Line Stats Comparison

First post on here so please be nice with me...

I'm on PlusNet and happy with the service but like most people hope for a little more speed on my 80/20 service.
Below are two snipets of stats. First from my ASUS DSL-AC68U working as an all in one solution and the second from an unlocked HG612 using the ASUS as a router only;

ASUS;

SNRMarginDown=6.1 dB
AttenDown=10.4 dB
SNRMarginUp=6.1 dB
AttenUp=0.1 dB
DataRateDown=60392 kbps
PathModeDown=FastPath
DataRateUp=16456 kbps
PathModeUp=FastPath
FECDown=716
FECUp=72
CRCDown=84
CRCUp=15
HECDown=0
HECUp=0
ADSLUpTime= 3:39, 53 secs
AttainUp=16527
AttainDown=75088

HG612;

Max: Upstream rate = 17046 Kbps, Downstream rate = 57924 Kbps
Bearer: 0, Upstream rate = 16983 Kbps, Downstream rate = 58396 Kbps
Down Up
SNR (dB): 6.1 6.0
Attn(dB): 17.3 0.0
Pwr(dBm): 13.2 6.9
Total time = 1 days 12 hours 21 min 19 sec
FEC: 0 1787
CRC: 1549 290
ES: 1217 279
SES: 0 0

Now the bit that gives me hope of more speed is the AttainDown=75088 on the ASUS. Why would this be so different to the 57924 figure on the HG612? I realise the actuals are similar(ish) 60392 vs 57924 but if the ASUS is to be believed surely I should be able to sync at a higher speed?

Just for info I'm on an ECI cab (also have an ECI/r modem) and have just installed a MK3 faceplate to replace the old MK1.

I appreciate your thoughts?

TIA, Mick

It is not known how the different VDSL2 modems handle the power masks, i.e. three different ways that VDSL2 is configured for the Openreach network and this means in some cases the estimates of maximum potential speed may be over stated.

Power masks exist to protect the performance of ADSL2+ so if ADSL/ADSL2+ was turned off one day higher speeds from VDSL2 are possible.

Thanks. From what I've read the HG612 info is considered to be quite accurate so it's likely then that the ASUS is overstating the MAX figure?

Also I believe that the ECI/r modem that I have will connect slightly faster (on an ECI cab) than the HG612? But I won't get any stats and I don't want to mess with it..!

Can you install DSLStats and then take a couple of snapshot images of the Tones graph with SNR overlay and of the QLN graph It would be interesting to compare the two versions to see where the differences actually are.

I've got DSLStats running now but thought it would only work with HG612 not the ASUS.?

Also (a potentially dumb question) how do I insert the graphs into a post?

You cannot upload to a post - just host them on a suitable site - photobucket, flikr &c and post the links.

OK. Is this what you mean? These are from the HG612.

https://flic.kr/p/zeiut5
https://flic.kr/p/ztA5xu

Not sure I can get them from the ASUS though..!!

Almost ... under the Tones tab there are two further tabs, the Bit loading would be useful too.

If you can get the ASUS ones, a comparison may shed light on the differences.

Okie Dokie.

Here is the other https://flic.kr/p/zuRDDj

I'll have a try later and see if I can get anything out of the ASUS.

Nothing obvious there!

Also worth a small tweak to settings:

Select the <Config> tab, then <Items to Monitor>, Under Others, select <SNR per tone>, <Show as separate graph> and <Include with bitloading> That will overlay Bitloading with SNR.

Thanks for that it now looks like this https://flic.kr/p/ztFFDm

As for gaining access to the ASUS just tried with no luck. DSLStats isn't compatible with it. Even tried using a few different models just in case but no joy it just says "no data returned".

Again, nothing obvious there all the way up to the top tone.


A pity about the ASUS, have you though about an email to the DSLStats team asking if they could look at including it - there is at least one other ASUS there already.

Thanks I'll try emailing them there's no harm in asking. I tried the other ASUS as well but no joy.

Probably not worth emailing them as the DSLStats website states "It works with DSL routers using Broadcom chipsets which have a fully functional telnet interface with access to the Broadcom CLI" and according to the Kitz website it's a MediaTek chipset in the ASUS.
Oh well..!!

In reply to a post by DamBuilder:

Thanks. From what I've read the HG612 info is considered to be quite accurate so it's likely then that the ASUS is overstating the MAX figure?

I mentioned on another post about a quirk in the calculation of the maximum attainable speeds - at least in broadcom chipsets, like the HG612.

The modem will normally sync with an actual speed that matches the attainable speed, at least if that speed is less than the purchased package limit. If old-style DLM intervenes, the sync speed drops by around 10% as interleaving and error correction are turned on. However, the max attainable value would often go up by as much as the actual speed went down. If DLM removes the intervention, both values return to normal.

What does this have to do with your ASUS?

Well, I notice that your HG612 is reporting a downstream FEC count of zero, alongside non-zero CRC and ES counts. That tells us that old-style DLM has not intervened. It also tells us that your line isn't currently subject to new-style G.INP either.

But the ASUS statistics *do* report a non-zero FEC value. That tells us that either old-style DLM intervention has happened, or that G.INP has been activated. You mention an ECI cabinet - if you're on one of those, then you don't have G.INP.

With old-style DLM intervention in place, you are probably running at a line speed reduced by 10-20%. If the ASUS follows the same quirk, then the maximum will be elevated by the same amount.

However, your ASUS is actually faster. That might show that the ASUS can hold a faster sync speed (better spec of receiver?), but leaves it subject to more errors ... resulting in a DLM slowdown.

One other thing I notice is that the reported attenuation is very different.

A value of 17dB from the HG612 probably equates to around 400m of 0.5mm copper; your speeds are on the slightly low side for this distance, IMO.

A value of 10dB from the ASUS is rather different ... suggesting it calculates very differently.

Thanks for that it's very interesting.

When I first went live with PN (using the ASUS as modem/router) interleaving kicked in after 4 days then went back to Fastpath on the up after another 2 days and finally Fastpath on the down 2 more days after that. About a week ago I switched to the HG612 and set the ASUS as router only.

Yes I'm definitely on an ECI cab and as the crow flies I'm 300 metres away but following the roads 350 metres so your estimate is not far off.

Well after looking in the ASUS GUI I noticed a couple of graphs (link below) and I've included a full set of current stats; The AttenDown=10.6 dB is def lower than the HG612..!!

https://flic.kr/p/zy66Ks

outDiscards=3589
inDiscards=16
outBytes=564759043
inBytes=4055801771
outPkts=3403951
inPkts=6552071
fwVer= FwVer:5.5.1.131_B_A60901 HwVer:T14.F7_0.2
lineState=up
Opmode=ITU G.993.2(VDSL2)
SNRMarginDown=6.1 dB
AttenDown=10.6 dB
SNRMarginUp=6.1 dB
AttenUp=0.1 dB
DataRateDown=60195 kbps
PathModeDown=FastPath
DataRateUp=17207 kbps
PathModeUp=FastPath
WanListMode=1
FECDown=708
FECUp=146
CRCDown=142
CRCUp=28
HECDown=0
HECUp=0
ADSLUpTime= 8:25, 42 secs
ADSLActiveTime=0 min, 20 secs
PowerDown=13.2 dbm
PowerUp=6.4 dbm
ATURID=26005443434e0000
ATUCID=b5004946544eb204
FarEndVendorID=Infineon(Lantiq)
AttainUp=17302
AttainDown=76864
ShowtimeStart=20
TotalStart=60
ATURANSIRev=0
ATUCANSIRev=0
ATURANSIStd=0
ATUCANSIStd=0
InterleaveDepth=1
AdslStandard=VDSL2
AdslType=ANNEX_B
mtenStandard=G.dmt.bisplus (Annex L)