Kitz ADSL Broadband Information
adsl spacer  
Support this site
Home Broadband ISPs Tech Routers Wiki Forum
 
     
   Compare ISP   Rate your ISP
   Glossary   Glossary
 
Please login or register.

Login with username, password and session length
Advanced search  

News:

Author Topic: Curious behaviour observed with DG834GT  (Read 2387 times)

sevenlayermuddle

  • Helpful
  • Addicted Kitizen
  • *
  • Posts: 5369
Curious behaviour observed with DG834GT
« on: February 18, 2009, 03:56:23 PM »

I've been using the DG834GT since mid-december and am absolutely delighted with it.  Using Netgear fw, and CLI (non GUI) tweaking.    However, I've noticed one curiousity that maybe of vague interest.  Don't anybody give this high priority as it's not a problem, just an observation...

Let's say I connect at 2pm on a dry Tuesday.  I've CLI tweaked the snr to 50, which results in an initial SNR  of target -3dB.  My target's 15, so I connect at 12.  It then rests at about 12dB until evening then, over the following 24 hours, the SNR dips overnight and then recovers during the ealry hours of daylight.  No surprises there, I know.

But here's the thing... although the SNR recovers again in daylight, it only gets back to about 10dB or so.  Thereafter each day follows the same general routine, dropping from 10dB down (to whatever, it's irrelevant to my point) at night, then back to 10dB when times are good, but it never seems to get back to the 12dB that it reported for the first hour or two after connecting.  It's very much as if the DG834GT's measurement of actual SNR grows slightly more pessimistic over the first 24 hours or so, before settling down to a more consistent story.

The process does seem to be repeatable.  I've just this moment done a manual reconnect at the same time of day as a few days ago.  It had been resporting reporting a bit over 10dB before I did so.  I reconnected at a similar speed, but a it reported an SNR improvement of about 2dB.  It's still saying 12, but I'm confident that by this time tomorrow it'll be reporting 10dB again, and the next day, and the next...

I considered the possibility that biswapping might not be happening, but I've discounted that by taking screen snapshots of the stats a few hours apart and they do change.  Still I wonder whether, despite the fact that bit-swapping does take place, could it be slightly sub-optimal?

Anyway, as I said, it's a great router and I don't want anybody to take this as any kind of criticism of it.
Logged

kitz

  • Administrator
  • Senior Kitizen
  • *
  • Posts: 33888
  • Trinity: Most guys do.
    • http://www.kitz.co.uk
Re: Curious behaviour observed with DG834GT
« Reply #1 on: February 18, 2009, 11:05:52 PM »

When I read your post, at first my mind also seemed to wander down the bit swapping track.. then I realised that you were meaning it drops once, recovers, but not as much as it should... but it in effect only does this the once.

for example? 12dB > 8dB > 10dB > 8dB > 10dB > 8dB > 10dB

I wonder if its somehow trying to compensate for the -3dB of Target SNR?... or thats what it appears?   
I honestly dont know - even though I tweaked my GT I didnt see it display that behaviour nor have I seen it elsewhere... and I cant think what could be causing it.

Non Full recovery of SNRM is sometimes something to do with bitswapping... and it could be the firmware.
Logged
Please do not PM me with queries for broadband help as I may not be able to respond.
-----
How to get your router line stats :: ADSL Exchange Checker

sevenlayermuddle

  • Helpful
  • Addicted Kitizen
  • *
  • Posts: 5369
Re: Curious behaviour observed with DG834GT
« Reply #2 on: February 19, 2009, 12:01:40 AM »

for example? 12dB > 8dB > 10dB > 8dB > 10dB > 8dB > 10dB

Yes, that's right, explains it much better than I did.  Why did I use all these words when a simple arithmetic sequence is so much more descriptive :-\

I'm pretty sure it happens regardless of SNR tweaking, but maybe I'll try it untweaked for a day or to for the sake of experiment. 

I did actually wonder... I've heard reports of that router hanging onto connection down to almost inexplicably low margins.  If it were the case that the quoted margin was actually pessimistic, then it might explain that inexplicability.  However, I suspect you folks with routerstats and all that good stuff would notice quickly if your 834Gts did the same, so I guess it's not likely.

Other than a characteristic of my specirfic fw version, my money's maybe on the possibility that it's a side-effect of bit-swapping.  Maybe the characteristics of my line are such that  bit-swapping is excercised more (or less) than usual?

Anyway thanks for your comments, there's no need to dwell upon it.



Logged

kitz

  • Administrator
  • Senior Kitizen
  • *
  • Posts: 33888
  • Trinity: Most guys do.
    • http://www.kitz.co.uk
Re: Curious behaviour observed with DG834GT
« Reply #3 on: February 19, 2009, 12:36:18 AM »

>> but maybe I'll try it untweaked for a day or to for the sake of experiment.

Would be interesting to know just out of curiosity..   but then again theres the stupid DLM which could leave you with a lower IPprofile for a few days :(

>> I've heard reports of that router hanging onto connection down to almost inexplicably low margins

It does seem to - sometimes even going into negative figures and still hanging on in there.



Logged
Please do not PM me with queries for broadband help as I may not be able to respond.
-----
How to get your router line stats :: ADSL Exchange Checker

orainsear

  • Reg Member
  • ***
  • Posts: 635
Re: Curious behaviour observed with DG834GT
« Reply #4 on: February 20, 2009, 04:50:15 PM »

I'm on ADSL Max and see exactly the same pattern of higher resync SNR margin which then drops to a lower maximum value, usually a couple of dB lower.  It seems to happen with all modems/routers that I have tried and I put it down to bitswapping.
Logged

sevenlayermuddle

  • Helpful
  • Addicted Kitizen
  • *
  • Posts: 5369
Re: Curious behaviour observed with DG834GT
« Reply #5 on: February 20, 2009, 05:25:36 PM »

Interesting.

The only other router I used for long enough to have noticed it was a rather old and painfully slow, but solidly built, 3Com Officeconnect.  I don't think that did it, but a lot's probably changed in router technology since that got built.

My hunch is that you'd be right about it being related to bit-swapping.
Logged
 

anything