In reply to a post by RobertoS:

For the second time, max figures and actual figures are related. With some DLM adjustments of the download they frequently move dramatically in opposite directions on a re-sync. I know you are bothered about the upload, but if that has happened it gives a strong indication of the type of change the DLM has made, or whether it has made no change at all.

That in turn gives a clue to what has happened to the upload. If you know the make of FTTC cabinet that can be useful as well.

ECI Cab with a HH5 Type B, i have a Type A but don't really want to put it back on. Switching between a MKIII and a MKII filter didn't do owt and i have run a quiet line test and the line is quiet so i am assuming everything is fine with the physical condition of the line and i reiterate actual my speeds are spot on http://www.speedtest.net/my-result/5399684714

6. Data Rate: 19993 / 79999
7. Maximum Data Rate: 21942 / 110149
8. Noise Margin: 6.6 / 14.2
9. Line Attenuation: 5.1 / 6.4

no idea then?

Sorry, I missed your previous post.

I think the drop in attainable could be just more of the same, more crosstalk. Your actual in both directions is still the maximum for your product.

Re your two speed tests, I hadn't looked at your early one using thinkbroadband. That Avg TBB single thread line suggests throttling or congestion somewhere. It should be very close to the multi-thread one. Speedtest.net and BT Wholesale tests only use multi-thread, and speedtest.net drops some of the slower periods in the test as well.

Real life speed can be affected by poor single-thread throughput, though whether most people would notice it is another matter.

I suggest you see if the tbb test is still doing the same, and if so start a new thread about that. Include in it your current actual stats and the test you give above, as well as whatever the new one shows.

Thank you for your reply, although if the dropping upload attainable was down to crosstalk wouldn't the max download attainable have dropped as well and not increased by 8Mbit?

As i have pointed out a couple of times i cannot moan about the speeds, but if the upload takes another dip on the next reboot i have problems.

If interleaving gets turned on, on higher attenuation lines the downstream sync drops by several Mbps, 10Mbps fall being typical, and the downstream max attainable rises by a similar amount. Now I have to guess that with your very low attenuation interleaving has been turned on but 79999 actual is still possible.

Has your base latency gone up by 8ms? First step interleaving causes that, higher levels 16ms or more addition.

It's not easy to be sure what is going on, as we don't often see interleaving on upstream and we need full stats to know if that is also in place. From what I remember this opposite movement of actual and attainable doesn't happen on upstream, but I could easily be wrong.

In reply to a post by RobertoS:

If interleaving gets turned on, on higher attenuation lines the downstream sync drops by several Mbps, 10Mbps fall being typical, and the downstream max attainable rises by a similar amount. Now I have to guess that with your very low attenuation interleaving has been turned on but 79999 actual is still possible.

Has your base latency gone up by 8ms? First step interleaving causes that, higher levels 16ms or more addition.

It's not easy to be sure what is going on, as we don't often see interleaving on upstream and we need full stats to know if that is also in place. From what I remember this opposite movement of actual and attainable doesn't happen on upstream, but I could easily be wrong.

When i got infinity in 2011 the fastest ping i got on speedtest was 15ms, 5 years later it is still 15ms so i don't know.

It was suggested on another forum that one possible cause is the upstream power may have been reduced, i am not very technical but that seemed as plausible as owt else i have heard.

This all happened shortly after a 14 day BT resync or whatever it does every 2 weeks. Less than 24 hours in to the next 14 day period and the router resyned between 2 and 3 in the morning and the speeds are what i have posted.

I am tempted to put my Type A back on seeing as i am on an ECI cab just to see if it makes a difference.

I put the Type A on and not a lot happened

Type A

6. Data rate: 19993 / 79906
7. Maximum data rate: 22158 / 112876
8. Noise margin: 6.4 / 14.5
9. Line attenuation: 5.2 / 5.3
10. Signal attenuation: 5.1 / 5.3

Type B

6. Data Rate: 19993 / 79999
7. Maximum Data Rate: 21986 / 110114
8. Noise Margin: 6.7 / 14.1
9. Line Attenuation: 5.1 / 6.4

the alternative TBB speedtest gives a different picture regarding the speed if i have read it right

http://www.thinkbroadband.com/speedtest/button/14665...

Put the Type B back on, the 5GHz side has better range than my Type A.

Dunno if this helps at all

current linestatus
state: UP
mode: VDSL2
mod type: Profile 17a
snr margin down: 141
snr margin up: 66
latn down: 64
latn up: 51
satn down: 0
satn up: 0
output power down: %2D107
output power up: 134
rate down: 79999
rate up: 19998
attainable rate down: 110091000
attainable rate up: 22014480
chantype: fast

Well god knows what that was all about, it did a resync at 05:20:55 and its all ok again

6. Data Rate: 20000 / 79999
7. Maximum Data Rate: 34192 / 109675
8. Noise Margin: 15.6 / 14.0
9. Line Attenuation: 5.2 / 6.4