[A83] Fwd: [TI-H] Re: TI-83P SE hardware, looking for thoughts has hardw

[Prev][Next][Index][Thread]

[A83] Fwd: [TI-H] Re: TI-83P SE hardware, looking for thoughts has hardware ports


Hope Dan doesn't mind (from the hardware list)

----------  Forwarded Message  ----------
Subject: [TI-H] Re: TI-83P SE hardware, looking for thoughts
Date: Wed, 25 Jul 2001 13:52:23 -0400
From: "Dan Englender" <dan@calc.org>


Well, I guess my thought was, why would they add specifically a 32.768 khz
crystal when whatever they had been doing previously (on the 83P) for clock
speed and interrupts should work just fine.  It actually boils down to a two
part question I guess:
A) Why add an xtal?
B) Why 32.768 khz?

As for software, the boot page is different on the SE than the 83 Plus,
however, the changes could be for a variety of things.  There are also a lot
more usable hardware ports on the SE.  At the moment there seem to be
between 20 and 30 ports that have some sort of function that I haven't
figured out yet.  This is compared with 2 unknown ports on the regular 83P.

On whether the main clock signal is generated from this crystal, I'm
tentativly guessing no, because of the hardware timers.  I think I mentioned
them, but anyhow, there's a set of hardware timers on the SE (not present on
the regular 83P).  To change the frequency that the timers run at, you have
to play with some values on the ports.  There's one bit that, if you set it,
causes the timers to run off the main CPU frequency.  There's another that,
if you set it, causes the timers to run off the 32.768k xtal frequency.  And
if you set both, it seems to run off a "mix" of the two frequencies.  That
seemed to indicate to me that they were seperate, but my say-so doesn't hold
much ground :)


-Dan

P.S. - I suppose it wouldn't hurt to append what information I've currently
found about the ports.
On the 83P, the following ports are used: 0,1,2,3,4,5,6,7,10h,11h,14h,16h
Functions:
0 - link port
1 - keyboard port
2 - hardware stuff (low battery indicator, etc).
3 - Power and interrupt enable
4 - Interrupt speed and type
5 - unknown
6 - memory bank0 page
7 - memory bank1 page
10h - LCD instruction port
11h - LCD data port
14h - Flash lock/unlock port
16h - unknown
....You may be wondering how I know these are the only used ports.  Well,
TIOS outputs to all these ports, plus, all other ports seem to be "shadows"
of currently existing ports.

On the SE:  All ports between 0-3Fh are either zero, or have some value.
Every port 40h-0FFh is set to 0FFh.  Thus all usable ports must be in the
00-3Fh, though that's no gaurantee that *all* ports in this range are used).
0 - link (direct line access)
1 - keyboard
2 - hardware version type, battery status, etc
3 - Power and interrupt enable
4 - Interrupt type and speed
5 - unknown (seem to behave different than 83P though)
6 - memory bank0
7 - memory bank1
8 - unknown (probably something to do with link port though)
9 - link port hardware assist status
Ah - link hardware assist input
Bh/Ch - unsure, but probably link hardware assist input buffer
Dh - link hardware assist output
Eh/Fh - unknown, intialized to zero by TIOS and never again used
10h - LCD instruction
11h - LCD data
12h - appears to be a shadow of 10h
13h - unknown, not used by TIOS at all
14h - Flash lock/unlock
15h - unknown, when read returns a value of 33h
16h/17h - unknown, not used by TIOS at all
18h-1Fh - seem to be a harware interface to help with MD5 routines
20h - bit0 = select CPU speed, bit1 = unknown.  Other bits not used
21h-26h - unknown, intialized to values by boot page, never used again.
Can't change values with an output to the port
27h-2Fh - unknown, have default values, values can be changed with no
noticable effect, except port 29h, where if bit4 is reset, the calculator
crashes.
30h-37h - hardware timers
38h-3Fh - unknown, not used by TIOS at all.