Speedtouch "extended reach"

[Duncan Sands]

On Friday 30 June 2006 18:41, Aurelio Arroyo wrote:
>  From file UK3012 Extended.zip\330 3012 Extended\Phonebook\Phonebook.ini
> > [...]
> > +	buf[0] = 0x0b;
> >   
> [Modem.POTS]
> ModemMode=11

Correct.

> > [...]
> > +	buf[0] = 0x10;
> > +	buf[4] = 0x20;
> >   
> [ModemSettings]
> ModemOption=1000000020

Correct.  In fact ModemOption can be up to 16 bytes long,
entirely determining the data sent in this urb.

> > [...]
> > +	buf[0] = 0xc0;
> > +	buf[1] = 0x1f;
> > [...]
> [USB]
> [...]
> BMaxDSL=8128

Correct.

In addition, the bytes sent by URB 147 and URB 148 are
determined by the parameters PeakDownStreamBitRate (ds) and
PeakUpStreamBitRate (us).  It's not clear exactly what the
relationship is: here's the data I gathered, I haven't had
time to analyse it yet:

	ds/us    U147  / U148
        ff/ff -> d4 ea / 05 00
        fe/80 -> 25 50 / 0a 00
        fe/1a -> 1c 50 / 32 00 (default)
        fe/10 -> 2e 50 / 52 00
        fe/02 -> 70 50 / 96 02
        fe/01 -> e0 50 / 2d 05
        fe/00 -> windows crashes (divide by 0?)
        fd/fe -> a0 ba / 05 00
        f0/f0 -> f2 f9 / 05 00
        ef/ef -> 5b ad / 05 00
        d0/1a -> 1c 81 / 32 00
        d0/10 -> 2e 81 / 52 00
        80/02 -> 70 12 / 96 02

It looks like ds determines the second value,
and us the other three.  Windows crashed when
us was set to 0, so presumably division is
involved.

A strange thing: the first byte of URB 149 and of URB 150
is set to 1 if vpi!=0, and 0 if vpi=0.  But why?!  And how
does this work if you have several connections open at once?
I was unable to work out what determines bytes 2 and 3 of
these URBs.

So the only remaining undetermined values seem to be in
URBs 149 and 150.

Here are the effects of some other parameters from the
registry:

BandwidthAlgo: can alter the sequence of altsetting calls for
interface 1.

DataAltSetting, MaxAltSetting: looks like the altsetting for
interface 1 is usually (see ForcedSelection) set to the minimum of these.

DriverType: not tested

EncapsulationMode: no visible effect

FCS (Frame Check Sequence?): no visible effect

ForcedSelection: some values cause BMaxDSL not to
be sent, and altsetting 3 to be selected for interface 1.

MTU: no visible effect

Name: not tested

vci: no visible effect

MaxSimConnections: no visible effect (however it creates a
new empty registry directory; based on the name it is similar
to the directory with DataAltSetting, EncapsulationMode,
FCS, MTU, Name, PeakDownStreamBitRate, PeakUpStreamBitRate,
vci and vpi in it; these are presumably all per-connection
values, but how are settings for other connections sent, and
does this mean that there is an upper bound on the number of
connections?).

FileVersion: not tested


Ciao,

Duncan.