB target, possibly just over 70Mb if the line made good use of the D3 band.
As more customers get added to the new VDSL Sidepod (which i don't think will be Vectored) you could see your downstream sync crash.
-
Interesting, I wonder why. Did your attempts at sync capping in the other thread noticeably reduce the FEC count?
Something I have noticed though is I get a slight drop (by 1.0 dB) in downstream SNRM and one single upstream CRC error occurs each time I pick up the phone, then the SNRM returns to normal and a second CRC occurs upon hanging up. This is not causing me any issues so I will not be doing anything about it, but I'm just wondering if this is normal on VDSL2? I'm using a known good (on my old ADSL2+ line anyway) VDSL Mk3 faceplate for filtering. Perhaps if I ever upgrade to the 80/20 product and end up running at 3 dB target SNRM it might cause a problem as I'd then be dropping from 3 dB to 2 dB instead of 13.4 dB to 12.4 dB?
Yeah if you look at that thread I went from something like 20,000+ FECs an hour to about 10,000. Sorry I slightly over-exaggerated with the millions, although it doesn't take many days to get into the millions.
It's probably just a noisy line to be honest, I can't really be bothered to go into the reasons why, it's always been slightly baffling why I sync quite high on the downstream side of things for my line length (750m) but a bit low on the upstream side.
Your line is a bit shorter than mine but you'll see you have a higher upstream sync but a lower potential downstream. Doesn't make sense to me!
I wonder if 10,000 FECs per hour is the magic number and I would simply need to cap my DS sync more to see a benefit, might try that if I get bored.
Regarding the SNRM drop, it shouldn't do that. Sounds like the start of an HR fault to me.
-
Ignore me, I can't read. Your line is shorter and so you see a higher expected DS and US as would be expected.
Curious to see more stats over time from another vectored line though.
You don't see your SNRM rise immediately after a resync then? It was an odd thing I always saw with certain modems, sync around 6dB and it would slowly rise over a few minutes to 7dB or more, almost like it took time for vectoring to activate.
-
Regarding the SNRM drop, it shouldn't do that. Sounds like the start of an HR fault to me.
I was worried you might say that, before my service was activated I was able to do the quiet line test on 17070 and did hear crackling. I won't be doing anything about it for now since it's not actually causing me any issues and since trying to get a HR fault fixed which made my ADSL2+ service completely unusable took 3 years and 12 Openreach engineer visits. Perhaps this, together with
As more customers get added to the new VDSL Sidepod (which i don't think will be Vectored) you could see your downstream sync crash.
means it is a good idea for me to stick with the 40/10 product!
I was half expecting my new order to be put on the new sidepod as my house has been empty for a good while and codelook reports 557 premises passed for my PCP and I imagine a large number of people choose VDSL due to the low ADSL speeds and lack of Virgin media or 4G signal, so I'd have thought the main VDSL cab (perhaps upgraded to 512 lines) is rather full. Perhaps my line was left jumpered in after the previous resident moved out.
Curious to see more stats over time from another vectored line though.
No problem, happy to post some in a few weeks time.
You don't see your SNRM rise immediately after a resync then?
No I don't, but didn't you find that the v2 DSL firmware stopped this behaviour? I've been on this firmware since day 1 since it was the last thing I flashed to my VMG8924-B10A when I was playing with building firmwares.
-
Well I'm using a BT Smart Hub now but don't have any issues on that either.
Actually on the upstream side it syncs very well.
-
Regarding the SNRM drop, it shouldn't do that. Sounds like the start of an HR fault to me.
Agreed. A joint is starting to show either HR or semi-conductive tendencies.
-
Actually on the upstream side it syncs very well.
Just noticed on your other thread the trough in your D0 band. I have that too. Is this a usual thing for VDSL2 (DPBO for the benefit of ADSL lines perhaps?), or perhaps something seen on vectored lines?
Agreed. A joint is starting to show either HR or semi-conductive tendencies.
How annoying, I thought I'd have left those issues behind at my old address! Hopefully it won't get worse to the point where it becomes service effecting.
-
Just noticed on your other thread the trough in your D0 band. I have that too. Is this a usual thing for VDSL2 (DPBO for the benefit of ADSL lines perhaps?), . . .
Correct, it is the DSPBO to allow for exchange based xDSL services.
It is actually in the DS1 band. Only the US bands are numbered started from zero.