Kitz Forum
Broadband Related => FTTC and FTTP Issues => Topic started by: dfects on September 23, 2016, 09:48:33 AM
-
Hiya,
We were finally able to order FTTC after being on adsl forever. We can only get Infinity 1 due to being about 1km in cable length from our cab and BT only offered a self install. The "smart" hub came with an adsl dongle so I replaced my xf-1e with this one thinking it may actually make a difference.
We've been having a few issues... each night it was resyncing and the sync speed on the down stream went from 36 to 37, 38 then 40 for two nights. The next night it dropped to 35 then to 32.4 :( The noise margin has also increased from 6.3 on initial sync to 9.1 and now 8.8 so I'm trying to find out why.
Anyway, two questions:
- Can I use my XF-1E with fttc?
- Could the bt filter be causing more interference which is affecting my sync speed/noise margin?
-
As you have only just had VDSL installed, it will still be in its training period. This means that you will likely see many resyncs as DLM works out the best place for your line. The increased SNRM implies that it has intervened to deal with a high error rate, probably with interleaving + RS error correction.
On a line with no faults, this shouldn't really have happened, so I would suggest that you try to get your hands on a DSLStats compatible modem or modem/router such as a HG612, or a ZyXEL VMG series (1312, and 8X24 series are good choices). This will allow us to see detailed information about your connection and work out if there is any cause for concern.
The training period is 10 days to my knowledge, so if you are still concerned it may be worth ringing BT and asking them to perform a GEA test to detect any faults from their end. Be careful though, as some faults, such as a rectified loop, are more often than not internal wiring issues, and if an engineer comes out and discovers this, there is a hefty fee, sometime up to £200.
Another thing you could do is go to this site and enter your phone number: https://www.btwholesale.com/includes/adsl/adsl.htm?s_cid=ws_furls_adslchecker (https://www.btwholesale.com/includes/adsl/adsl.htm?s_cid=ws_furls_adslchecker)
This will allow you to find out which cabinet you are on, and your estimated speeds. You can then enter your phone number on this website: https://www.telecom-tariffs.co.uk/codelook.htm (https://www.telecom-tariffs.co.uk/codelook.htm)
This site will return a result showing your exchange. If you click on this, then click on 'List all x fibre cabinets', then find your cabinet in the list to show an approximate location and brand. The brand is the important thing in this case, as it will indicate whether or not you have access to a technology called G.Inp, which can be used as an alternative to interleaving + RS error correction on more stable lines.
On another note, what is your current ping? A high ping can indicate the use of interleaving on a line, so a ping higher than around 20ms can sometime be an indicator.
-
Considering you are using a plug-in filter rather than a filtered faceplate, the most important aspect is whether or not you have any other telephone sockets, even with nothing plugged into them. Unfiltered extension wiring acts as a "bridged tap", the length of internal wiring may not have mattered much for ADSL frequencies, but it's much worse for VDSL2 frequencies.
On the general subject of filter quality, see MR-261 (https://www.broadband-forum.org/marketing/download/mktgdocs/MR-261.pdf).
-
Thanks for the replies. From my understanding from research, I thought on vdsl provisioned lines only the first 2-3 days were considered the training period of sorts? There is a lot of confusing information on this.
Its a new build house, only 3-4 years old. There are extensions everywhere, but when we had our phone activated I asked the OR to disconnect all the extensions and just leave the master socket connected which he did. The grey box is literally the other side of the wall from the faceplate, so the cable length shouldn't be more than maybe 50 cm between them.
On speed tests and often in games my ping is still around the 18-20ms mark even though we're currently banded to 32.4 mb. The upstream is still on 6.2 noise margin and syncing 9.68, which is about 2mb above the speed estimate!?
I did have a few bits of equipment close to the home hub/faceplate including an Intel Nuc pc, a raspberry pi, two external hdds and an old wireless (probably not dect) cordless phone. I've since removed all these in case its causing interference? Not sure how likely this is. My DSL has stayed connected since (nearly 2 days) with a margin of 8.8 but not sure if this is due to the lower speed banding or the interference being removed.
I've tried a quiet line test, but only into the adsl filter as I don't want to risk killing the DSL sync again in case DLM can improve our line. There is a very very faint constant hum on the line but no crackling, pops, etc. I'm not sure if the hum is the cheap corded phones speaker/microphone or an actual issue, but its very quiet.
The cabinet we're connected to (only became active maybe 3-4 weeks ago, we're likely to be the first on it) is listed as Huawei.
Does DLM slowly recover the noise margin? Or if the interference was removed would the noise margin drop straight away on the next update? I can see every day at 8.54am (no idea why) it gets a transmission on TR69 and has adjusted the noise margin then (from 9.1 to 8.8).
-
For a G.993.2 (VDSL2) circuit provided by Openreach there is no training period.
Assuming that the circuit was active with a state of synchronism between the modem and the DSLAM on the day of provision, then the DLM will begin monitoring (and adjusting circuit the parameters, if it considers such adjustment is necessary) from the second day following the circuit's provision. The DLM will then continue to monitor (and adjust) the circuit from that day onwards.
-
Thanks for the reply :) That makes sense! The strange thing is after the first 3-4 days our speed actually increased for 2 days to 40mb which it stayed synced at for 24 hours each time? Then after those two days it dropped to 35 then to 32.4 each night.
The only thing that changed is I checked the hub with a contactless (infrared) thermometer as it has restricted airflow where it is and in doing so, I noticed the adsl line wasn't fully inserted into the hub and pushed it in until it the clip clicked into place!? Surely inserting the cable properly didn't cause extra interference! It didn't desync when I did this it stayed connected.
Really hoping DLM drops the noise margin and removes the banding in the next few days after removing all the sources of interference I can but will monitor. Will look at getting a modem that gives actual stats too.
-
If the DLM decides to manipulate a circuit's parameters, it will usually do so once per 24 hour period. However if there is severe disturbance/disruption to the circuit then there may be multiple DLM intervention events within a 24 hour period, in an attempt to attain stability.
The overall aim of the DLM is to try to keep a circuit operating in a stable fashion. Whilst it is adjusting the circuit parameters the synchronisation speed may fall or rise . . .
-
Do you know if DLM will slowly lower the noise margin if its going to? Or can it assess the noise on the line and make large adjustments straight away? I noticed when it increased it went to 7.9, 8.8 then 9.1 so hoping it steps down the noise margin over the next few days/weeks and maybe readjusts the banding. Feel in an awkward situation now where I don't know if the problems resolved or whether there is actually a line issue.
-
Do you know if DLM will slowly lower the noise margin if its going to? Or can it assess the noise on the line and make large adjustments straight away?
The DLM does not directly adjust the SNRM. There is a target SNRM defined for the circuit, which is 6 dB. (Though there has been some testing of a 3 dB target SNRM.) Other influences will cause the instantaneous, moment by moment, SNRM to vary from the defined target. The difference between the average, instantaneous, SNRM and the target SNRM should be regarded as a "buffer zone" which arises as a result of those other adverse influences upon the circuit.
I noticed when it increased it went to 7.9, 8.8 then 9.1 so hoping it steps down the noise margin over the next few days/weeks
That could happen. Or there could be a reason to induce a resynchronisation event, in which case a "step change" in the SNRM may be observed. It is important that all of the circuit parameters are considered and not to focus on just one of them.
and maybe readjusts the banding.
Has it been conclusively determined that your circuit is banded? :-\
-
do you have access to a modem that gives true connection stats? like the white openreach Huawei HG612? I'm pretty sure your connection isn't banded.
DSL uptime:
1 Days, 5 Hours 15 Minutes 10 Seconds
Data rate:
9.68 kbps / 32.40 kbps
Maximum data rate:
9678 / 39063
Noise margin:
6.1 / 8.8
Line attenuation:
25.1
Signal attenuation:
VPI / VCI:
0/38
Modulation:
G_993_2_ANNEX_B
Latency type:
Fast Path
I have no idea if "Maximum data rate" is the current sync, or max attainable. Either way neither of those figures are banded figures.
A banded circuit will sync at the exact same figure every time. It may not be possible to conclusively tell with notsosmart hub
-
You're right, I don't know for certain. Its based mainly on the speed dropping from speeds like 40.78mb, 40.48mb to exactly 35.00mb then exactly 32.4mb, both of which seem to be banding limits on https://community.plus.net/t5/Library/FTTC-DLM-What-it-is-How-it-works/ba-p/1322799 .
I don't know either what the exact stats shown on the smart hub are, other than the "Data Rate" is actually the sync speed as its shown on the "basic" page, it shows in kbps on the tech log page due to a bug but is actually mbps. The Maximum data rate has varied with each sync, but the lowest i've seen is 3750ish:
here are the current stats I can get from both pages:
(https://lh3.googleusercontent.com/MdWsoEhNyM8YjBb6Yl5lyjoG4Al03u3Vy7VYWa17N95yyjLsjAxGzwpXOQsU7i8k9DfjmfXhzRIw-09Mw8XpA0gknU__mzXKRSqPi44y_pO0vTsKeRmQvsd2G03pGgvgkvq2k4R9gb2rcOjKNBOZgldM4E6FXlgOkMeubLSFLcwb_jzf_jD1E0SL4S2SYgXln6Xma1He6hszXDbmiPZY5J5Qc_Ksav9HhpHrUrLJpoGXOmRS6MyGcIKNJdlydCKWhaYKBjixTBDfDAg1EZkzWXqVIgK3KawjDANRtXqO0pt3VjLNcxJDt4-cLSIvXSdQXQmGK9ajk8Tqkpl9fR4yODX4L3zPQ0LXK4JGse7IAL2OYP4toTfJ-Tm-EYPEvlnGMn-IzeX8EFBU8PUHLKOevXUL6uGuFZ7asJpM9q5uRv8Y6vIofaMzmFgWqkPTXCOesghAz5cKBlwoTRZ7qYxSXgGczbGXsqT1YsDUleAh5DN-IsRLuJ72ysw3jq9Ms4W-E5_donG5UGIlkkoipsV_bIa-CphmVNBWYXQ6-7sEHnxkVfDR7gir_zIWVRQ8YsJQ2WKwzOZ3syKC5W6pP3xIFp9KGstB-zbPImv60IUkAwpqXAC58A=w1084-h554-no)
-
ahhh, I see the confusion. The term banding/banded we are referring to is where DLM puts a limit on the sync speed. The connection will never be able to sync above this, until a DLM reset is performed. What you point to on Plusnet appears to be something different that in all honesty I'm not familiar with. Is that not related to ip profiles?
-
ahhh, I see the confusion. The term banding/banded we are referring to is where DLM puts a limit on the sync speed. The connection will never be able to sync above this, until a DLM reset is performed. What you point to on Plusnet appears to be something different that in all honesty I'm not familiar with. Is that not related to ip profiles?
I'm confused, is the "maximum data rate" actually the sync rate and what the hub is calling the sync rate some form of ip profiling based on the noise/other variables? Sorry I'm pretty new to all this. Our ADSL connection was pretty stable for 3 years with a draytek vigor 120 on this line I so didn't touch it and before that I've only dealt with docsis on virgin connections.
-
The "maximum attainable net data rate", labelled as "Maximum data rate" in your screenshots, is your modem's estimate of what the maximum speed might be under the current conditions, it takes into account the target SNRM, but ignores the banding.
The Plusnet library article about FTTC DLM sounds like a reasonable summary of FTTC banding. The bands can set a minimum and maximum speed. The article is about two years old, so the exact bands may have changed since then.
The IP Profile is something completely different. The IP Profile is simply a number, about 96% of the data rate (sync speed), to allow for some of the overheads (such as the PPP layer), and give an idea of the throughput (file download speeds) possible for the data rate shown by the modem.
-
Thanks, that makes it a lot clearer. So the chances are my line is actually banded? Not sure what the best option is here. To wait and see if removing all the electronics from near the hub/faceplate have made a difference and DLM improves our speed over the next few days/weeks or take the hit and disconnect to do a quiet line test.
Will DLM ever actually move me up a band?
-
I guess it's banded. The FTTC DLM seems to apply banding fairly commonly, unlike the ADSL DLM which adjusted the target SNRM first, and only uses banding in the worst cases.
Interleaving+FEC can result in the max attainable rate being moderately higher than the actual rate, but it wouldn't cause the increased SNRM. In a similar way as increasing the target SNRM results in a lower rate, limiting the rate results in a higher SNRM.
The DLM may eventually relax or remove the banding, how long it takes, no-one really knows. But whatever caused the banding in the first place generally needs to no longer be present.
-
Going from ADSL to VDSL2 the internal wiring needs to be almost perfect on longer lines and same for the master socket with a new NTE5 & SSFP MK3 installed it's a pity you never had Openreach to do the FTTC install.
As this is a new FTTC line I would wait a few more days but if your are not happy after that you can call your ISP and ask them to send an OR engineer out it should be free
-
You know, I think I might of missed something. I installed an extra 12v 120mm pc fan in my cabinet that houses that hub and the other electronics I mentioned on Sunday night. I didn't notice until just, but the next time it synced on Monday morning the hub dropped from 40.48mb to exactly 40.00mb. This could possibly be a band? The plusnet doc says so. The next 2 nights it dropped to 35.00mb, then 32.40mb both of which are bands according to that doc too.
The fan was only about 10-15cm from the hub, and maybe 30cm from the faceplate :( Could this of been the cause of all the interference? Its all been removed now, so the only thing close to the hub and faceplate are my asus router and the hub itself.
-
ask your ISP to run a GEA line test. this will show if your line is banded or not.
-
Regarding the gradual decline since getting the service, is there any possibility you were one of the first to be connected after the cabinet went live?
If so then as neighbouring lines get connected up so crosstalk will increase and your speeds will drop. I was pretty much first in my village to get FTTC and initially saw around 50Mbps. Over the next few months as the word spread, and the rest of the village switched over, I progressively lost almost half of that, getting about 30Mbps last time I looked.
-
We probably were the first on there. I knew the cab had been fitted as I drive past it every day so I was checking daily until it became available, so I signed up straight away and took the first slot I could ;D Glad to be free of 6mb adsl.
It was only activated on the 14th of Sept, 10 days ago today. Maybe it is cross talk as I did tell my neighbours as I saw them, at least the ones who mentioned how frustrated they are with the slow net we had.
I know our cab serves our 27 houses, a small industrial estate of maybe 10 units and a couple of older houses. Most the rest in our area are on another cab from my checks. Our adsl cab is a very small rusty old cab a new shiney double cab, and the fibre one fitted is smaller than most in the area too.
I managed to test my corded phone on my nans bt connection today, but her line has constant clicking and hissing so it wasn't much help (luckily shes 85 and can barely hear so she doesn't care). I tried it in my parents virgin phone line by lifting the receiver and pressing one number so it was quiet and it sounded almost identical to on our bt line on the quiet line test... that is a very faint hiss. There may be a tiny bit more of a buzz on our bt line, but its hard to tell as it was tested several hours apart. Either way, its not really told me if its the cheapo phone or the line thats making the noise :(
Our hub had two TR69 (ACS?) connections this morning, one that set the downstream noise margin from 8.8 to 8.6 at 8.28am then one that set it back to 8.8 at the same 8.54am fixed time!? Not sure if that tells me anything at all :(
-
A cheapo old-fashioned wired POTS telephone should be fine for testing, surely. If you can hear hiss, then that’s presumably not good (?)
-
I'm using this http://www.argos.co.uk/static/Product/partNumber/3924358.htm and I have to admit, I'm trying it through the adsl filter as I don't want to disconnect the hub at the moment which i know is less than ideal.
Its hard to describe the noise, its constant, monotone and quiet... as in if there is any noise at all while listening you won't hear it. If its silent, and you concentrate its there. My parents virgin line was pretty much the same with this handset, so not sure if its the low quality speaker or maybe both are crap lines. Are your guys lines honestly dead silent? so you can't hear a difference between the test, and the line being dc'ed as in the receiver was down?
-
@sevenlayermuddle
It seems unlikely to be crosstalk, because crosstalk is basically noise, so if it were crosstalk, the speed would be decreasing while the SNRM stays the same, or the crosstalk would be seen by a drop in the SNRM if the crosstalk appeared and you remained connected. But here we've got decreasing speeds, and a raised (from the usual default of 6 dB) SNRM.
Our hub had two TR69 (ACS?) connections this morning, one that set the downstream noise margin from 8.8 to 8.6 at 8.28am then one that set it back to 8.8 at the same 8.54am fixed time!? Not sure if that tells me anything at all :(
The TR69 remote management won't be setting the SNRM, but it might cause the Hub to reboot or reconnect. The current SNRM will vary slightly over time, this is normal.
-
Hmm it seems strange, as I say the TR69 connections seem to come in at 8.54ish every day, so I've had chance to check before and after it definitely seems to be when the noise margin alters on the hub but I haven't had any resyncs or reboots for just over 3 days now. Maybe its just a coincidence. Today it changed twice, once at 8:27 and then again at 8:54, both times it changed as I was keeping an eye on it before I went out. I think the 8:27 one coincides with when I did a fault check on bt's site oddly.
08:54:31, 24 Sep.
TR69 connectivity to (pbthdm.bt.mo) has been closed
08:54:31, 24 Sep.
TR69 sending GetParameterValuesResponse message
08:54:31, 24 Sep.
TR69 receiving GetParameterValues message
08:54:31, 24 Sep.
TR69 connectivity to (pbthdm.bt.mo) has been initiated
08:54:31, 24 Sep.
TR69 receiving InformResponse message
08:54:30, 24 Sep.
TR69 event found : 6 CONNECTION REQUEST
08:54:30, 24 Sep.
TR69 event found : 4 VALUE CHANGE
08:54:30, 24 Sep.
TR69 sending Inform message
08:54:29, 24 Sep.
TR69 event found : 6 CONNECTION REQUEST
08:54:29, 24 Sep.
TR69 event found : 4 VALUE CHANGE
08:54:28, 24 Sep.
TR69 sending Inform message
08:54:27, 24 Sep.
TR69 creating new session with ACS
08:54:25, 24 Sep.
TR69 ConnectionRequest: processing request from ACS
08:54:25, 24 Sep.
TR69 ConnectionRequest Failed
08:54:25, 24 Sep.
TR69 ConnectionRequest: processing request from ACS
08:54:23, 24 Sep.
TR69 connectivity to (pbthdm.bt.mo) has been closed
08:54:22, 24 Sep.
TR69 connectivity to (pbthdm.bt.mo) has been initiated
08:54:22, 24 Sep.
TR69 receiving InformResponse message
08:54:22, 24 Sep.
TR69 event found : 4 VALUE CHANGE
08:54:22, 24 Sep.
TR69 event found : 2 PERIODIC
08:54:21, 24 Sep.
TR69 sending Inform message
08:54:20, 24 Sep.
TR69 event found : 4 VALUE CHANGE
08:54:20, 24 Sep.
TR69 event found : 2 PERIODIC
08:54:19, 24 Sep.
TR69 sending Inform message
08:54:19, 24 Sep.
TR69 creating new session with ACS
08:51:37, 24 Sep.
TR69 ConnectionRequest Failed
08:51:37, 24 Sep.
TR69 ConnectionRequest: processing request from ACS
08:46:02, 24 Sep.
HTTP UserAdmin timeout from 192.168.1.7
08:34:16, 24 Sep.
HTTP UserAdmin login from 192.168.1.7 successfully
08:34:09, 24 Sep.
HTTP UserBasic login from 192.168.1.7 successfully
08:28:37, 24 Sep.
TR69 connectivity to (pbthdm.bt.mo) has been closed
08:28:36, 24 Sep.
TR69 connectivity to (pbthdm.bt.mo) has been initiated
08:28:36, 24 Sep.
TR69 receiving InformResponse message
08:28:36, 24 Sep.
TR69 event found : 4 VALUE CHANGE
08:28:35, 24 Sep.
TR69 sending Inform message
08:28:34, 24 Sep.
TR69 event found : 4 VALUE CHANGE
08:28:33, 24 Sep.
TR69 sending Inform message
08:28:33, 24 Sep.
TR69 creating new session with ACS
08:27:40, 24 Sep.
TR69 connectivity to (pbthdm.bt.mo) has been closed
08:27:39, 24 Sep.
TR69 sending SetParameterValuesResponse message
08:27:39, 24 Sep.
TR69 receiving SetParameterValues message
08:27:39, 24 Sep.
TR69 connectivity to (pbthdm.bt.mo) has been initiated
08:27:39, 24 Sep.
TR69 receiving InformResponse message
08:27:39, 24 Sep.
TR69 event found : 6 CONNECTION REQUEST
08:27:39, 24 Sep.
TR69 event found : 4 VALUE CHANGE
08:27:38, 24 Sep.
TR69 sending Inform message
08:27:37, 24 Sep.
TR69 event found : 6 CONNECTION REQUEST
08:27:37, 24 Sep.
TR69 event found : 4 VALUE CHANGE
08:27:36, 24 Sep.
TR69 sending Inform message
08:27:36, 24 Sep.
TR69 creating new session with ACS
08:27:34, 24 Sep.
TR69 ConnectionRequest: processing request from ACS
08:27:34, 24 Sep.
TR69 ConnectionRequest Failed
08:27:33, 24 Sep.
TR69 ConnectionRequest: processing request from ACS
08:27:31, 24 Sep.
TR69 connectivity to (pbthdm.bt.mo) has been closed
08:27:30, 24 Sep.
TR69 sending SetParameterValuesResponse message
08:27:30, 24 Sep.
TR69 receiving SetParameterValues message
08:27:30, 24 Sep.
TR69 connectivity to (pbthdm.bt.mo) has been initiated
08:27:30, 24 Sep.
TR69 receiving InformResponse message
08:27:30, 24 Sep.
TR69 event found : 6 CONNECTION REQUEST
08:27:30, 24 Sep.
TR69 event found : 4 VALUE CHANGE
08:27:29, 24 Sep.
TR69 sending Inform message
08:27:28, 24 Sep.
TR69 event found : 6 CONNECTION REQUEST
08:27:28, 24 Sep.
TR69 event found : 4 VALUE CHANGE
08:27:27, 24 Sep.
TR69 sending Inform message
08:27:27, 24 Sep.
TR69 creating new session with ACS
08:27:25, 24 Sep.
TR69 ConnectionRequest: processing request from ACS
08:27:25, 24 Sep.
TR69 ConnectionRequest Failed
08:27:24, 24 Sep.
TR69 ConnectionRequest: processing request from ACS
08:27:12, 24 Sep.
TR69 connectivity to (pbthdm.bt.mo) has been closed
08:27:11, 24 Sep.
TR69 sending GetParameterValuesResponse message
08:27:11, 24 Sep.
TR69 receiving GetParameterValues message
08:27:11, 24 Sep.
TR69 connectivity to (pbthdm.bt.mo) has been initiated
08:27:11, 24 Sep.
TR69 receiving InformResponse message
08:27:10, 24 Sep.
TR69 event found : 6 CONNECTION REQUEST
08:27:10, 24 Sep.
TR69 event found : 4 VALUE CHANGE
08:27:10, 24 Sep.
TR69 sending Inform message
08:27:09, 24 Sep.
TR69 event found : 6 CONNECTION REQUEST
08:27:09, 24 Sep.
TR69 event found : 4 VALUE CHANGE
08:27:08, 24 Sep.
TR69 sending Inform message
08:27:07, 24 Sep.
TR69 creating new session with ACS
08:27:05, 24 Sep.
TR69 ConnectionRequest: processing request from ACS
08:27:05, 24 Sep.
TR69 ConnectionRequest Failed
08:27:05, 24 Sep.
TR69 ConnectionRequest: processing request from ACS
08:26:23, 24 Sep.
TR69 ConnectionRequest Failed
08:26:23, 24 Sep.
TR69 ConnectionRequest: processing request from ACS
If the DLM has set some form of offset for the noise margin, is it likely to see variations around that still? Not sure if it going from 8.6 back to 8.8 is an indication its definitely seeing noise on my line still or if its something still in place by DLM from previous errors/dc's.
-
The TR69 messages will be coincidental with any changes in the current SNRM. If you keep watching the current SNRM, you're bound to see it go up or down slightly, because the noise level will change, from all sorts of electrical equipment, environmental conditions, crosstalk from other lines maybe even ADSL ones.
-
Thanks for your help. So would you say the noise margin currently set at 8.8db being higher than 6db being an indication there is a noise problem currently? Or could it be a precaution from DLM that might drop over time due to previous errors/noise.
Sorry for all the questions.
-
I think the elevated SNRM is a result of the DLM capping (banding) the speed. It could be due to the previous noise from all those electronics near the modem that you've removed, or it could be due to something else, some issue with the line. It's difficult to tell without more detailed stats.
I suppose you could re-instate all the electronics near the modem, check the SNRM before switching them on, then switch them on, and re-check the SNRM. If you see the the SNRM has decreased after switching all that kit on, then it would indicate that they are indeed a source of noise affecting your VDSL2 connection. However, some sources of noise may affect the error counts, which you can't see, more than they affect the SNRM, which you can see.
It could all actually be simply the DLM finding a stable setting for your line, in which case the DLM won't increase the speed unless the line is improved somehow.
-
Thanks for the response. I'm hoping its a sign that the line originally started at 36mb, then slowly increased over 5 days to 40.48mb before diving to 32.4mb where it is now that it can actually be stable at higher speeds, even if not at 40+mb. Unless for some reason it tests higher speeds even if the error count is high on lower speeds. Only thing that really changed was me installing that extra fan, checking temps with an infrared thermometer and seating the adsl cable fully. Other than that it can only be an external fault that developed recently or interference from a neighbours new line if it increases only when stable.
Is it normal to see the upstream noise margin remain around 6 even when the downstream increases? I'd of thought both would be affected. Its still punching above its max sync estimate of 7.7mb with 9.68mb atm.
-
No, I don't think so, I don't think there's anything to suggest a fault developed recently.
You start off on some default settings, and it takes a few days for the systems to detect that your modem supports G.INP, and then it enables G.INP, perhaps resulting in a slight increase in speeds, compared to interleaving+FEC. But it may never have been stable at the initial 36Mb or the subsequent 40Mb, and so the DLM made further changes.
If it had been operating for a few weeks at one speed, then the speed decreased, it might be different. But for only a few days, right at the start of the service being activated, I don't think there's much significance to it. Sure, if your speeds keeps getting slower and slower and slower, you would be able to report a fault, but if it's still operating within estimates, getting six engineers out to try and find a couple of extra Mb might not really be called for.
The FTTC DLM adjusts the downstream and upstream independently.
-
The upstream and downstream noise margins do vary independently, as different things may be happening in the frequency bands used by the two.
-
I haven't spoken to BT, nor have I reported a fault. I don't want to unless there is definitely a fault as I don't wish to waste bt or open reach's time unless I know something requires fixing.
Is there a modem that people generally recommend around here? I used a draytek vigor 120 when on ADSL and it was rock solid. Only thing I hated was not being able to see connection details without setting a static ip and connecting directly to the modem. I'm very happy with my asus ac66u at the moment so would ideally keep that for routing duties, but as long as i can dmz that should be okay.
-
Actually, I'm taking this way off topic, will post in the hardware forum.
In a rough attempt to bring this back on topic :D Would using my old adslnation XF-1e work, and is it likely to make any difference over the bt supplied one?
My master socket is well, I think its an extension socket? It was already installed when we purchased the new build, and the open reach engineer came around and checked the wiring and also removed the extensions from this after the photo was taken at my request so the only thing connected is this faceplate.
Photo:
(https://forum.kitz.co.uk/proxy.php?request=http%3A%2F%2Fi.imgur.com%2FoMQoZfY.jpg&hash=81c49ab05d92485faab120916051fd8610ca39d2) (http://imgur.com/oMQoZfY)
-
We seem to be missing the photo.
-
The old adslnation XF-1e should work OK. adsl filters are essential the same as those called vdsl or xdsl.
-
Thanks :) Have fixed the image link... It was on google photos, I did try it in a browser I wasn't logged into and it worked so who knows! On imgur now.
I don't think its a problem now after doing a bit of research... We have a grey box outside the house directly opposite which I guess contains the NTE hardware so basically all our internal stuff is just extensions. It seems they only fitted these briefly, typical we've ended up with one so now we don't have a test socket at all. The only thing connected to ours now though is this faceplate, as I say the open reach installer disconnected the other extensions (looks like they're connected in series) at the first point.
-
Could you also take a photo of the grey box outside? It could be an external NTE, or it could just be a junction box where lots of wires are (or were) connected.
Anyway, the photo of inside that socket shows one problem - there appears to be some extension wires connected and going somewhere (presumably back outside, so not very far). Possibly acting as a bridged tap. So basically, there's still at least one problem with your wiring. You can see there are two wires connected to each numbered terminal, but there needs to be only one wire on each terminal to fully disconnect the extension wiring. Also the orange bell wire could probably be disconnected. Presumably all you need to do is carefully pull the top wires out of those IDC terminals, assuming those are the extension ones. And disconnect both orange wires.
-
As I mentioned, that photo was taken BEFORE the open reach engineer came to our property when the line was activated and I specifically asked him to remove the other extension wires to reduce possible interference, which he did. They go behind the plasterboard and are wired to the other extension panels wired throughout our house. So now, only one set of wires will be connected. I just checked the other phone points in the house, they're as dead as a dead thing :)
(https://forum.kitz.co.uk/proxy.php?request=http%3A%2F%2Fi.imgur.com%2FraDgflv.jpg&hash=e2ffad14ac8792c7fc99ca097c1cca3b3c419121) (http://imgur.com/raDgflv)
-
Right, so it is an external NTE, and there's probably no point in disconnecting the bell wire on the inside, if it's still connected, because it tends to not be connected to anything at that end anyway, some microfilters only have two pins present in the plug that goes into the phone socket.
-
I know the adsl cable going to my hub at the moment (supplied by bt for the self install) only has 2 pins on it, where the one I was previously using had all 4. Not sure if that's good enough to bypass interference or not. I have no idea if the BT supplied adsl filter only has 2 pins for both, the xf-1e has all 4 though.
Do you know if its technically legal for me to open the grey external nte to check if the bell wires even connected?
-
As far as I'm aware, external NTE's should not be used for DSL connections, they should be bypassed.
-
Hmm it was fitted in around 2011, would they have not of thought of that if they were not compatible? I really wish we could of got the speeds for Infinity2 so an actual OR engineer could of done the install for us and hopefully have checked it all over first. Would love to have a standard NTE5. Is there any chance we could get BT to remedy this for us?
-
You'll need an Openreach technician I think. Here's an old thread about it http://forum.kitz.co.uk/index.php?topic=10663.0
-
Hmm :( Not sure I feel comfortable doing that myself. Any idea how I'd go about trying to get an OR engineer to sort this for me?
-
I wasn't aware of anything in particular about external NTEs that they must be removed for DSL services. At the time when FTTC was only available with an engineer install, sure, the external NTE would have been bypassed or removed in order to fit an internal NTE5a and filtered faceplate.
The advantage of a modern master socket and centralised filter is mainly so that it can filter off all the extension wiring at one point, so if you don't have any extension wiring, this isn't applicable.
What are your estimates speeds? Is there really anything that needs sorting out?
-
The relevant SIN 470 (http://www.btplc.com/sinet/SINs/pdf/470v2p1.pdf) states in Section 5 --
5. Ceasing usage
From November 2011 the External NTE will no longer be provided at NewSites. Housing
developers will instead install an internal NTE5. All existing External NTEs will remain in
place until faulty or removed if an End User changes to an incompatible service.
-
The relevant SIN 470 (http://www.btplc.com/sinet/SINs/pdf/470v2p1.pdf) states in Section 5 --
5. Ceasing usage
From November 2011 the External NTE will no longer be provided at NewSites. Housing
developers will instead install an internal NTE5. All existing External NTEs will remain in
place until faulty or removed if an End User changes to an incompatible service.
Any idea what exactly an "incompatible service" is? Most importantly, is that FTTC?
-
Informal discussions with both current and ex Openreach technicians has established that an xDSL service can be regarded as an "incompatible service".
I am fairly sure that when Black Sheep has been tasked to fault-find a poorly performing xDSL service and there was an XNTE is situ, one of the first steps would have been to check if that XNTE was still in circuit or was just acting as a big waterproof housing for a pair of gel-crimps.
-
But BT SIN 470 also says:
standard inclusion of a filter preventing electrical interference in DSL installations;
I assume that's referring to the filter on the bell wire, as found in a modern NTE5a.
-
I assume that's referring to the filter on the bell wire, as found in a modern NTE5a.
Yes, your assumption is correct. It is just a 22 mH choke in series with the connection (from IDC3) to the junction between the series connected 1.8 microF capacitor and the 470 k resistor that shunts across the pair.
-
Does that mean that disconnecting my bell wire won't actually make any difference, as its already filtered in the external nte? Pretty sure mine will be wired "correctly" as there is no internal NTE5, so I'd be without an NTE at all if its not connected.
I tried to open the external NTE box to take a look with quite a lot of strength, but the screw won't turn. It was flexing most the box, so I stopped to stop the risk of any damage. Must of been tightened by he-man, unless its a screw with a reverse thread :-\ I'm only in my thirties and although no muscle man, i'm not weak.
-
A "bell wire" is really unnecessary with any modern telephone and is just a relic of times past.
The other designated purpose of the series connected resistive-capacitive shunt is to provide the correct loading impedance to the exchange equipment when no telephone is connected to the circuit.
I tried to open the external NTE box to take a look with quite a lot of strength, but the screw won't turn. It was flexing most the box, so I stopped to stop the risk of any damage. Must of been tightened by he-man, unless its a screw with a reverse thread :-\ I'm only in my thirties and although no muscle man, i'm not weak.
How odd. It should be fairly straightforward to open. I can't explain the difficulty you have experienced. ???
-
Yeah, just to put it in perspective I recently built several self standing shelving units from some old crates for the shed and in the past some for under the stairs from raw materials... done plenty diy. I have decent tools and know how to use a screwdriver ;) But that xnte did not want to let me in :no: Not without force I didn't want to exert on plastic that was already flexing.
Is it still worth trying to dc the bell wire to make sure? Assuming it doesn't matter if I do this internally on the faceplate.
-
Is it still worth trying to dc the bell wire to make sure? Assuming it doesn't matter if I do this internally on the faceplate.
Just do it at the socket. I.e. remove (pull out, don't cut) all wires in IDC3.
-
Just do it at the socket. I.e. remove (pull out, don't cut) all wires in IDC3.
As advised 3 times in the BT forum ;)
-
As advised 3 times in the BT forum ;)
Oh wow sorry I didn't even notice the usernames! What a dingbat. I was reading the info on here about the choke at the same time and wasn't sure if that negated the need to remove the wire. Apologies. I suffer from anxiety and ocd, sometimes that switch that tells me when I'm worrying too much about things just doesn't work. Thank you for your help
-
The reason why it doesn't matter about disconnecting the bell wire at that end, is because it'll be connected to nothing at that end anyway! So pulling a wire out of a terminal that doesn't connect to anything, doesn't change anything, it'll still be a disconnected end of a piece of wire. That's assuming you have a decent microfilter, which will make no connection to that pin in the phone socket. A bad microfilter might just pass the bell wire straight through to the telephone, completely unfiltered.
-
Ha well I just had a shock. I went to go remove the bell wire, and guess what... I noticed my faceplate looks like this (please excuse the dust its behind a cabinet):
(https://forum.kitz.co.uk/proxy.php?request=http%3A%2F%2Fi.imgur.com%2FetWmRfC.jpg&hash=7529b7df7db7fe978e204ca31612c35da03147ac)
That's an NTE5, isn't it. What a tool... I remember speaking to the guy the entire time he was here, I don't remember him swapping the faceplate at all! It was in 2013, so guessing he probably bypassed the xnte and fitted this instead? I can't remember if I saw him messing with the xnte but he may of before knocking. I haven't popped it open as I don't want to dc the hub as its been 5 days stable now.
-
D'oh! :doh:
So you just need to fit a Mk 3 SSFP to the NTE5/A.
-
Informal discussions with both current and ex Openreach technicians has established that an xDSL service can be regarded as an "incompatible service".
I am fairly sure that when Black Sheep has been tasked to fault-find a poorly performing xDSL service and there was an XNTE is situ, one of the first steps would have been to check if that XNTE was still in circuit or was just acting as a big waterproof housing for a pair of gel-crimps.
Just spotted this, Mr Cat …… as always, spot on …….. protocol (at least on my area) is to straight-crimp through the UG feed directly to the internal wiring. :)
-
Will order the MK 3 SSFP, should be super easy to fit as there are no extensions :)
Assuming if the OR engineer had left the xnte connected I'd of had a whole host of issues with two nte's on the line? Shame I couldn't open the xnte to check.
-
Well, in that case there would have been two resistive-capacitive shunts across the pair. I guess that remote testing would deduce that the line length is either longer or of a different make-up then reality. :-\
-
Hmm interesting :) I noticed some of the businesses connected to our cab were able to order fibre a few weeks before we were while we (the entire estate of 27 houses based on the few I checked) just got the following message:
"Your cabinet is enabled for Superfast fibre however you're not able to order fibre just yet. This might be because the length of the line is too long to get Superfast speeds. We're actively looking at other options. If you'd like us to let you know if fibre becomes available, register your details."
Pretty sure that's just a coincidence though as https://www.telecom-tariffs.co.uk/codelook.htm and Black Sheep (who was super kind enough to check a LOT for me) stated it wasn't showing as active, where it was shortly after I was able to order. Not sure if its usual for business premises to get fttc first :-\
-
Now don't expect a massive change in stats dfects when the SSFP MK3 is installed it's very subtle in how it works but over a few weeks there should be less errored second & CRC events in the router logs :fingers:
-
I don't expect miracles :) If it improves connection speed/stability that's more of a bonus. Be nice to not have to wonder if the microfilter i'm using is hampering anything.
-
I just had another go at opening the xnte to check it was bypassed, man that screw will not turn :lol: 3 different phillips #2 screwdrivers and it will not budge with literally all my strength to the point my arms were shaking. Thinking the guys superglued it in or something! I did notice the other end of the screw tunnel is open, not sure if there is a captive nut in there that's rusted tight or something. ???
Think the only way I'll ever see inside is by removing the rubber bung on the side and peering in with a torch :P
-
@burakkucat Don't you always have a double ring capacitor and resistor in place with a filtered faceplate. Both the back of the nte5 and the filter plate have one. I have aways hoped that 470k was big enough to ignore?
-
As much as I would like to categorically say either "yes" or "no" I am restricted by not having access to the circuit diagrams of the SSFPs -- the original, the Mk 2, the Mk 3 and the Mk 4. (I have a suspicion that the Mk 3 & the Mk 4 are electronically identical and are just different physical presentations thereof.)
The more I think about the SSFPs, the more uncertain I become. I expect that an analysis of the components used will show a low-pass filter between the "network side" and the telephony port. I would not be surprised to know that there is also a high-pass filter between the "network side" and the xDSL port. I would be very surprised to know that there is yet another resistive-capacitive shunt across the "network side" as that is a necessary definition of the network termination. (I forget which SIN defines that requirement.) I would not be surprised to know that the "bell wire", originating within the NTE5, is provided with a further degree of decoupling before it transits to the lower front face-plate where it passes through a 22 mH choke before appearing on "pin 3" of the telephony port.
Hmm . . . Somewhere in the grotto I have both a broken Mk 2 and a broken Mk 3 SSFP. Given a bright light, a steady hand and good eyesight I guess it should be possible to sketch out the circuit. Application of ~50V DC across the "network side" and usage of a DMM should aid the identification of unknown circuit components. (The other alternative is to gain access to Grimbledon Down -- or whatever is the current name for Martlesham Heath / Adastral Park -- and look in the Wizards' archive of circuit diagrams.)
:-\ :hmm:
-
You helpfully uploaded some photos a while ago http://forum.kitz.co.uk/index.php/topic,14381.msg269386.html#msg269386. You need the board in your hand to see the circuit but the capacitor is clear and my recollection is that having junked the NTE5 bell wire by only re connecting up the a and b the bell wire is regenerated directly between the incoming a and b. My recollection is just that, as I was not really checking for that.
p.s. As an aside and re http://forum.kitz.co.uk/index.php/topic,14381.msg269651.html#msg269651 I think the Mk3 choke is bifilar albeit not very well wound. The one in my Mark 3 is tidier than the picture but it looks as though the second layer of bifilar winding spreads out unevenly over the first layer.
-
The choke, to which you refer, is not for the "bell wire" but is for common-mode signal rejection. Of the two windings on that toroid, one is in series with the A-wire and the other is in series with the B-wire. Now that you have reminded me, I do recall looking at that choke on both the Mk 2 and the Mk 3 SSFPs.
With your assistance I do also recall that the plug of the SSFP only carries two connections, that are made via the "test" socket of the NTE5. Consequently the signal for the "bell wire" has to be regenerated on the SSFP. Hopefully that capacitor is connected to the downstream side of the low-pass filter and not the upstream side . . . :-\
We really need --
- to see the circuit diagram of the Mk 2 SSFP.
- to see the circuit diagram of the Mk 3 SSFP.
- an experimenter to remove the (above discussed) capacitor from the Mk 2 SSFP and check the resulting performance of the SSFP.
- an experimenter to remove the (above discussed) capacitor from the Mk 3 SSFP and check the resulting performance of the SSFP.
-
Here are pictures of both sides of a mark3. I have yet to puzzle the circuit fully out regarding the regenerated bell wire. It does look like your correct about the bell wire being regenerated after the filter but at first sight one end of the capacitor does not seem connected to anything! :'(
[attachment deleted by admin]
-
After my last post, yesterday evening, I went routing through the what-not (within the grotto) and extracted both the Mk 2 and the Mk 3 PCBs.
A quick look (with the aid of a magnifying glass) enables me to say that there is not a (duplicate) resistive-capacitive shunt across the pair. The "big yellow capacitor" just connects "terminal 3" to "terminal 2" (using the understood numbering as of the IDCs at an extension socket).
Interestingly I saw that the capacitor was marked "1u8K250V". I can translate the "1u8" to 1.8 micro-Farad and the "250V" to be the maximum voltage rating but the mention of Kelvin (the upper case "K" leaves me mystified. (Unless the "K" indicates the capacitor type . . . it's not electrolytic, so polyester would be my guess.) :-\
-
I am puzzled over what is providing the bell wire voltage unless that comes back from the phone. It definitely does not come from the NTE5 and I agree that the capacitor " just connects "terminal 3" to a filtered "terminal 2" ". I assumed that something ought to be raising terminal 3 to terminal 5 voltage via a high resistance. I have one wired phone that definitely needs a bell wire and it certainly rings with the MK3. Does that imply that just the capacitor (which is the right value for the ring), connected between 2-3 is enough for the ring and the phone has the needed charging resistor?. As I said I am puzzled.
-
Internet resyced for the first time in about a week today, upload speed has increased to exactly 10 megabit (from 9.6), download speed still 32.40 exactly. Upload noise margin increased a little as did the downstream. Attenuation is .1 less :P
DSL uptime:
0 Days, 8 Hours 55 Minutes 44 Seconds
Data rate:
10.00 mbps / 32.40 mbps
Maximum data rate:
10000 / 38508
Noise margin:
6.7 / 8.9
Line attenuation:
25
We're about 1km in cable length from the cabinet (thanks to black sheep for looking this up), upload speed is over performing the estimate by nearly 3mb, download is at the lower end of the estimate :-\
-
The MKT 82 is a Metallized-Polyester-Film Capacitor
I believe the k refers to a tolerance of 10%
-
Your download is capped at 32.4Mbps and will stay that way until (if) DLM relents and removes the banding.
-
The MKT 82 is a Metallized-Polyester-Film Capacitor
I believe the k refers to a tolerance of 10%
Thank you.
-
I am puzzled over what is providing the bell wire voltage unless that comes back from the phone. It definitely does not come from the NTE5 and I agree that the capacitor " just connects "terminal 3" to a filtered "terminal 2" ". I assumed that something ought to be raising terminal 3 to terminal 5 voltage via a high resistance. I have one wired phone that definitely needs a bell wire and it certainly rings with the MK3. Does that imply that just the capacitor (which is the right value for the ring), connected between 2-3 is enough for the ring and the phone has the needed charging resistor?. As I said I am puzzled.
No resistor is required. The "ring" is due to the superimposed AC voltage across the pair (the A- & B-wires). In the telephone, the sounder is connected between the A-wire and the "bell-wire". The "bell-wire" returns to the 431A plug via the line-cord. At somewhere in the "local circuit" the "bell-wire" is connected to a 1.8 micro-Farad capacitor, the other end of the same capacitor is connected to the B-wire. Hence when "ringing current" is applied, there is an AC path as follows:
A-wire <---> sounder <---> capacitor <---> B-wire
-
Your download is capped at 32.4Mbps and will stay that way until (if) DLM relents and removes the banding.
I concur. Without the banding, I would expect to see a DS of close to the package limit (Openreach 40/10 Mbps package provisioned).
-
Kinda thought as much :) Seemed to resync around the time I left the house this morning which is an odd coincidence.
Assuming its unlikely the OR engineer would of fitted the NTE5 and left the XNTE in circuit... and if he had it would have more of an affect then what I'm seeing?
-
Assuming its unlikely the OR engineer would of fitted the NTE5 and left the XNTE in circuit...
Agreed. Unless, of course, he was unable to open it! :D (If the latter was true, I would have expected the case to have been broken open, the remnants removed, a replacement BT66 to be fitted and the network service feed being joined to the internal wiring by a pair of gel-crimps.)
and if he had it would have more of an affect then what I'm seeing?
I don't think what you are seeing could be attributed to a circuit with double shunts.
-
Thank you very much sir :) Pretty sure its rusted up, I can see rusted screws in photos on google and it seems to screw into a metal threaded sleeve. Its been 3 years since the OR engineer was here, plenty time for it to of rusted shut. Maybe it was wet when he came, wish my memory was better ??? I was going to try some wd40 to try and free it but did a test on the bottom corner and its discoloured the plastic so thrown that idea out.
Going to leave it alone and see if anything improves. Thanks again.
-
Have you purchased the SSFP MK3 and slapped it over the NTE5 as you may well need it on the longer line trust me :)
-
Have you purchased the SSFP MK3 and slapped it over the NTE5 as you may well need it on the longer line trust me :)
Its in royal mails hands ATM might show up today :) it worth me fitting it straight away at the risk of DLM seeing another drop?
-
Just turn the modem off before you start and leave it off for 30 minutes.
-
Thanks to everyone on this thread. I followed the advice given above and fitted an VDSL faceplate MK3 this morning after switching the router of for over 30 minutes.
There are no extensions as all the telephones are cordless.
The faceplate was fitted to the box where the telephone line enters the house beside my computer and router.
For the record the info before and after:-
Download Before 25.56Mbps improved to 29.64Mbps and certainly in line with speed one would expect at around 1Km from the cabinet.
Upload Before 3.19Mbps improved to 3.49Mbps - Lack of information as to what one would expect at 1Km.
Noise Margin 6.2/6.5 6.0/6.2
Lin attenuation 28.2 28.1
-
For the record the info before and after:-
Download Before 25.56Mbps improved to 29.64Mbps and certainly in line with speed one would expect at around 1Km from the cabinet.
Upload Before 3.19Mbps improved to 3.49Mbps - Lack of information as to what one would expect at 1Km.
Noise Margin 6.2/6.5 6.0/6.2
Lin attenuation 28.2 28.1
The Noise Margin and Line Attenuation values are as expected.
As for a typical synchronisation speed for an approximately 1km line to the DSLAM, take a look at Bald_Eagle1's circuit statistics.
-
Finally plucked up the courage to remove the adsl filters and install the MK3. As Dray helpfully suggested I left it turned off for 30 minutes while I fitted it. Its reconnected now and is looking slightly better in terms of the noise margin:
Before (remembering kbps is actually mbps due to a bug):
(https://forum.kitz.co.uk/proxy.php?request=http%3A%2F%2Fi.imgur.com%2F7THEhOW.png&hash=b99d7d92987e731c5a2ac8568213ad172012f19b)
After:
(https://forum.kitz.co.uk/proxy.php?request=http%3A%2F%2Fi.imgur.com%2FCwODpMo.png&hash=17530b5fdc8f620e118f9e18d0281a175c5d524d)
I haven't seen the downstream noise margin below 8.5 since our speed was banded to 32.4mbps.
I didn't expect the downstream sync speed to increase as I'm pretty sure we're banded at this point, but at least its no worse and the noise margin is less!
Our next door neighbour recently had fibre activated too, and hes apparently only getting 16mbps down! My 32.4mpbs doesn't seem so bad at the moment. Will be trying to help him.
-
The noise margin being less isn't a good thing, it's a bad thing (with the capped speed), and the max attainable speed is also slightly worse. However, the error rate, which you can't see without stats, may be better with the MK3.
-
I thought the ideal was being closer to 6? The max attainable swings between 37 and 39 with each sync, although this is marginally the lowest its been so far.
-
If there is something limiting the speed, such as the limit of the product (40/10 or 55/10 etc.), or DLM imposed banding, then a higher SNRM is better, as it indicates greater potential for more speed if the speed cap were removed.
Normally, your modem would connect at the best speed it can manage, with the SNRM at the target of 6 dB. If the current SNRM subsequently increases, that's a good thing, as it indicates there's less noise present, and if your modem re-connected at that moment, it should get a slightly higher speed.
-
As the sync rises the snrm drops. The target is 6dB but as you're banded it's above that. The snrm dropping without the sync rising usually means extra noise on the line. Seeing as it's a result of a new SSFP, it suggests the MK3 wouldn't sync as high if you weren't banded. If the snrm was higher with the same sync that would be good as it would suggest a higher attainable sync.
As ejs stated it may have a lower error rate, but you would need a modem with detailed stats to see this. A lower snrm is only a good thing if it has a higher sync along with it.
-
Ahhh that makes sense. So essentially I could of made my connection worse by fitting the mk3 ☹
-
Would I be best to leave it and see if it improves or swap back to the old setup?
-
Leave it alone for several weeks to see if the banding ever gets lifted. Or obtain a modem with access to more detailed stats.
-
OK so time for an update.
The HUB stayed connected pretty stably for a month or two, only rebooting once every 14 days which apparently is a normal thing for these? Until last weekend, where we dropped down another band to 30mbps. I'd observed the average SNR dropping down from around 8 down to into the 6's before the banding drop. We were one of the first on fibre on our estate, so maybe the effects of crosstalk from new sign ups?
Our current stats:
Data rate: 9.36 mbps / 30.00 mbps
Maximum data rate: 9359 / 38103
Noise Margin: 5.1 / 9.1 (downlink stays around 8.9-10 atm)
Profile speed (from the bt wholesale speed checker): 27.76 / 10
Speed test results: 27.12 / 9.1
Ping: consistently about 15ms on speed tests
Adsl checker results (I find it interesting the upstream has always exceeded these):
(https://forum.kitz.co.uk/proxy.php?request=http%3A%2F%2Fi.imgur.com%2FInKbhZd.png&hash=ed158fd685f3a66b7f2e21b8b02242f84b68d8d8)
Quiet line test either in the test socket or main socket (we have no extensions anyway) has no pops or crackles, just a very quiet hum that increases as my hand gets close to/tightens around the coiled handset wire which I think is more a case of poor shielding/components than line noise? Wonder if you can get any line testing devices rather than just using a phone?
I have a HG612 coming from ebay that I couldn't resist at £12.98 delivered, but I'm going to be a bit hesitant about installing as from what I've read if it causes us to drop another band, DLM recovering it will probably never happen.
-
Is the external NTE still connected?
-
Yes, hopefully crimped inside rather than connected. I never managed to get into it due to the screw apparently being rusted in. NTE5 master socket with a mk3 VDSL faceplate.
-
Sorry to drag this post up from the depths, but its already got most the info in. The hub rebooted again last night after 14 days as seems standard with these, but the last couple of days the snr had dropped from 6/9.8 ish to 3.8/6.8. Max attainable speed tanked from roughly 38,000 each sync to 32,808 and we lost maybe 1mb of upload sync speed. Still capped at 30mb with a profile of 27.76 on the downstream.
One of our nearest neighbours knocked the other day to ask about our broadband, hes only getting 16mb out his line. Our next door neighbour only gets 18mb and one just down the street said hes seen 23mb on his line, but its currently about 19mb! Looks like we have one of the more stable lines on the street, although all the ones I know of are under performing. I guess this is likely to be either lots of crosstalk or some interference somewhere :( I thought it being a new estate with new wires these kind of things would of been avoided :(