During the past month only four calculators were added or updated in the database, which brings us to a grand total of 41.
While this reduced amount of entries already lets take some useful conclusions, like the "introduced in production lines" dates on http://prizm.cemetech.net/index.php/Hardware_Revisions , it is still too insignificant for any kind of more complex statistical analysis - one or two entries can completely change the outcome.

I ask that those with ID stickers that are almost unreadable, but still containing some remains of ink, please take the best possible photo / flatbed scan of them and post them here or privately send them to me, along with all the other information you can gather, most importantly the original OS version and hardware revision. It appears we're not going to get many data points, so let's at least make each point is as complete as possible.
Did you have any duplicates ever - the reason I ask is because I posted a serial from a photo of it posted on ebay a couple of weeks ago - so someone else could have done it as well before or after me.
If the website doesn't say something about updating an existing calculator's information, then that's because it's the first time that ID is being submitted. We don't store duplicates.

Some thoughts from IRC:
Quote:
18:49 < gbl08ma> Assuming serial numbers count from the beginning, and are not per month, then the 393rd Prizm came with OS 01.00, but the 3209th one had 01.01 already.
18:50 < gbl08ma> Matches with the theory that 01.00 had really nasty bugs they quickly fixed
18:50 < gbl08ma> (I think TeamFX said 01.00 was present only in the ones sent to teachers)
18:53 < gbl08ma> First production date for fx-CG10: September 2010
18:54 < gbl08ma> First production date for fx-CG20: January 2011
18:55 < gbl08ma> Last production date for fx-CG10: March 2013
18:55 < gbl08ma> Last production date for fx-CG20: September 2014
19:00 < gbl08ma> If the first OS was built on September 2010, as well as the bootloaders, and in October 2010 calcs were coming out with HW rev 001V03...
19:01 < gbl08ma> That means 001V02 barely saw any units, and there may not even be earlier mass-produced revisions
19:01 < gbl08ma> That means, the only hardware revisions we need to care about are 001V03 and 001V04. Also explains why nobody has submitted a 001V02 calc, they might not exist at all
Quote:
Matches with the theory that 01.00 had really nasty bugs they quickly fixed

Back in 2010, there were these teacher events in early October where you could get a Prizm. Casio simply had no time to fully test the OS 1.00, but I could not find any serious bugs.

Also, there are just two weeks between OS 1.00 (2010.0922.1713) and OS 1.01 (2010.1005.1420). My best guess is that OS 1.00 broke some specific Prizm feature (which I do not know).

There are a lot of changes between OS 1.00, OS 1.01 and OS 1.02 (fx-CG20 preinstalled) and this is another hint that general OS development had not been completed when the fx-CG10 was released.

Quote:
Also explains why nobody has submitted a 001V02 calc, they might not exist at all

The member Vinx from casiocalc.org has a revision 001V02 Prizm. He got it from ebay from a teacher.
I will ask him for the product code if he still has that model.
Thanks for your insights, TeamFX.

TeamFX wrote:
The member Vinx from casiocalc.org has a revision 001V02 Prizm. He got it from ebay from a teacher.
I will ask him for the product code if he still has that model.


In this case, even more interesting than the product ID would be to get a good quality photo of the PCB. I came to the conclusion that I was crazy when I mentioned the possibility that 001V02 had not been mass produced, because flyingfisch's first Prizm (which became a brick) had this hardware revision, and there's even a low quality photo of the PCB: http://s.lowendshare.com/1/1423611015.469.prizm-screw.png

Unfortunately this picture is too low-quality to do any sort of comparison with other revisions, and flyingfisch doesn't have a better quality copy (nor the product ID of the calculator in question).
Quote:
In this case, even more interesting than the product ID would be to get a good quality photo of the PCB.

There is one in my archive.

Also note, if the hardware revision number 001V0x does not change, then the PCBs schould be identical. Casio even increases this number when just moving some of the text printed on the PCB.

They added some electrical component in revision 001V03. I'm not sure what it is. It's black and has a letter "S" printed on it. It can also be found in the ClassPad 330+ and 400. Seems to be some kind of capacitor.


PS: All fx-CG20 have a slightly better LCD controller (3A36-2 instead of 3A36-1).
This controller provides a better white color. On early fx-CG10 models, white is a bit yellowish.

The latest conroller is a 3A36B-1 that comes with OS 1.04.0201 models. This controller provides better color values (red is no longer pink).
TeamFX wrote:
There is one in my archive.

Hmm yes, it's the fx-CG 10 that's in there. My bad for only looking at fx-CG 20 pictures (without realizing that probably not a single 001V02 CG 20 was produced).

TeamFX wrote:
They added some electrical component in revision 001V03.


Is it near the keyboard connector? If yes, I've found it. Might be a small capacitor, yes.

Talking about capacitors, my capacitor-less "bad repair job" CG 20 still works fine...

And it's interesting to know there is a difference in fx-CG 20 (before 001V04) besides the two flash bytes and PCB markings... and makes me wonder why they were including the better controller on the fx-CG 20s only.

Do you know if the CG 10s with 001V04 mainboard also have the new "red fixed" 3A36B-1 controller (which means that all 001V04 calculators have the same LCD controller), or do they keep using the old "yellow white" one?
Quote:
Is it near the keyboard connector?

Yes. Next to the inductor.

Quote:
makes me wonder why they were including the better controller on the fx-CG 20s only.

It should be included in later fx-CG10 models as well.

Quote:
Do you know if the CG 10s with 001V04 mainboard also have the new "red fixed" 3A36B-1 controller?

I think so. Just ask someone here on Cemetech if the red color looks ugly.
Executing "SolveN(0)" in the Run-Matrix menu gives a nice red warning.
TeamFX wrote:
Executing "SolveN(0)" in the Run-Matrix menu gives a nice red warning.


In fact, executing (almost?) any command involving SolveN will result in a nice red warning (really, one of the most annoying things in the OS).

Users willing to test this by seeing much red can also use the "Color light" mode of the Utilities add-in, set to red. Or without any add-ins, by pressing Menu -> 4 -> Up -> Left -> Shift -> 5 -> 2 -> 3 -> 3 -> 1 -> EXIT -> Down -> Right.
On that wiki page there is the text:

Quote:
001V04
...
Differences near the USB port - new test contact;
New test contact near the 3 pin port;

That is a bit inaccurate.

First, if you take a closer look, you will notice that these contacts are permanently shorted and because of this, they don't really matter.

Second, both contacts are connected to ground and more importantly they are connected with each other on the other side of the PCB.

So you may add the following sentence:
USB ground and 3Pin ground are directly connected.
Thanks, corrected. Please let us know if you find any other inaccuracies.
Do you know that the fx-CG10 hardware revision 001V03 uses the old 3A36-1 LCD controller?

I guess, there could be a few models that have the old controller, but since fx-CG10 and fx-CG20 use the same hardware, there should be no further differences.
Hmm, I had misinterpreted your comment as "fx-CG 20s all have controller ABC, therefore all fx-CG 10s have controller DEF", which is obviously not logically correct.

So yes, it's possible that only early CG 10s have the 3A36-1, and that after the CG started being produced they switched all models to the 3A36-2 up until HW 001V04. This would greatly simplify things and make more sense (why keep different circuits around?). I know geekboy has a dead fx-CG 10 (001V03)... and I just asked... and:
Quote:
21:45:26 (C) gbl08ma: geekboy: could you please see if the markings on the LCD controller are 3A36-1 or 3A36-2?
21:46:21 (C) gbl08ma: (the LCD controller is the tiny circuit at the top, which you have detached from the main board in your pictures)
21:46:44 (#) geekboy: 3A36-2
21:47:39 (C) gbl08ma: thanks a lot. Looks like some more wiki edits are needed...


So I'll edit the wiki again... and at the same time I'm happy I was wrong, this makes things simpler.

EDIT: edits done. From my point of view, 3A36-1 came with 001V02 (which encompasses fx-CG 10s only, as fx-CG 20s only began production well after 001V03 was introduced), 3A36-2 came with 001V03 (fx-CG 10 and 20), and 3A36B-1 comes with 001V04 (fx-CG 10 and 20).

In hindsight, I should have added a field for the LCD controller version to my survey. Anyway, the amount of people who fill the HW revision field is already so low, I doubt it would help much. People interested in contributing that additional detail can always do so using the "comments" field.
I figured out a few things on the OS versioning scheme. You may want to add this too.

OS version "xx.xx.YZxx"

[Region]
Y = 0 : n/a
Y = 1 : Australia
Y = 2 : France
Y = 3 : USA and Canada
Y = 4 : China
Y = 5 : Singapore

[Math input/output features]
Z = 0 : n/a
Z = 1 : Slim
Z = 2 : All
Z = 3 : Reduced
Z = 7 : None

"Slim" refers to the fx-9860G Slim. This model supports "Syntax help", but does not have all the math input/output features. So it is also a "Reduced" variant.

"All" refers to models supporting all math input/output features.

"Reduced" refers to the fx-9860G and the fx-9860GIIs (a special GII model). Some functionality is missing like results as multiples of pi.

"None" refers to models that do not have any math input/output features. This is true for the fx-7400GII and French Graph 25+ Pro, as well as the fx-9750GII.

The French Graph 35+ USB (fx-9750GII) supports math input/output and has the same OS version as the Graph 75/95 (fx-9860GII). However, both OSes are totally incompatible...

Possible meaning of the last two digits:

OS version "xx.xx.xxAB"

[Build option 1]
A = 0 : Standard
A = 1 : Development

"1" is set in some fx-9860G prototype emulator.
"1" remains to be set when OS 1.00 is declared final and there are no other changes (fx-CG10, fx-FD10 Pro).

[Build option 2]
B = 0 : Standard
B = 1 : Special

"1" is set on fx-9860GII OSes compiled for the SH-4A hardware.
"1" is set on special OS builds (Slim OS 1.10, fx-CG OS 1.04).

PS: The fx-CG20 CN comes with OS 1.04.4200 preinstalled. It seems that on these models the default message and menu language is Chinese rather than English.
TeamFX, please see if you agree with http://prizm.cemetech.net/index.php/Versions_and_versioning#Versioning
Product code from Vinx is 755AW09MA000351
Hi, I tried inputting the following id from http://www.ebay.co.uk/itm/Casio-fxCG10-Colour-USB-Graphic-Calculator-with-Cover-and-Utilities-CD-fxCG20-/171737644671 (FX CG 10 but in the UK which is unusual and with OS version being differnt to what I expect) but it did not work - maybe one of you guys can read the ID better than me and input it in the database. Kind regards
That ID had already been submitted in the beginning of March, but I've updated the information associated with it anyway.

At this point it is pretty certain that the six digits at the end are a unique count and not a monthly count as previously thought, so I'm updating the website and the Prizm Wiki to reflect that.
It's not unique. That value overflows from time to time.
The serial number in the flash ROM seems to be better.

By the way, the blue Prizms have a 755CW* product code and they have the text "School Property" on the front.
Here are some photos of that model: http://shaiwu.smzdm.com/p/29827
The value only overflows when it gets to 999999, right? (Which is yet to happen, and probably won't happen in a long time) Because if it has overflowed already, then in more than 40 IDs there's no sign of it happening.

And by the "serial number in the flash ROM", do you mean the USB ID? Because that doesn't appear to be serial, just a random string. Of course, since we haven't been collecting it, we may be missing some pattern in it...
  
Register to Join the Conversation
Have your own thoughts to add to this or any other topic? Want to ask a question, offer a suggestion, share your own programs and projects, upload a file to the file archives, get help with calculator and computer programming, or simply chat with like-minded coders and tech and calculator enthusiasts via the site-wide AJAX SAX widget? Registration for a free Cemetech account only takes a minute.

» Go to Registration page
Page 2 of 3
» All times are UTC - 5 Hours
 
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum

 

Advertisement