Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Black Sheep]

mine came today also, but I have left it wrapped as I have a line fault currently.

Do we have them before black sheep?

Not only do you have them before us, to all intents and purposes, we are unawares they even exist !! There has been no communication at all that the MK3 is service-ready and available. You have to bear in mind, we are a communications company. 

[les-70]

  A me too!.  I put the mark 3 in place late this morning.  It is too early to give any reliable performance indications, just first impressions. The attainable  and sync speed went up roughly 1 Mb/s and the error rates seem about 30% lower.  This could just be a lucky time of day change so I will report on long term results in day or so.

 I attach the snr/per tone graph just before and after the swap. With the mark 3 the highest frequencies are less attenuated and carry more bits.

The graphs are with the line capped at 64 Mb/s (which can be judged in relation to the lowest recent attainable speed of 73mb/s).  At the time the reported attainable was 76 with an HG612- first swapped to a Billion 8800NL (with latest firmware) which gave 78 and then the with the billion and mark 3 - 79  .  Note that line conditions may have changed during the swap and as noted above a longer test is needed for a live connection.

 

[burakkucat]

In his rush to try to do too much in too short a time yesterday, b*cat forgot to mention that his Mk 3 SSFP also arrived on Saturday. Damaged in the post. 

The supplier was very quick to acknowledge my (eBay) message (complete with photographs), apologised for the mishap and promised to dispatch a replacement to The Cattery on Monday. 

If he does not want me to return the broken one, I will take the PCB out of the case and photograph it alongside a Mk 2 SSFP PCB. 

[NewtronStar]

My gut tells me that most members are more interested in looking whats inside the MKIII than actually testing weather it can deliver the goods on what's it's supposed to do.

I wanna see long term graphs (1 month) to see if it works or fails to deliver 

[les-70]

  A long term graph would only show the state of the connection and most likely some change with time as other lines get services.  A true test would involve comparisons under fixed but realistic conditions.    In the absence of test conditions i tend to rely looking for consistency of changes with swaps in and out. That is how I have evaluated modems and other components.  I will try to alternate the mark 3  with a mark 2 a few times over the next week or two to see if differences seem consistent each time.  Today shows a hopeful gain from one insertion.  If that gain is true I expect a loss on going back to the mark 2 and a repeat gain on returning to the mark 3.  Long term is only correct in the sense that enough time is needed each time to ensure varying situations  are coped with. 

[NewtronStar]

Long term is only correct in the sense that enough time is needed each time to ensure varying situations  are coped with.

Yeah les-70 I like your thinking and another way to judge the difference for each device over time is looking to see the how the error averages look in dslstats 7 days with MKIII and 7 days with MKII that should be a good realistic test 

[NewtronStar]

Not only do you have them before us, to all intents and purposes, we are unawares they even exist !! There has been no communication at all that the MK3 is service-ready and available. You have to bear in mind, we are a communications company. 

If thats the case BS someone is being fraudulent and just changing the logo of the MK2 to MK3 and putting them on fleebay hope thats not the case, now I can unterstand why members need to have a look at the PCB 

[Black Sheep]

Hmm ..... I hadn't thought of that angle ??

TBH, my own thoughts are that they will probably be the genuine article, they just wont have gotten round to let the engineers aware of their availability ?? It's happened before, you wont be surprised to hear!!

[roseway]

For reference, here are pictures of the inside of an SSFP Mk2:

[les-70]

 Unfortunately my master socket is behind a heavy cupboard which needs emptying to move it.  I have now had severe complaints     re leaving the cupboard away from the wall, the room full of the contents, and possibly marking the carpet.  I have therefore done another swap test before putting it all back with the mark 3. I won't be swapping as much as I planned to.

I just briefly went  back to the mark 2 and then returned to the mark 3.  The attainable increase of about 1 Mb/s with the mark 3 occurred again so I am tending to believe it.  Judging from the previously posts bit loading's this improvement may only occur if your attainable is above about 60Mb/s and your using the highest frequency 17a tones.

The error rate over 24 hours with the mark 3 continued,  from a couple of hours, to be about 30% less than the smallest normal 24 hour average, but as suggested by NewtronStar, I am sure a longer test is needed to judge error rates.  I will now do a 7 day test with the Billion and then one with an HG612 as I am used to the error rates with both of those so a rough evaluation may be possible if thunderstorms allow!

[Chrysalis]

personally I think these are genuine, just some guy who got access to stock or whatever selling a bunch off

[NewtronStar]

personally I think these are genuine, just some guy who got access to stock or whatever selling a bunch off

Its Just me being paranoid again CHRY if the MK3 is able to reduce the errors by say 8% then I'll puchase one, but my findings our if the user has a low to medium exisiting HR fault then the MK2 or 3 will not help to much.

I will now do a 7 day test with the Billion and then one with an HG612 as I am used to the error rates with both of those so a rough evaluation may be possible if thunderstorms allow!

Even in our home for example a pull cord switch for the light in the bathroom looks like a T/Strike on my line stats 20 to 25 CRC's it only happens when its pulled to the on position never to the off position.

I guess in most UK homes there will always be mains switch arcing, Can't see the MK3 being any different than the MK2 when it incounters mains switch arcing unless your home is a new build.

I think most CRC and ES errors come from inside the EU premises via the 240 mains supply, so my thinking you would be better of with an expensive mains Conditioning Filter to your Modem and Router.

[les-70]

   I managed to get a stay of execution of one day prior to fully putting the cupboard back.  I also try to never have more than 3 re-syncs per day when testing as that does not seem to have ever upset the DLM.  I therefore decided to use the resyncs on a HG612 instead of the Billion for 24 hours. The mark 3 gave a smaller speed increase with the HG612 of about 0.5 Mb/s and over 24 hours errors were about 20% less than the usual good day average.  I would need to repeat this a few times to be fully convinced but it looks hopeful. 

  On its own the Billion 8800NL with the latest 039 dsl driver firmware seems to make better use of the higher frequencies and as the mark 3 seemed to have less attenuation at the highest frequencies it may be a favorable combination.  This may be consistent with switching back to the Billion with the mark 3 place giving an attainable increase of 3.5Mb/s c.f the HG612 (which may or may not be significant). A just under 3Mb/s increase was previously seen with the mark 2. 

  I have now definitely started a long test with the Billion.

[oddlegs]

I note that the SSFP Mk2 faceplate has been apparently replaced by the Mk3, but does anyone know what was supposed to be wrong with the Mk2 ?
I am on an EO line of around 3.5 km with a longish drop wire located the best part of 1000 feet above sea level on the top of a hill.
I have always suffered from a large variation in day/night noise margin. This meant that my line would lose sync at regular intervals.
I partially got around this with the excellent Billion 7800n with altered default noise margin and the Phyr option but my usable sync peaked at around 7000 kb/s.
About 2 months ago I replaced my V10 faceplate with a Mk2 and the difference was amazing. My sync is now well above 8000 kb/s and the noise margin hardly drops at all overnight. Speed tests are above 7200 kb/s.
I have been connected now without sync drops for several weeks.
Strangely enough, some time ago I tried a RF3 filter in the modem adsl lead and that made things much worse rather than better.
If the only difference with the Mk2 over the older faceplate is the integrated RF3 common mode choke why did it not work before ?
I shall stick with the Mk2 for now I think.

[roseway]

I don't know the answer to your question, but I think it's most probable that the difference between the Mk 2 and Mk 3 is only in the higher VDSL2 frequencies. There was no recognisable difference when I made the change on my current ADSL2+ connection.