Author |
Message |
|
critor
Member
Joined: 04 Feb 2009 Posts: 132
|
Posted: 03 Oct 2009 03:09:53 pm Post subject: |
|
|
thepenguin77 wrote: Well that's interesting. But I have to break the trend with the P-0508M FAIL 120.
I've updated the table with your result.
So the screen fps problem seems more complicated than the extra-RAM problem.
But maybe we'll see something if we have more tests. |
|
Back to top |
|
|
critor
Member
Joined: 04 Feb 2009 Posts: 132
|
Posted: 03 Oct 2009 03:14:23 pm Post subject: |
|
|
Well, as we've changed the thread page, I'm pasting the table here.
Code: End of Serial |128K RAM test |LCD FPS test |Boot Code |CPU |Type |ASIC
|(OK/FAIL) |(120/240/280) |(1.00/1.02) |(MHz) |(BE/SE)|(CPU/RAM chip)
----------------+----------------+---------------+------------+------+-------+------------------
S-0304 |OK | |1.00 | |BE |TI-REF 83PLUSB/TA2
S-0504 |OK |120 |1.00 |13.6 |BE |TI-REF 83PLUSB/TA2
S-0504 |OK | |1.00 | |SE |TI-REF 83PLUSB/TA2
S-0704 |OK | |1.00 | | |TI-REF 83PLUSB/TA2
S-0704A |OK | | | | |
S-0804 |OK | |1.00 | |BE |TI-REF 83PLUSB/TA2
----------------+----------------+---------------+------------+------+-------+------------------
S-0305C |OK |120 |1.02 |14.5 |SE |TI-REF 84PLUSB/TA3
S-0505D |OK |120 |1.02 |14.6 |SE |TI-REF 84PLUSB/TA3
S-0605C |OK | |1.02 | | |TI-REF 84PLUSB/TA3
S-0605D |OK | |1.02 | |SE |TI-REF 84PLUSB/TA3
S-0306F |OK | |1.02 | | |TI-REF 84PLUSB/TA3
S-0506F |OK |120 |1.02 |15.2 |SE |TI-REF 84PLUSB/TA3
S-0606F |OK | |1.02 | |BE |TI-REF 84PLUSB/TA3
----------------+----------------+---------------+------------+------+-------+------------------
K-0906B |OK |120 |1.02 |15.4 |BE |
----------------+----------------+---------------+------------+------+-------+------------------
S-0207G |OK |240 |1.02 |15.1 |SE |TI-REF 84PLUSB/TA3
S-0307G |OK |120 |1.02 |15.3 |BE |TI-REF 84PLUSB/TA3
S-0307G |OK |120 |1.02 |15.6 |SE |TI-REF 84PLUSB/TA3
S-0407G |OK | |1.02 | |SE |TI-REF 84PLUSB/TA3
----------------+----------------+---------------+------------+------+-------+------------------
S-0507H |FAIL |240 |1.02 |15.7 |SE |
S-0607H |FAIL | |1.02 | | |
----------------+----------------+---------------+------------+------+-------+------------------
K-0308J |FAIL | |1.02 | | |
----------------+----------------+---------------+------------+------+-------+------------------
P-0108K |FAIL |240 |1.02 |15.6 |SE |TI-REF 84PLCR/TA1
P-0308J |FAIL |240 |1.02 |15.9 |BE |TI-REF 84PLCR/TA1
P-0308K |FAIL | |1.02 | | |TI-REF 84PLCR/TA1
P-0308L |FAIL | |1.02 | |SE |TI-REF 84PLCR/TA1
P-0408L |FAIL |120 |1.02 |15.9 |SE |TI-REF 84PLCR/TA1
P-0508M |FAIL |120 |1.02 |15.7 |SE |TI-REF 84PLCR/TA1
P-0508M |FAIL |240 |1.02 |15.7 |SE |TI-REF 84PLCR/TA1
P-0808M |FAIL | |1.02 | |SE |TI-REF 84PLCR/TA1
P-0509M |FAIL |240 |1.02 |15.9 |SE |
Please, help us!
Report the RAM-test result (and other things if you want) with the last 7 characters of your serial (very important).
For now, it seems that all TI-84+ made in the P-factory (China) fails the extra-RAM test.
Last edited by Guest on 05 Oct 2009 04:27:12 pm; edited 1 time in total |
|
Back to top |
|
|
calc84maniac
Elite
Joined: 22 Jan 2007 Posts: 770
|
Posted: 03 Oct 2009 03:22:04 pm Post subject: |
|
|
All three of mine have S, and they work. |
|
Back to top |
|
|
thepenguin77
Advanced Newbie
Joined: 17 Jul 2009 Posts: 72
|
Posted: 03 Oct 2009 03:28:44 pm Post subject: |
|
|
Mine was boot 1.02 with 15.7 mhz. SE |
|
Back to top |
|
|
critor
Member
Joined: 04 Feb 2009 Posts: 132
|
Posted: 03 Oct 2009 03:29:02 pm Post subject: |
|
|
Thanks penguin77 -> updated.
calc84maniac wrote: All three of mine have S, and they work.
Good!
Thank you calc84maniac!
Can you please post the last 7 characters of your serials, so I can update the table?
It may help us to determine "when" TI did change some hardware revisions, or when TI did begin or end the "factory-switching".
Thanks.
Last edited by Guest on 03 Oct 2009 04:16:35 pm; edited 1 time in total |
|
Back to top |
|
|
calc84maniac
Elite
Joined: 22 Jan 2007 Posts: 770
|
Posted: 03 Oct 2009 04:41:57 pm Post subject: |
|
|
S-0704A
S-0605C
S-0207G |
|
Back to top |
|
|
critor
Member
Joined: 04 Feb 2009 Posts: 132
|
Posted: 03 Oct 2009 04:54:40 pm Post subject: |
|
|
calc84maniac wrote: S-0704A
S-0605C
S-0207G
Thank you -> updated
Now some TI-84+ manufactured between 2007/02 and 2008/05 would be very interesting...
That's the "critic" period, where the same hardware revision may have been manufactured by both factories: S and P.
Other TI-84+ are still welcomed though
Now that we're allmost sure that TI-84+ with a "faulty" RAM comes from the P factory in China... can we email TI-Care, in order to see what stupidity they're able to answer in front of the facts?
Last edited by Guest on 03 Oct 2009 04:55:21 pm; edited 1 time in total |
|
Back to top |
|
|
DrDnar
Member
Joined: 28 Aug 2009 Posts: 116
|
Posted: 03 Oct 2009 05:22:30 pm Post subject: |
|
|
They'd tell you that the TI-84 has 24K of RAM; the TI-84 had always been at war---I mean had 24K of RAM. |
|
Back to top |
|
|
ztrumpet
Active Member
Joined: 06 May 2009 Posts: 555
|
Posted: 03 Oct 2009 06:50:07 pm Post subject: |
|
|
How do I test the CPU speed? |
|
Back to top |
|
|
thepenguin77
Advanced Newbie
Joined: 17 Jul 2009 Posts: 72
|
Posted: 03 Oct 2009 06:56:31 pm Post subject: |
|
|
ztrumpet wrote: How do I test the CPU speed?
Run this program and press 1. |
|
Back to top |
|
|
calc84maniac
Elite
Joined: 22 Jan 2007 Posts: 770
|
|
Back to top |
|
|
critor
Member
Joined: 04 Feb 2009 Posts: 132
|
Posted: 04 Oct 2009 03:12:42 am Post subject: |
|
|
Updated with one new record.
I'll have 3 "small" groups of students on tuesday morning.
I may test their TI-84+, but most of them are new.
So I'll probably get P-XXXXM FAIL.
Let's burn the P-factory! |
|
Back to top |
|
|
critor
Member
Joined: 04 Feb 2009 Posts: 132
|
Posted: 04 Oct 2009 06:27:56 am Post subject: |
|
|
Table updated with an interesting calculator: a K-revision.
(right during the interesting period)
It's failing....
Guess what!
It's a P, of course! :roll:
So for now, march 2008 is the first month we're getting TI-84+(SE) manufactured by P.
(there's been a big hole since april 2007)
And during march 2008, P has manufactured 3 different hardware revisions: JKL (and J was allready failing).
More tests are needed during that period, before we can be "sure" and start disassembling similar P/S hardwares under a microscope.
Last edited by Guest on 04 Oct 2009 06:33:43 am; edited 1 time in total |
|
Back to top |
|
|
critor
Member
Joined: 04 Feb 2009 Posts: 132
|
Posted: 04 Oct 2009 07:45:33 am Post subject: |
|
|
I've check the thread history.
There is only 1 thing which doesn't match, if I haven't missed anything.
ztrumpet has reported a L-revision which passed the X-RAM test.
And for now, we've got no TI-84+(SE) manufactured by S after 04/2007.
So either:
- S has been manufacturing a minority of the TI-84+(SE) after 04/2007
- P did randomly (and very rarely) manufacture some TI-84+(SE) with a working 128Kb X-RAM
- the mentionned calculator has been sent for service, and it's hardware has been replaced - so the serial doesn't mean anything any more
- we're wrong about the S/P and X-RAM correlation (I don't think so)
|
|
Back to top |
|
|
critor
Member
Joined: 04 Feb 2009 Posts: 132
|
Posted: 04 Oct 2009 09:28:02 am Post subject: |
|
|
I've found it! I've found it!
I think I now have a more complete explanation to the problem.
I've opened some of my calculators, and I've seen a difference between P and S.
Yes: the ASIC (CPU/RAM combo chip) is simply different.
I've gathered some more information, by looking at motherboards on datamath.org.
Here are my hypothesis:
- Until 04/2007, TI-84+(SE) are manufactured by S, are passing the X-RAM test, and are using the ASIC TI-REF 83PLUSB/TA2 or TI-REF 84PLUSB/TA3.
- From 03/2008, TI-84+(SE) are manufactured by P, are failing the X-RAM test, and are using the ASIC TI-REF 84PLC/TA1.
Yes, the CPU/RAM chips used by the P-factory have a quite different tag, and probably are different or work differently.
I still have some TI-84+ to open...
Table updating in progress...
Serials and tests are still welcomed: we've got a gap of allmost a year between both hypothesis.
And if you want to open your 84+(SE) and report the ASIC type, no problem
Last edited by Guest on 04 Oct 2009 09:36:43 am; edited 1 time in total |
|
Back to top |
|
|
ztrumpet
Active Member
Joined: 06 May 2009 Posts: 555
|
Posted: 04 Oct 2009 10:17:35 am Post subject: |
|
|
Nice find, critor!
I think I can provide some key information because a lot of the calcs in my grade were bought around August 2007. So all the 07s I posted are from about August.
The L that dosn't quite follow has never been sent to Ti since Ti originally shipped it, so no hardware has ever been replaced.
Does anyone know how this happened? ztrumpet wrote: 2) Friend 1-bought new in 2007
84+
Version: B
Os: 2.40
Boot Code: 1.02
X-Ram: Yes
ALCDFIX: Yes
EDIT: What about reporting the full serial number? Then you/we could put all the calcs in order and it would help place oddities.
Last edited by Guest on 04 Oct 2009 10:24:03 am; edited 1 time in total |
|
Back to top |
|
|
critor
Member
Joined: 04 Feb 2009 Posts: 132
|
Posted: 04 Oct 2009 10:41:59 am Post subject: |
|
|
Here are some photos of the TI-84+(SE) ASICs (CPU/RAM combo chip)
1) The TI-REF 83PLUSB/TA2 (36x36 pins)
(1st hardware, 03/2004 - 08/2004, factory S, passing X-RAM test)
2) The TI-REF 84PLUSB/TA3 (36x36 pins)
(hardware F - G, 06/2006 - 04/2007, factory S, passing X-RAM test)
3) The TI-REF 84PLCR/TA1 (25x25 pins)
(hardware J - M, since 03/2008, factory P, failing X-RAM test)
Other serials/tests will help us filling the time-gaps between those chips.
Edit 1: Anybody understanding something from the writings on the chips?
Edit 2: Oh my God... just count the pins on the chips... the 3rd one simply cannot work the same way...
Last edited by Guest on 04 Oct 2009 10:56:32 am; edited 1 time in total |
|
Back to top |
|
|
thepenguin77
Advanced Newbie
Joined: 17 Jul 2009 Posts: 72
|
Posted: 04 Oct 2009 11:21:30 am Post subject: |
|
|
calc84maniac wrote:
My program does not take into accout the hardware port. But I don't think it would matter. My program waits for the LCD to say that it is finished before it writes again. This means that it updates a fast as the LCD can take.
The only way that that port would mess up the results is if for some reason you over-set it. Say your calculator needs a delay of 20 but you decide to tell it to wait for 50, then my program would say that it is slower then it actually is.
Edit: That wasn't the port I thought it was, but still, a 1 clock delay shouldn't mess up the results. Even if it changed it by 1 FPS you would still round it to to the right group because all calcs group around 120, 240, or 280.
Last edited by Guest on 04 Oct 2009 11:49:19 am; edited 1 time in total |
|
Back to top |
|
|
calc84maniac
Elite
Joined: 22 Jan 2007 Posts: 770
|
Posted: 04 Oct 2009 01:29:03 pm Post subject: |
|
|
thepenguin77 wrote: calc84maniac wrote:
My program does not take into accout the hardware port. But I don't think it would matter. My program waits for the LCD to say that it is finished before it writes again. This means that it updates a fast as the LCD can take.
The only way that that port would mess up the results is if for some reason you over-set it. Say your calculator needs a delay of 20 but you decide to tell it to wait for 50, then my program would say that it is slower then it actually is.
Edit: That wasn't the port I thought it was, but still, a 1 clock delay shouldn't mess up the results. Even if it changed it by 1 FPS you would still round it to to the right group because all calcs group around 120, 240, or 280.
No, I'm talking about the CPU speed test. That would certainly be affected if the opcodes are taking longer than you would expect. |
|
Back to top |
|
|
thepenguin77
Advanced Newbie
Joined: 17 Jul 2009 Posts: 72
|
Posted: 04 Oct 2009 02:35:42 pm Post subject: |
|
|
Wow I never thought about that. The code writes to ram 3 times per loop. So thats 3 t-states longer. 15770000/97*100=16250000 for my calc.
That seems really fast though... |
|
Back to top |
|
|
|