Kitz Forum
Broadband Related => Broadband Hardware => Topic started by: graevine1 on December 30, 2007, 11:37:43 AM
-
Currently I have a Netgear DG834G V3 an exceptionally good router especuially as this site has provided the information to access the extended diagnostics for instance FEC, HEC errors, Interleave etc etc etc all power to those involved.
However its time to upgrade ie this will also enable us to take advantage of must greater in particular Wireless speed and Mimo technologyetc etc.
So as a trial we have purchased a DG834N but alas would like a little assistance to achieve the full diagnostics that are not available from the menue given as standard in that router, or should we be deciding on another model in the Netgear range (The choice for us is only one from the Netgear range of routers bit restrictive for us but thats life.) So has anyone got an answer to the patches necessary to use this 270Mbps N class router to provide the diagnosis of the old twisted pair feed. [Thanks to all and great New Year to all without too much Contention!! being added]
-
Not really what you were asking (but you probably will soon) - you are aware that the "270Mbps" is extremely misleading yes? Best case is that you'll get maybe 40Mbps sustained data transfer as 802.11n is designed specifically to support two concurrent HD streams which run around 18Mbps each. I have tried most 802.11n access points and best case is around twice the speed of 802.11g (with WPA/WPA2 encryption). Re MIMO - well I haven't seen any improvement at all in range - outside yes; inside no.
Hopefully someone else has a DG834PN and can help with the menus.
-
I am aware that it aparently will not achieve what appears to be the stated claimed speed, but it should be the highest that may be achieved.
What is most important is that we can discover the full measurements of the twisted copper pair and the service that we hope for and believe we are paying for.
That was the reason for my question.
I was led to believe that the 834N superceeded the 834 PN. I have also put in a call to Netgear and hope also they come up with some requested answers . It will all then be posted up for all to see.
Thanks for your interest, lets see how it goes.
-
802.11n is unlikely ever to be ratified IMHO. In terms of wireless connectivity I do not believe there is a difference between the Pre-N and what Netgear now sell as the N. There may be a difference in the ADSL chipset but I don't believe so.
-
I dont have a DG834N therefore I could be wrong... but from what Im led to believe it is a similar chipset to the DG834NB/DG834PN.
If so then its already on the router stats page under Netgear DG834GT / DG834PN/DG834N
http://www.kitz.co.uk/adsl/frogstats.htm#anchor26b
Also worth looking at is
http://www.kitz.co.uk/tute/DMTv8.htm
-
oh.... and if it is a broadcom chipset - then you may be interested in this for advanced configuration for the broadcom based routers.
http://www.kitz.co.uk/tute/voyager_2100_cli.htm
-
Thanks Kitz, and we wish you a Happy Healthy Prosperous Year, in addition I hope your fingers dont get any shorter like mine being worn away by the keys !! Lol.
Attempted to Telnet and got the following message.
Connecting to 192.168.0.1 ... Could not open connection to the host, on port 23: connect failed
I temporally disconnected the Norton Firewall but to no avail so I guess could this be a necisity to open the Router firewall, any ideas please.
-
Did you put the router in debug mode first?
~ Enable debug via the web interface
http://192.168.0.1/setup.cgi?todo=debug
-
Thanks Kitz, no I didnt on this occasion, put that down top a lack of the seasons necessary lubricant !!!!
Well for your files this is the result
BusyBox v1.00 (2006.10.04-06:55+0000) Built-in shell (ash)
Enter 'help' for a list of built-in commands.
# help
Built-in commands:
-------------------
. : alias bg break cd chdir continue eval exec exit export false
fg hash help jobs kill let local pwd read readonly return set
shift times trap true type ulimit umask unalias unset wait
#
#
So what do you suggest for the next stage as we wish to see all we can concerning the route to the exchange etc, in fact as much detail as we can obtain. Thanks !!
-
type the following at the command line when you're telnetted in:
/usr/sbin/adslctl info --stats
And see what it outputs.. this is pretty much all you'll get from it. You won't see anything about "route to the exchange" (not even sure what you mean by that), but the router knows nothing about that, other than it's connected!!
-
Thanks Mr Chris . This is most of what we are after as you will see
/usr/sbin/adslctl info --stats
BusyBox v1.00 (2006.10.04-06:55+0000) Built-in shell (ash)
Enter 'help' for a list of built-in commands.
# /usr/sbin/adslctl info --stats
/usr/sbin/adslctl: ADSL driver and PHY status
Status: Showtime Channel: INTR, Upstream rate = 448 Kbps, Downstream rate = 358
4 Kbps
Link Power State: L0
Mode: G.DMT
Channel: Interleave
Trellis: ON
Line Status: No Defect
Training Status: Showtime
Down Up
SNR (dB): 9.3 19.0
Attn(dB): 47.0 25.5
Pwr(dBm): 18.8 11.8
Max(Kbps): 3872 988
Rate (Kbps): 3584 448
G.dmt framing
K: 113(0) 15
R: 16 4
S: 2 4
D: 16 8
ADSL2 framing
MSGc: 1 1
B: 113 15
M: 2 4
T: 1 1
R: 16 4
S: 2.1403 4.2500
L: 912 128
D: 16 8
Counters
SF: 114341 114339
SFErr: 1 0
RS: 3887606 1943763
RSCorr: 137977 0
RSUnCorr: 1 0
HEC: 1 0
OCD: 0 0
LCD: 0 0
Total Cells: 16430867 0
Data Cells: 35944 0
Drop Cells: 0
Bit Errors: 0 0
ES: 1 0
SES: 0 0
UAS: 11 0
Total time = 32 min 42 sec
SF = 114341
CRC = 1
LOS = 0
LOF = 0
ES = 1
Latest 1 day time = 32 min 42 sec
SF = 114341
CRC = 1
LOS = 0
LOF = 0
ES = 1
Latest 15 minutes time = 2 min 42 sec
SF = 9558
CRC = 1
LOS = 0
LOF = 0
ES = 1
Previous 15 minutes time = 15 min 0 sec
SF = 52923
CRC = 0
LOS = 0
LOF = 0
ES = 0
Previous 1 day time = 0 sec
SF = 0
CRC = 0
LOS = 0
LOF = 0
ES = 0
15 minutes interval [-30 min to -15 min] time = 15 min 0 sec
SF = 51860
CRC = 0
LOS = 0
LOF = 0
ES = 0
#
#
However , there is quiet some variation in the figures obtasined using different routers,
-
>> However , there is quiet some variation in the figures obtasined using different routers,
Something I noticed many years ago - way before even maxdsl came in.
-
Thanks Kitz, but I am also interested in why the figures obtained for example simply in relation to Line Attn and SNR vary from router to router.
Well you may finf this usefull
/usr/sbin/adslctl info --stats
BusyBox v1.00 (2006.10.04-06:55+0000) Built-in shell (ash)
Enter 'help' for a list of built-in commands.
# cat proc/cpuinfo
system type : 96358GW-16
processor : 0
cpu model : BCM6358 V1.0
BogoMIPS : 299.00
wait instruction : no
microsecond timers : yes
tlb_entries : 32
extra interrupt vector : no
hardware watchpoint : no
VCED exceptions : not available
VCEI exceptions : not available
#
#
Little doubt whats inside this Netgear DG834N
My programing knowledge is years and years out of date, so must learn to read a book again and also learn to swot (and retain)
Want to find FEC errors, LCD errors, OCD errors, Line up Count, up time or show time, line up time, all will be explained later.
Ealier today using the dg834g Ver3 I found that my exchange is Ti so it would be good to find this info also. There will be more Ive no doubt.
-
Thanks Kitz, but I am also interested in why the figures obtained for example simply in relation to Line Attn and SNR vary from router to router.
Its consumer grade equipment so all the passive components will be +/- 10% and they will vary depending on temperature. The filters you use are frankly junk (I used to design notch filters for military radar amongst other things - Rapier if anyone has heard of that) and so are your extension cables (usually). The figure you see for noise margin is (sort of) an average across the frequency range anyway.
In addition you will get increased noise at night (Heaviside layer in the ionosphere reflects more MF); the real attenuation on the line will also vary due to temperature, humidity and about 10% of UK phone lines are under water at some point in the year - I've even seen cables frozen in ice before. Then there's crosstalk, induced and conducted noise........etc etc etc
Frankly I'm amazed ADSL works at all given the output power and I suspect my ADSL2+ service is only going to get worse. SRA makes it much better but I can't see BT going for that anytime soon.
That probably doesn't help a lot :D
If anyone out there is doing a maths course then do pay attention to Fourier and Laplace transforms for they aren't esoteric nonsense that nobody uses - they are in fact one of the main reasons your ADSL works at all.
-
Thanks to Rizla and Mr Chris and I wonder if you have any further pointers, it would be good if we could also establish from this 834N what is in use in the exchange, and the points I have raised above. Linux is not in my experience list so if anyone knows it would save so much time.
A great new year to all.
-
These days what is "in use at the exchange" is almost entirely irrelevant regarding your speed/attenuation/etc. I would go so far as to say that it is irrelevant for <8Mbps ADSL connections. For ADSL2+ it may be a factor in some instances (usually "power users" wishing to tweak to the max) but even then its not really important as 99% of them have no idea what they are doing anyway.
Even if you do know the chipset in use on the DSLAM/MSAN then it's not that useful anyway as you will not be able to determine the firmware in use.
Frankly if you wanted info from the modem then Netgear is about the worst choice you could have made as something which doesn't have enough memory to even display bitloading/tone is even less likely to provide remote info.
-
Hi Rizla, Well Im not exactly certain because the 834 G Ver3 certainly enables you with the Ti AR7 chipset to determine what is at the exchange. AND it appears to be faster on line than this 834N with its Broadcom chipset. I do however want access to asome more facts such as FEC errors and "line up counts" etc
I hope this can be found. I see no reason why not. at any rate Im finding out a lot more today by further interrogation of the broadcom foirmware etc.
Have a great year.
-
FEC errors are not errors. FEC errors by definition is the number of corrected pakets by the DSlam when Interleaving is enabled. HEC errors are errors and need to be monitored. The N series do have issues but for a different reason. The new series router uses the Broadcom chipset for the modem code and then utilises the Atheros chipset for the wireless side of the router.
This was supposed to enable the new wireless standard of 108mb/s however it has been known to cause intermittenty connection as the code form both chipsets react with eachother in uncertain ways.
This is why the SpeedTouch routers are still better for wireless however the new Belkin series uses both the Atheros and the Broadcom chipsets but they do it so the code in the middle buffers the WAN and the LAN sides.
-
Thanks Azzaka, Im sorry but to use the word error was wrong of me, you are of course correct. You have hilighted a concern of myself re the internal chipsets, but i believe you have spotted where I am coming from and where I am leading to. I believe with a number of twisted pair connection problems that no doubt will increase then it is imperative that one can by whatever means be in a position to interrogate the route by means of the router/modem. I do not believe that the twisted copper network is being maintained to the standard expected or to the standard that it may have been a few years ago. I have not had the oportunity as yet to examine the Belkin but may do soon. Thanks for your interest. Any further observations would be greatfully received.
-
it is imperative that one can by whatever means be in a position to interrogate the route by means of the router/modem.
What information would you be looking for here? When you say interrogate the route, what exactly do you mean by this?
-
I was attempting not to be too technical, with my choice of words. I am firmly of the opinion that 98% of the problems that users of broadband over the twisted pair are affected by, or are down to "the route" ie from the exchange via the 'E' side cable to the CAB and the 'D' side cable to the dwelling/premises, so the more information that we can extract using the modem/router the better and easier it is to establish the problem and therefore point the finger at the maintainer/supplier of the route pair (or the one wire). More will come to light by next week I hope. However, there does appear to be surfacing a possibility of incompatability of some chip set/software combinations, this is starting to leak out. This will only affect some few % of instalations, and may not be able to be rectified by firmware flashes in all cases. There may be cases of errors/service difficulty introduced between exchanges and router/modem, and/or wi/fi processor/modem processors or the buffers/translators between them. I hope no one would dissagree that the time has come whereby the user should be able to interrogate a router with ease and be able to produce a print of performance for records or maintenance. After all you dont buy a new car and find it stops working for even a minute a day, or buy a gallon of petrol and go away happy with being told that the six pints is really the one gallon you have just believed you have paid for !!!!!!.
-
Hiya,
If I read you correctly, what you're wanting from a modem is something akin to the kit that comms engineers use to determine line length, or where there is a break, etc. Unfortunately you're being very vague about what you would require from a modem. You keep mentioning that you would like to see "more information", but you do not say exactly what information you would like to see. Without actually expanding upon this, I can see this thread going round in circles and everyone simply getting frustrated, with no decent outcome.
Modem chipsets, as I'm sure you know, can tell you information about ATM cell fault counts, etc... and the calculated "attenuation" and "SNR margin" (I quote these because they are of course a sort of average, as the SNR and attenuation will change by frequency).
A modem chipset cannot perform enhanced diagnostics in the same degree as the expensive kit used by BT engineers for line testing can. They are consumer grade, for one, and they simply do not have the capability of the testing built into them. This would vastly inflate the price of a router, and statistically, the number of lines which DO work well with cheap modems is much higher than the number of lines which are faulty.
More will come to light by next week.... please expand upon this, I'm interested to know what you mean. Oh, and be as technical as you like, I'll only ask what you mean if I don't understand, I'm not proud :)
-
All to be said at the moment is that there is a lot of information to be obtained from the imbedded commands and routines in most of the routers, that can be used as a guide as to aid with difficulties that users of broadband have in many cases, there are a lot who just accept whatever they receive as a working circuit, and a lot who are being "oversold" broadband which they are not using to a level that they believe they are having to pay for. So lets see what can be retrieved from these routers.
I fully accept that results ie attnuation & SNR etc are rather wide in their tolerance as disclosed. However, there are thoise of us who can use the information retrieved to point pretty accuratly as to where a problem may exist. I myself used to diagnose all types of faults over the twisted pair, coax, and waveguide, but I had £10's of thousands of kit for the purpose. BUT you have to be able to use it, and, correctly! even then you have an experience factor built in to many of the resulting figures.
Im making my batch of marinaded Mango chutney at the moment so must go for a stir and taste. I should be able to produce some more next week, as I said but this is a side project to the main route. Till i have a little more time.
Have a great day.
-
>> I am firmly of the opinion that 98% of the problems that users of broadband over the twisted pair are affected by, or are down to "the route" ie from the exchange via the 'E' side cable to the CAB and the 'D' side cable to the dwelling/premises,
But the "twisted pair" is either the D side or the E side anyhow. I'm really not sure what you are trying to get at here.
>> All to be said at the moment is that there is a lot of information to be obtained from the imbedded commands and routines in most of the routers, that can be used as a guide as to aid with difficulties that users of broadband have in many cases, there are a lot who just accept whatever they receive as a working circuit,
Again Im not sure what you are trying to get at here. If you have high attenuation as reported by your router there sure as hell isnt much you or BT can do about it.
As already explained to you on numerous occasions.. the vast amount of poor speeds are either down to
a) User incorrect set up or malware/unsecured networks
b) Contention at the local exchange but bear in mind that we pay for a contended product to make adsl affordable
c) Contention at the ISP - central pipes cost a huge amount of money and to make adsl affordable then users have to share the available bandwith. We dont pay for a 1:1 connection - otherwise the cost would be nearing £1000 per month.
d) Maxdsl will always try and give the user the best speed possible according to the line conditions. If you have a long line - then theres not much that anyone can do about this. adsl frequencies deteriorate over distance. Longer lines are actually more costly than shorter lines.
>> However, there are thoise of us who can use the information retrieved to point pretty accuratly as to where a problem may exist.
There are some of us that can perhaps do so.. but you cant expect Jo Average to be able to.. the aim of sites like this is to help users understand it.
There are many of us on forums just like this who give their time freely without pay to try and help users diagnose where their problems lie. Users like us offer some of the best free advice in the country.
>> I myself used to diagnose all types of faults over the twisted pair, coax, and waveguide, but I had £10's of thousands of kit for the purpose.
Exactly - therefore you cant expect the same from a £25 router.
A router can help tell ATM line faults but no way can it tell your exact routing - do we really need to know that?
I'm sorry but I really am not sure what information you are wanting.
If you are simply wanting router line stats (http://www.kitz.co.uk/adsl/frogstats.htm) - This site was actually one of the first to supply this information to the UK general public for a wide range of routers. In fact this site has often been copied by others (with my permission / acknowledgement).
-
Kitz, to say in a few words. I have no compaint about what you are attempting in some cases, and it must be noted achieving in a lot of cases, to resolve users difficulties. You and your friends on this site are a credit to users. The majority of ISP's do attempt, some more than others to assist in sorting out service problems. However I must take issue with your comment " If you have high attenuation as reported by your router there sure as hell isnt much you or BT can do about it" Well there certainly is, as appart from the cable suppliers BT in one way or another own the twisted copper pairs. If BT want to issue an invoice for a service then they should ensure that the network is maintained, I will not be drawn any further on this point at present, other than to say a number of lines are extremly "ropy" and should have been "uplifted" long ago. If it is known for instance that its own Engineers have been putting in A55's for "uplift" to be carried out, because a cable is known to be "flat" then heads should and must "roll" if it is not completed in reasonable time. The majority of the public are not in a position to understand what is the problem, they simply want a constant useable service if they are expected to pay. For you and other members of this site have to beg and pray for, for instance Interleaving to be switched "on" or "Off" is simply a joke, when it is clearly not being done at request, especially when the customer is informed it will be done. Its such a waste of your time in the continual requests for instance. To not be open with what "Traffic management" is in force on a customers line, without the customer being made aware in real time is I believe an unfair practice. I could go on and on.
I was simply after what information could be obtained that was "hidden" inside the router that maybe of use in diagnosis. Especially that which was not readily available.
Im getting there and will be soon.
Does the above help or does the "mystery" which I dont understand !! deepen into something else that I am not aware of..
-
There appears to be a little confusion as to what chipset is in which Netgear router
Thanks to this site and the assistance given.
I believe it would be of value to put onto your router page the following
Im not saying its going to work in all cases but it has in all that I have looked into.
First telnet the router in question, that is on the site
then:-
at the command line against the #
(you type) cat proc/cpuinfo (then you press enter)
BusyBox v0.61.pre (2007.04.16-10:12+0000) Built-in shell (ash)
Enter 'help' for a list of built-in commands.
# cat proc/cpuinfo
processor : 0
cpu model : MIPS 4KEc V4.8
BogoMIPS : 211.35
wait instruction : no
microsecond timers : yes
extra interrupt vector : yes
hardware watchpoint : yes
VCED exceptions : not available
VCEI exceptions : not available
#
#
This is for the DG834G Version 3 and the cpu is clearly identified
similarily with the DG834N which has the broadcom chipset you receive
BusyBox v0.61.pre (2007.04.16-10:12+0000) Built-in shell (ash)
Enter 'help' for a list of built-in commands.
# cat proc/cpuinfo
system type : 96358GW-16
processor : 0
cpu model : BCM6358 V1.0
BogoMIPS : 299.00
wait instruction : no
microsecond timers : yes
tlb_entries : 32
extra interrupt vector : no
hardware watchpoint : no
VCED exceptions : not available
VCEI exceptions : not available
#
#
I hope this helps against which chipset is in which router.
I have chosen the netgear routers as they are generally known to work better on lines with a greater line attenuation etc.
-
>> I believe it would be of value to put onto your router page the following
That info is already up on the page that I linked you to several days ago which I said you may find useful - under the section busybox shell commands (http://www.kitz.co.uk/tute/voyager_2100_cli.htm)
By combining the cat and proc command we can get certain info from the router.
# cat proc/cpuinfo
system type : RTA770W
processor : 0
cpu model : BCM6345 V0.0
BogoMIPS : 92.97
wait instruction : no
microsecond timers : yes
tlb_entries : 32
extra interrupt vector : no
hardware watchpoint : no
VCED exceptions : not available
VCEI exceptions : not available
I've asked people to use that command before now to find out what chipset is in which router... and thanks to the kind co-operation of several zen users + zen reps with various different routers, its how we established a list of which chips were in which Netgear routers. "Narsty" then also updated the wiki AR7 chipset list with the results of this info.
Its also been previously mentioned in several threads on this forum when trying to find out what chipset is in the netgear routers which has been topical of late in respect of the AR7 issues.