Print

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

[banger]

After I was initially connected G.INP did not kick in but Banding did. eg I was banded before G.INP was applied. Surely it should be the other way round, G.INP applied before possible banding. I have seen someone else with this on another forum and it happened exactly the same way.

As it happens G.INP gave me another 10m sync and banding was only removed by an engineer reset, a very helpful engineer at that.

After banding had been applied G.INP was applied which took my SNR margin to about 9dB and no amount of reboots of the modem would shift sync from 44m @ 9dB. It took a test engineer to remove banding.

What do you think?

[NewtronStar]

Can only give you my experience with G.INP once the line becomes unstable due to 16 resyncs a day the DLM will turn off G.INP and your forced onto Interleaving and if the resyncs continue banding will then come into play, So from this you must remember that G.INP is just another DLM tool to lower the errors yet Interleaving is still the default by the DLM when things go very wrong.

[WWWombat]

After I was initially connected G.INP did not kick in but Banding did. eg I was banded before G.INP was applied. Surely it should be the other way round,

I agree entirely. Not only does DLM employ banding too readily, DLM also seems to stick banding in place forever. For some people, at least. And we've seen cases where banding gets improved in steps ... but never quite gets taken away fully.

In fact, in olden days, DLM employed banding very rarely. The behaviour you describe has only really been happening over the last year.

Around springtime of the last 3 (?) years, it seems that DLM has had an overhaul related to G.INP in some form - with some changes related to coping with devices that cannot handle G.INP very well, and some changes related to the DLM patent dispute with Assia.

It seems to me that last year's update got banding wrong. I hope it changes this year.

[nivek1612]

Banger

Yes I reported the same on another forum. Think your referring to me as we exchanged posts

It does seem that DLM will try banding before G.Inp which is a bit strange. However it could at least remove banding once g.inp is applied an restart the clock for monitoring

Engineer visit and reset and g.inp back on after just 48 hours means I'm stable and 10MB better off on sync

Trouble is most people don't have understanding isps who won't organise a BTO visit if the line is in the bands defined by the checker