| Author |
Message |
|
calcdude84se
Member
Joined: 09 Aug 2009
Posts: 207
|
 Posted: 30 Sep 2009 03:56:29 pm Post subject: |
|
|
| I got mine in '08. I guess TI never did give up the bad LCD after all... |
|
| Back to top |
|
|
DrDnar
Member
Joined: 28 Aug 2009
Posts: 116
|
| Posted: 30 Sep 2009 06:54:57 pm Post subject: |
|
|
| Sure, replace one problem with another. |
|
| Back to top |
|
|
ztrumpet
Active Member
Joined: 06 May 2009
Posts: 555
|
| Posted: 30 Sep 2009 08:08:50 pm Post subject: |
|
|
Alrighty.
Today I spent my math class collecting data! Here it is:
1) My calc-bought new 9/9/09
84+se
Version: M
Os: 2.43
Boot Code: 1.02
X-Ram: No
ALCDFIX: No
2) Friend 1-bought new in 2007
84+
Version: B
Os: 2.40
Boot Code: 1.02
X-Ram: Yes
ALCDFIX: Yes
3) Friend 2-borrowed from school
84+se
Version: D
Os: 2.30
Boot Code: 1.02
X-Ram: Yes
ALCDFIX: No
4) Friend 3-bought new in 2007 (I think)
84+
Version: G
Os: 2.40
Boot Code: 1.02
X-Ram: Yes
ALCDFIX: No
5) Friend 4-bought used in 2007
84+se
Version: G
Os: 2.43
Boot Code: 1.02
X-Ram: Yes
ALCDFIX: No
6) Friend 5-bought new in 2007
84+se
Version: G
Os: 2.40
Boot Code: 1.02
X-Ram: Yes
ALCDFIX: Yes
7) Friend 6-bought new in 2007 (I think)
84+se
Version: L
Os: 2.43
Boot Code: 1.02
X-Ram: Yes
ALCDFIX: Yes
8) Teacher
84+se with extra link port on top
Version: C
Os: 2.30
Boot Code: 1.02
X-Ram: Yes
ALCDFIX: No
I don't know if the 84+'s have xram, but Dr. D'nar's test said they did.
Also, is the os and Base Code the same? |
|
| Back to top |
|
|
calcdude84se
Member
Joined: 09 Aug 2009
Posts: 207
|
| Posted: 30 Sep 2009 08:39:50 pm Post subject: |
|
|
| The OS shouldn't affect xram. The 84+ and 84+SE are the same except for the additional flash pages. Base code should always bee 1.02 or maybe 1.00 |
|
| Back to top |
|
|
DrDnar
Member
Joined: 28 Aug 2009
Posts: 116
|
| Posted: 01 Oct 2009 01:45:03 pm Post subject: |
|
|
| I have another idea for a test. Unfortunately, it involves resetting your RAM. The idea is to test each of the RAM pages for execution. Normally, pages 81h, 83h, 85h, and 87h should be executable, while trying to run code from 80h, 82h, 84h, or 86h should cause a reset. Try each of these pages with the attached program. If it crashes on pages 82h--87h, then the extra RAM is linked to a protected page; if not, the RAM is linked to an unprotected page. We can also see if changing the memory execution privileges affects which pages are executable. |
|
| Back to top |
|
|
thepenguin77
Advanced Newbie
Joined: 17 Jul 2009
Posts: 72
|
| Posted: 01 Oct 2009 03:17:57 pm Post subject: |
|
|
Well kind of humorous results. The program worked just like a normal calculator would.
81 yes
83 yes
85 yes
87 yes
80 no
82 no
84 no
86 no
This was one of those programs where grouping is very helpful. |
|
| Back to top |
|
|
critor
Member
Joined: 04 Feb 2009
Posts: 132
|
| Posted: 01 Oct 2009 04:14:32 pm Post subject: |
|
|
I've got five TI-84+.
I've tested the RAM pages with BANKTEST and MEMTST83.
And it seems I've got results which will make our problem a little more complicated...
1) TI-84+
Boot Code 1.00
OS 2.41
Hardware A
Date 05/2004
OK
2) TI-84+SE
Boot Code 1.02
OS 2.43
Hardware C
Date 03/2005
OK
3) TI-84+SE
Boot Code 1.02
OS 2.43
Hardware F
Date 05/2006
OK
4) TI-84+
Boot Code 1.02
OS 2.41
Hardware G
Date 03/2007
OK
5) TI-84+
Boot Code 1.02
OS 2.41
Hardware J
Date 03/2008
FAIL
Yeah, here's the problem.
I've got a TI-84+ revision J, which fails both tests.
And strangely, 4 posts ago, a revision L did pass the test, and a revision M didn't...
For now, let's sum up the last posts for bank-testing:
A: OK (critor)
B: OK (ztrumpet)
C: OK (critor + ztrumpet)
D: OK (ztrumpet)
F: OK (critor)
G: OK (3x ztrumpet)
J: no (critor)
L: OK (ztrumpet)
M: no (ztrumpet + thepenguin77)
Either my J-chip is broken...
Either TI did change somthing for the J-revision, then cancelled the modifications, and then put them back...
Or either, there is no 100% correlation between the extra-RAM and the hardware revision...
It may depend on something else too...
Last edited by Guest on 01 Oct 2009 04:23:26 pm; edited 1 time in total |
|
| Back to top |
|
|
FloppusMaximus
Advanced Member
Joined: 22 Aug 2008
Posts: 472
|
| Posted: 01 Oct 2009 05:21:31 pm Post subject: |
|
|
I suppose I might as well add to the cacophany. My 84+ is a revision C and works fine.
The funny thing about the RAM execution protection is that it's based on the value you output to the paging port, rather than directly based on the physical RAM address. As I recall, most of the experiments I did were on my 83+ SE, but what I found was that under some circumstances it would, say, allow execution on page "82", but not on page "8A", even though those represent the same physical RAM page.
So I would guess that the same is true here, that if you write 3 to port 5 and jump to C000, it won't reset, whereas if you write 2 to port 5 and jump to C000, it will - despite the fact that the same physical page is being used in both cases.
Last edited by Guest on 01 Oct 2009 05:23:48 pm; edited 1 time in total |
|
| Back to top |
|
|
calcdude84se
Member
Joined: 09 Aug 2009
Posts: 207
|
| Posted: 01 Oct 2009 06:56:12 pm Post subject: |
|
|
I got exactly the same results as thepenguin77 did. :biggrin:
thepenguin77 wrote: Well kind of humorous results. The program worked just like a normal calculator would.
81 yes
83 yes
85 yes
87 yes
80 no
82 no
84 no
86 no
This was one of those programs where grouping is very helpful.
We have a new, albeit strange, direction, provided the source code is right. I didn't look at it yet, I will now... |
|
| Back to top |
|
|
DrDnar
Member
Joined: 28 Aug 2009
Posts: 116
|
| Posted: 01 Oct 2009 07:33:08 pm Post subject: |
|
|
| Actually Flop's comment kind of kills this idea. I suspected that there might be a disconnect between the execution privileges and the physical memory. |
|
| Back to top |
|
|
calc84maniac
Elite
Joined: 22 Jan 2007
Posts: 770
|
| Posted: 01 Oct 2009 07:56:11 pm Post subject: |
|
|
FloppusMaximus wrote: I suppose I might as well add to the cacophany. My 84+ is a revision C and works fine.
The funny thing about the RAM execution protection is that it's based on the value you output to the paging port, rather than directly based on the physical RAM address. As I recall, most of the experiments I did were on my 83+ SE, but what I found was that under some circumstances it would, say, allow execution on page "82", but not on page "8A", even though those represent the same physical RAM page.
So I would guess that the same is true here, that if you write 3 to port 5 and jump to C000, it won't reset, whereas if you write 2 to port 5 and jump to C000, it will - despite the fact that the same physical page is being used in both cases.
Elaborate please. Execution should NEVER be allowed on page 82. |
|
| Back to top |
|
|
ztrumpet
Active Member
Joined: 06 May 2009
Posts: 555
|
| Posted: 01 Oct 2009 08:03:34 pm Post subject: |
|
|
critor wrote: And strangely, 4 posts ago, a revision L did pass the test, and a revision M didn't...
I going to do more testing tomorroe on J,K,L, and M (If I can find more). |
|
| Back to top |
|
|
FloppusMaximus
Advanced Member
Joined: 22 Aug 2008
Posts: 472
|
| Posted: 01 Oct 2009 08:20:46 pm Post subject: |
|
|
calc84maniac wrote: Elaborate please. Execution should NEVER be allowed on page 82.
It depends on what value is set in port 21. I wrote about this on WikiTI, just yesterday, as it happens.  (It's not all that interesting, really; from what I've seen, there's no way of enabling execution on page 0, nor any way of enabling more than 4 pages simultaneously.) |
|
| Back to top |
|
|
fullmetalcoder
Member
Joined: 01 Aug 2009
Posts: 139
|
| Posted: 02 Oct 2009 05:01:27 am Post subject: |
|
|
FloppusMaximus wrote: calc84maniac wrote: Elaborate please. Execution should NEVER be allowed on page 82.
It depends on what value is set in port 21. I wrote about this on WikiTI, just yesterday, as it happens. (It's not all that interesting, really; from what I've seen, there's no way of enabling execution on page 0, nor any way of enabling more than 4 pages simultaneously.)
your addition puzzles me : why do you mention pages $89-$8F ? are they straightforward aliases to $81-$87 or is there something more subtle? |
|
| Back to top |
|
|
ztrumpet
Active Member
Joined: 06 May 2009
Posts: 555
|
| Posted: 02 Oct 2009 02:37:31 pm Post subject: |
|
|
I didn't have much time to test today so I only have 2 more, yet both fails.
9) Friend 7-bought new in 07
84+se
Version: H
Os: 2.40
Boot Code: 1.02
X-Ram: No
ALCDFIX: No
10) Friend 8-bought new in 08
84+se
Version: M
Os: 2.43
Boot Code: 1.02
X-Ram: No
ALCDFIX: No |
|
| Back to top |
|
|
thepenguin77
Advanced Newbie
Joined: 17 Jul 2009
Posts: 72
|
| Posted: 02 Oct 2009 02:48:00 pm Post subject: |
|
|
So when I said all those I tested were M's I was wrong. But I did some more testing today with bankTest. It failed every time which could indicate something faulty but.
11)
84+se
Version: H
Os: 2.43
Boot Code: 1.02
X-Ram: No
ALCDFIX: No
12)
84+se
Version: H
Os: 2.43
Boot Code: 1.02
X-Ram: No
ALCDFIX: Yes
13)
84+se
Version: J
Os: 2.43
Boot Code: 1.02
X-Ram: No
14)
84+be
Version: L
Os: 2.43
Boot Code: 1.02
X-Ram: No
ALCDFIX: No
Cpu Clock: 16.3 mhz
15)
84+be
Version: L
Os: 2.43
Boot Code: 1.02
X-Ram: No
ALCDFIX: No
cpu Clock: 16.1 mhz
Revision L has a crazy fast CPU. Most CPUs are around 15.7 mhz. I also found 84+be's with no hardware letter. One was omitted and the other had a 0. |
|
| Back to top |
|
|
ztrumpet
Active Member
Joined: 06 May 2009
Posts: 555
|
| Posted: 02 Oct 2009 02:50:29 pm Post subject: |
|
|
So the ones that have failed are:
H,J,L,and M but a J and a L have also passed.
Someone added on to my number system...
Last edited by Guest on 02 Oct 2009 02:51:42 pm; edited 1 time in total |
|
| Back to top |
|
|
DrDnar
Member
Joined: 28 Aug 2009
Posts: 116
|
| Posted: 02 Oct 2009 05:23:24 pm Post subject: |
|
|
I assume "X-RAM" means extra RAM, not cross-linked RAM or bad RAM. If not, then this table is wrong.
Code:
Unit label Hardware Bad RAM Bad Screen Notes
New 9/9/9 M SE Yes No
Friend 1 (07) B B No Yes
Friend 2 D SE No No
Friend 3 (07) G B No No
Friend 4 G SE No No
Friend 5 (07) G SE No Yes
Friend 6 (07?) L SE No Yes
Teacher C SE No No
Friend 7 (07) H SE Yes No
Friend 8 (08) M SE Yes No
11 H SE Yes No
12 H SE Yes Yes
13 J SE Yes
14 L B Yes No 16.3 MHz
15 L B Yes No 16.1 MHz
In other news, tabs are just plain broken on this forum. |
|
| Back to top |
|
|
ztrumpet
Active Member
Joined: 06 May 2009
Posts: 555
|
| Posted: 02 Oct 2009 10:06:52 pm Post subject: |
|
|
| Also, critor published some 11 posts up. |
|
| Back to top |
|
|
calcdude84se
Member
Joined: 09 Aug 2009
Posts: 207
|
| Posted: 03 Oct 2009 09:32:24 am Post subject: |
|
|
So there is not a strong correlation between hardware version and the availability of XRAM. It appears there is no easy way to say whether it is available without running a program.
Also, have I missed something, or have we already decided that these "no XRAM" calculators only have 48KB of RAM? Did someone physically verify the number or RAM chips, or are we going on without solid information? |
|
| Back to top |
|
|
|
Calculator Tech Support => Technology & Calculator Open Topic
