Yeah I'd managed that part but it didn't seem to be doing NAT!

I went back to 5.x and then forward to 6 again and now it works fine?!

I can report your SNR/Modem Option table seems to be correct.

For eg. I have a 6db Noise margin profile and using modem option 40 gives me 4.5db = -1.5db

As for the 536 vs Belkin 7633 I can report there is pretty much nothing in it so far, although it appears the Speedtouch is producing less errors. Sync speeds vary a little but they are +- 3% of each other at default settings.

If this will hold onto 4.5db where the Belkin couldn't, I'll be up another 512k profile step!

Thanks for the effort by the way, appreciated

DMT 7.11 now with sound effects.

I think because it's not a proper UK version of the firmware it might have a few issues, nothing that bothers me tho.

Edited by JinxterX2 (Tue 23-May-06 21:31:12)

Cool

In reply to:

---

If this will hold onto 4.5db where the Belkin couldn't, I'll be up another 512k profile step!

---

You want to monitor SNR fluctuation per tone in DMT, if there's too much negative flux across many tones (especially the mid-range to high) you know your sync rate is too high and will most likely fail. That's what I've found on my line anyway.

Don't be too ambitious on your SNR targets to start off with.

Edited by JinxterX2 (Tue 23-May-06 22:09:16)

I've already done most of the messing with my Belkin 7633 and this proves to be pretty much the same beast.

I can run a low of about 5db at night, which equates to about 7db in the day. I tried lower but it looks like the 536 cannot hang on either during the late night noise period.

Just lost about 1db instantly, so I switched back to the Belkin and that was down 1db too. Very similar performance!

Have to decide which one to keep now, I like DMT tool but the Belkin has a much much faster web interface (plain html vs some cgi on the Speedtouch)

I should not be greedy, I see you have to run a much higher SNR to keep it stable which is a pity
I've had two weeks uptime at 5-7db on the Belkin, so my line is not too bad I guess (57db Att - 4500-4700kbps)

Well it seems my max sync stable rate is 5600Kbps, any higher (even 256kbits more) and it craps out during the noise period, which is a pity because it means I can't get the Data Rate I wanted (5696Kbps).

The SNR per tone fluctuation readings in DMT are really interesting when you try different values for modemoption00, great program

Doesn't the Belkin 7633 use the BCM6348 chipset? If so it's not suprising it performs similar to the Speedtouch.

Edited by JinxterX2 (Wed 24-May-06 04:58:43)

Bit of messing about with the values (and lots of retrains at various times of day/night ) has yielded a sync of 5696Kbps using modemoption00=5d, which will hopefully keep the target aimed at between 15-15.5db (during the day) and 8-13db (during the night). If it holds the line I should get the 5120Kbps Data Rate at some point (got the figure wrong in my last post, oops). Those gainQ2 red dots in DMT graphs look intriguing...

i can report the DMT tool works for the speedtouch 580 aswell

DMT v7.12 out