I have used the commands below, but the router is pretty adamant that it's going to use 15db, despite the commands being accepted and the line re-synch'd.

Could this be because I'm so close to the exchange that a higher SNR down is usual?

In reply to a post by DSLPete:

Could this be because I'm so close to the exchange that a higher SNR down is usual?

34.5db is not a short line. My 24.5db line is reported to be 1700m, and that must be wrong; given my flat is behind some 1920's houses, and the exchange is at the end of the road of 1920s build - the BT cables must run down that road, then into mine, which is a 1km distance.

does the sync speed stay exactly the same? could be your sync speed is capped.

or it also could be the setting doesnt work with your provider, or even it doesnt work on g.dmt. (I use it on adsl2)

here is my line using the 70 setting.

Parameters
DSP Firmware Version E.25.41.55 A
Connected true
Operational Mode G.Dmt.Bis
Annex Type AnnexA
Upstream 1024000
Downstream 5456620
Elapsed Time 1 day 0 hr 20 min 3 sec
SNR Margin(Upstream) 7.0 dB
SNR Margin(Downstream) 7.25 dB
Line Attenuation(Upstream) 28.0 dB
Line Attenuation(Downstream) 53.5 dB
CRC Errors(Upstream) 0
CRC Errors(Downstream) 0
Latency(Upstream) Fast
Latency(Downstream) Fast

the command I used was slightly different because of the modulation.

port a1 set BisAForceSNRMarginDn 70

There is AnnexA which I assume is for adsl1
BisA adsl2 adsl2+
BisM adsl2 adsl2+ using annexM

Edited by Chrysalis (Sun 21-Mar-10 22:37:55)

here I am going to reveal some testing I did on the router with the bitloading

bear in mind this is a lot of numbers.

so the way this is layed out is the number on the left is the bin # of the first value, then the 2nd value is one bin higher and so on, so if it says bin32 then the first value is for bin32 and then the next bin 33 and so on. 15 is the highest possible bitloading per bin. The numbers drop off as the bin's go higher as that is on the weaker part of the line, but also notice the first few bin's are not maximum due to crosstalk with the upstream. These will show the affect that reducing the upstream power has on these bin's.

so this below is auto coding gain (conservative) with default full upstream power on my line. I cannot remember if the noise margin was 6db or 7db on these tests.

Bin 32 = 5 6 6 6 7 7 8 8 8 9 9 10 10 11 9 11
Bin 48 = 11 11 11 12 12 12 12 12 2 12 12 12 12 11 11 11
Bin 64 = 12 11 11 11 11 11 11 11 12 12 11 12 11 12 12 12
Bin 80 = 12 12 12 11 11 11 11 11 11 11 11 11 11 11 11 11
Bin 96 = 11 11 10 11 11 11 11 10 10 10 10 10 10 10 10 10
Bin 112 = 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10
Bin 128 = 10 10 10 10 10 10 10 10 10 9 9 9 9 9 9 9
Bin 144 = 9 9 8 8 9 8 8 8 7 7 7 7 6 5 5 3
Bin 160 = 1 0 2 4 5 5 6 6 6 7 7 6 7 7 7 7
Bin 176 = 7 7 6 6 6 7 6 6 6 6 5 5 5 5 4 3
Bin 192 = 1 0 0 0 0 2 2 2 2 2 2 2 0 0 2 2
Bin 208 = 2 2 0 0 2 3 3 3 5 4 4 0 3 3 4 4
Bin 224 = 3 3 2 3 3 3 0 1 0 2 2 2 2 3 2 2
Bin 240 = 0 2 1 0 0 0 0 0 0 0 0 0 0 0 0 0

now this one is the same with auto coding gain but with the upstream power reduced by 3db, so using the tx_attenuation bis3_0db setting. See the improvement on the 1st line, its an important improvement since those tones wont drop at night they stay at full strength on the strong part of line.

Bin 32 = 8 8 9 9 10 10 11 11 11 12 12 12 12 12 9 12
Bin 48 = 12 2 12 12 12 12 12 12 12 12 12 12 12 12 11 12
Bin 64 = 12 12 11 11 11 11 11 11 11 11 11 12 11 12 12 12
Bin 80 = 12 12 12 11 11 11 11 11 11 11 11 11 11 10 11 11
Bin 96 = 11 11 10 11 11 11 11 11 10 10 10 10 10 10 10 10
Bin 112 = 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 10
Bin 128 = 10 10 10 10 10 10 10 9 10 9 9 9 9 9 9 9
Bin 144 = 9 9 8 9 8 8 8 8 7 7 7 6 6 5 4 3
Bin 160 = 1 0 2 4 5 5 6 7 7 7 7 7 7 7 7 7
Bin 176 = 7 7 6 6 6 7 7 6 5 6 5 5 6 5 4 4
Bin 192 = 2 0 0 0 1 3 3 3 3 3 2 2 1 0 2 2
Bin 208 = 2 3 0 0 1 3 3 3 4 3 4 0 3 3 4 2
Bin 224 = 3 2 3 3 3 3 1 1 0 2 2 2 2 3 2 2
Bin 240 = 2 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0

increment +1.0 coding gain reduced upstream power by 3db

Bin 32 = 8 9 9 9 10 10 11 11 12 12 12 12 12 12 9 12
Bin 48 = 12 12 2 12 12 12 12 12 12 12 12 12 12 12 12 12
Bin 64 = 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12
Bin 80 = 12 12 12 12 12 12 12 12 11 11 11 11 11 11 11 11
Bin 96 = 11 11 11 11 11 11 11 11 11 11 11 11 11 11 11 11
Bin 112 = 11 11 10 10 10 10 10 10 10 10 10 11 10 10 10 10
Bin 128 = 10 10 10 10 10 10 10 10 10 10 10 10 9 9 9 9
Bin 144 = 9 9 8 9 9 9 8 8 8 7 7 7 6 5 5 3
Bin 160 = 0 0 3 4 5 6 6 7 7 7 8 7 8 7 8 8
Bin 176 = 7 7 7 7 7 7 7 7 5 7 6 5 6 6 5 4
Bin 192 = 3 0 0 0 2 3 3 3 3 3 3 2 0 0 3 3
Bin 208 = 4 4 3 0 4 5 4 4 5 4 4 0 4 4 4 3
Bin 224 = 3 3 4 4 3 3 1 1 0 2 1 2 2 3 2 3
Bin 240 = 0 2 1 0 0 0 0 0 0 0 0 0 0 0 0 0

as above but after gone dark and lost 1000kbit speed

Bin 32 = 8 9 9 10 10 10 11 11 12 12 12 12 12 12 9 12
Bin 48 = 12 12 13 12 12 12 12 2 12 12 12 12 12 12 12 12
Bin 64 = 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12
Bin 80 = 12 12 12 11 12 12 12 12 12 12 11 11 11 11 11 11
Bin 96 = 11 11 11 11 11 11 11 11 11 11 11 11 11 11 11 11
Bin 112 = 11 10 10 10 10 10 10 10 10 10 10 10 10 10 10 9
Bin 128 = 10 10 10 10 10 10 10 10 10 10 9 9 7 9 8 9
Bin 144 = 6 9 8 8 7 8 8 8 7 7 6 6 6 5 4 2
Bin 160 = 0 0 0 3 4 0 5 6 6 5 7 0 6 5 5 0
Bin 176 = 4 4 2 0 0 2 4 0 0 0 0 4 0 4 0 0

Seems I didnt record the tests that got tones loaded to 15, but I think I had to reduce the noise margin to get tones to 15 bits.

here is the SRA problem I refer to, I plan to flash the firmware after it next happens. I marked with a black diamond shape the time on the graph where SRA stops working. You see it works again after I restart the sync.

noise margin
sync speed

Edited by Chrysalis (Mon 22-Mar-10 00:11:20)

ok just synced up on this firmware. I dont see any obvious changes to the dsl stuff, so I am hoping it has stability fixes.

DSP Firmware Version E.25.41.64 A
Connected true
Operational Mode G.Dmt.Bis
Annex Type AnnexA
Upstream 1024000
Downstream 6472489
Elapsed Time 0 day 0 hr 0 min 38 sec
SNR Margin(Upstream) 7.5 dB
SNR Margin(Downstream) 7.05 dB
Line Attenuation(Upstream) 28.0 dB
Line Attenuation(Downstream) 51.5 dB
CRC Errors(Upstream) 0
CRC Errors(Downstream) 0
Latency(Upstream) Fast
Latency(Downstream) Fast

Edited by Chrysalis (Mon 22-Mar-10 17:31:11)

ok so far no actual SRA freeze up, but the router resynced earlier today. That itself is unusual and bad since with SRA I dont expect my sync to drop at all.

looking at the tbb quality graph I see a 'lot' more red than usual which either means today my line has been very noisy (this red even occured after the resync which was also at higher speed) or that the new firmware is not handling noise as well.

Compared to the .55 version I am running at the same dsl settings which are inc1_0 coding gain (higher than default but not very agressive) and snrm of 7db (more conservative than the 6db isp setting). As well as -3db of upstream power and disabled upstream SRA.

so there was a dip in sync speed at 8am, probably noise burst, then no more dips until the router lost sync, however even tho the snrm was always around 7db there was errors occuring due to background noise on my line, then when the speed jumps up at just after 1pm is the router resyncing, after that various dips show because of more noise bursts. Then a gap in the graph where I dropped the sync to change a setting.

Speed graph
SNRM graph

Edited by Chrysalis (Tue 23-Mar-10 18:00:33)

I have moth balled my 7402GXL for now and are following your threads and hope you work out theses new chipset 7402 .
Even thou i got a spare 7402r2 and 7402VGP i would like to use the new feature set you have discovered on the new 7402

well after changing coding gain to 6 things seem quite a bit better with the same sort of synch speed as 1_0.

billion also replied to me today with a url to my requested b1.64 firmware, so I think things are looking up on this router now, I wont flash tho until my next problem occurs.

If anyone wants the url msg me.

todays graphs.

Speed
Noise Margin

Edited by Chrysalis (Thu 25-Mar-10 19:42:29)

I been using the s6.b1.64 firmware now, works fine just as good as the older .64 firmware.

Also I now configure my line differently and the result pleasantly surprised me.

1 - I reduced coding gain to auto.
2 - I decreased target noise margin to 4db.
3 - I increased rein checking to the max 120pps.

Result is a sync that hovers between about 6-6.15mbit second and crc errors are suprisingly much improved.

The bit loading on this setup is that the higher(weaker) tones are used much less (meaning less night time speed loss) and the stronger tones have higher bitloading around the 13/14 mark instead of 11/12. Coding gain appears to get its extra speed by loading up bits on the weaker tones.

On my older setting sync would start initially at around 6.5-6.7mbit (higher if coding gain 7) but at the first night this would drop to well below 6mbit and then only recover to maybe just over 6mbit. So daytime speeds were perhaps a little less but night time speeds were more less. Since there is now less variance SRA has less work to do.

I have also discovered what one of the unknown settings does, and the setting allows you to disable adsl2/+ low power modes, it is also linked to SRA functions, ie. SRA supposedbly works better without the low power modes. I havent touched this setting tho as I am all for reducing crosstalk which is what the low power modes do. In these graphs, the first gap is me changing the settings, flashing firmware etc. There is a 2nd small gap and that was not a sync problem, the graphing server had a connectivity issue to the uk for a short period.

Speed - auto coding gain, reduced upstream power, 4db margin
Noise Margin - auto coding gain, reduced upstream power, 4db margin

DSP Firmware Version E.25.41.64 A
Connected true
Operational Mode G.Dmt.Bis
Annex Type AnnexA
Upstream 1024000
Downstream 6032048
Elapsed Time 0 day 20 hr 50 min 54 sec
SNR Margin(Upstream) 7.5 dB
SNR Margin(Downstream) 4.45 dB
Line Attenuation(Upstream) 28.0 dB
Line Attenuation(Downstream) 50.5 dB
CRC Errors(Upstream) 0
CRC Errors(Downstream) 0
Latency(Upstream) Fast
Latency(Downstream) Fast

Edited by Chrysalis (Sun 28-Mar-10 02:51:05)