Threaded Mode | Linear Mode

ThomasF · (This post was last modified: 08-03-2023 06:37 PM by ThomasF.)

(08-03-2023 06:14 PM)J-F Garnier Wrote:
(08-03-2023 05:56 PM)ThomasF Wrote: I noticed that the (really) high number decreased when I entered a small program ...
Something to dig into ...

Try to reset the calculator after switching to the 'BigMem' mode, and check again MEM.
If I'm not wrong, it may clarify things.

J-F

The 15C shows number of register in MEM, but BigMem shows a single value, which I guessed is free memory, but registers or bytes (probably)? If so, then a lot of code could be entered ... Big Grin

But maybe I'm on the wrong track ...
Thanks for the tip, will check accordingly ... Wink

Cheers,
Thomas

Voldemar · 08-03-2023, 06:36 PM

(08-03-2023 05:40 PM)Raymond Del Tondo Wrote: Found 16C mode, too:-)

Calculator off.
Hold [g] [ENTER], press [ON]
On screen
1.L 2.C 3.H
Then press [4]
On screen
15 15.2 16

Massimo Gnerucci · (This post was last modified: 08-03-2023 06:37 PM by Massimo Gnerucci.)

(08-03-2023 05:50 PM)Divasson Wrote: This is where a good overlay could come handy:
(designed and ordered a batch - it will arrive later in August)

Niiiice! Why can't I order one yet? Big Grin

Voldemar · 08-03-2023, 06:58 PM

(08-03-2023 06:36 PM)Voldemar Wrote:
(08-03-2023 05:40 PM)Raymond Del Tondo Wrote: Found 16C mode, too:-)
Calculator off.
Hold [g] [ENTER], press [ON]
On screen
1.L 2.C 3.H
Then press [4]
On screen
15 15.2 16

And then press [7]
Big Grin

Divasson · 08-03-2023, 06:59 PM

(08-03-2023 06:36 PM)ThomasF Wrote:
(08-03-2023 06:14 PM)J-F Garnier Wrote: Try to reset the calculator after switching to the 'BigMem' mode, and check again MEM.
If I'm not wrong, it may clarify things.

J-F

The 15C shows number of register in MEM, but BigMem shows a single value, which I guessed is free memory, but registers or bytes (probably)? If so, then a lot of code could be entered ...
But maybe I'm on the wrong track ...
Thanks for the tip, will check accordingly ...

Cheers,
Thomas

This is one of the reasons 15.2 is not standard: you are getting the same display as with 96 registers - just there is one more digit in the non-allocated registers due to the bigger memory: the allocated registers and unallocated registers may not be separated on the display, depending on the MEM split! This leads to confusion unless there was a heavy patching of the original firmware (leading to more bugs...etc.). It was decided to play safe.

Valentin Albillo · 08-03-2023, 07:08 PM

.
Hi, José,

(08-03-2023 06:24 PM)Divasson Wrote: The original firmware only guarantees inverting matrix of up to 8x8. Therefore we need a method to invert up to 13x13. Valentiiiiiiiiiiiín ! We need you !

I'll post tomorrow or the day after that (i.e. Saturday Aug, 5th) a couple of HP-15C programs I've written which can be used to compute the matrix inverse and the determinant of an NxN matrix (2 ≤ N ≤ 16 depending on available RAM,) which will run on any HP-15C version, including the original, the LE (patched for extra RAM or not), the CE (with 96 regs or more), the DM15's various versions and firmwares, as well as various true emulators (patched for extra RAM or not).

It runs quite fast, e.g. using the CE/96 regs it computes the inverse of a 9x9 matrix in 0.8 sec., and the determinant in ~ 0.2 sec.

I'm just completing the docs, including several worked examples, so stay tuned.

Best regards.
V.
P.S.: I wrote NxN Matrix Inversion programs in the distant past, e.g. have a look at this one for the HP-41C.

Voldemar · 08-03-2023, 08:03 PM

(08-03-2023 06:24 PM)Divasson Wrote: However, resetting is good procedure just in case.

I don't screw the battery cover anymore Smile

**rprosperi** · 08-03-2023, 08:06 PM

(08-03-2023 08:03 PM)Voldemar Wrote:
(08-03-2023 06:24 PM)Divasson Wrote: However, resetting is good procedure just in case.
I don't screw the battery cover anymore

I believe I won the bet. Smile

I predicted less than one week before someone posts that they've removed the screw.... lol. I will remove it immediately upon opening the unit I use, but will keep it in-place on the collection unit that shall remain in its box.

Voldemar · 08-03-2023, 08:29 PM

So interesting!
I haven't learned 15C yet, I have to learn 16C
Smile

Raymond Del Tondo · 08-04-2023, 08:09 AM

(07-27-2023 11:08 PM)rprosperi Wrote: The keys for the 15c CE are definitely not exactly the same as the LE as that was made in China and the 15c CE made by a different company in Philippines. The newest 12Cs also come from Philippines and the 2 units I have from there have been perfect with no missed or repeated strikes.

Hi Bob,
I received my 15CE yesterday, and yes, the keys feel noticably different and much better than those on my 12C 30th Anniversary Edition, which IIRC uses the same hardware as the LE. Back then I also had a 15C LE .

ThirdPoliceMan · 08-04-2023, 10:05 AM

Yesterday I received the HP-15C-CE.

Many thanks to Mr. Divasson for the logistics and to all those who realised the Collectors Edition project.

-----

S/N: 00917

-----

First and second impression (mainly compared to a DM15L)

Excellent printed user manual.

The Calculator looks very valuable, noble workmanship. Feels good and stable in the hand. Very good choice of colour for the labelling of the secondary functions (yellow - more of an Indian yellow). Just as good a feel as that of the DM15L. Back plate (made of metal?) with imprint very legible (much better than on the DM15L).

The 7-segment display is great to my taste (like a Casio G-Shock with digital display...). Very easy to read, looks very tidy. The contrast is not as good as that of the DM15L - under moderate lighting conditions, the DM15L is easier to read.

Significantly faster than a DM15L (in 12MHz and 48MHz mode).

Unfortunately, not all those status indicators are used as on the DM15L. This is disconcerting during some operations (e.g. no indication when to store in a register).

The keyboard is supple with a slightly springy tension, defined touch. Keys can be pressed in quick succession for fast input. A little too far to the pressure point, but that's a matter of taste.
The two coloured second function keys are slightly loose and sometimes twist a few degrees - not bad, but this disturbs the tidy image of the keyboard - at least for me.

BUT!

About every 40th keystroke is lost. Absolutely reproducible with continuous use! Sometimes (less often than unregistered keystrokes) entries are also repeated. This often affects the keys '8' and '6'. You can use the keyboard quickly because of its smoothness and good response, but there is now no reliability because of the missing or double registration of the keystrokes.
This is very annoying.

How can this be remedied?

Regards
Thomas

laurent_maury · 08-04-2023, 11:14 AM

Back to 15C mode???

GO TO 16C
Calculator off.
Hold [g] [ENTER], press [ON]
On screen
1.L 2.C 3.H
Then press [4]
On screen
15 15.2 16
Press [7]

Back to 15C
Calculator off.
Hold [g] [ENTER], press [ON]
On screen
1.L 2.C 3.H
Then press [4]
On screen
15 15.2 16
Press [4]
Press [e^x]

Thomas Radtke · 08-04-2023, 11:32 AM

Send it back for a replacement. You're the second customer to report this, but it's definitely not normal.

Voldemar · 08-04-2023, 11:59 AM

(08-04-2023 10:05 AM)ThirdPoliceMan Wrote: BUT!

About every 40th keystroke is lost. Absolutely reproducible with continuous use! Sometimes (less often than unregistered keystrokes) entries are also repeated. This often affects the keys '8' and '6'. You can use the keyboard quickly because of its smoothness and good response, but there is now no reliability because of the missing or double registration of the keystrokes.
This is very annoying.

How can this be remedied?

After two days of use I have also noticed that there are unregistered key presses, the blue g key sometimes does not register.
I did as in this advice:

(07-27-2023 11:08 PM)rprosperi Wrote: Also, more often these days than in classic "HP times", there is dust/dirt and oils on the material surfaces which can and do interfere with precise key clicks. Most often pressing a troublesome key 20-50 times "cleans" the dirt/oil/etc. and the key settles down and behaves more consistently as expected. I have no idea at all if this is the case here (I don't have a CE yet) but it is advice well worth trying, as it may clear the problem and will do no harm if that's not the cause.

It helps for now.

Voldemar · (This post was last modified: 08-04-2023 01:41 PM by Voldemar.)

(08-04-2023 11:14 AM)laurent_maury Wrote: Back to 15C mode???

GO TO 16C
Calculator off.
Hold [g] [ENTER], press [ON]
On screen
1.L 2.C 3.H
Then press [4]
On screen
15 15.2 16
Press [7]

Back to 15C
Calculator off.
Hold [g] [ENTER], press [ON]
On screen
1.L 2.C 3.H
Then press [4]
On screen
15 15.2 16
Press [e^x]

Go to 15.2C
Calculator off.
Hold [g] [ENTER], press [ON]
On screen
1.L 2.C 3.H
Then press [4]
On screen
15 15.2 16
Press [1/x]
Smile

Voldemar · 08-04-2023, 12:22 PM

On screen
15 15.2 16
Press [e^x] or [10^x] -> 15C
Press [y^x] or [1/x] -> 15.2C
Press [CHS] or [7] -> 16C

Voldemar · (This post was last modified: 08-04-2023 02:46 PM by Voldemar.)

So far, I don't see any difference in performance between 15C mode and 15.2C mode. The only thing I've noticed that is different is the memory ( [g] [RCL]).
The initial status
15C:
19 78 00-0
15.2C:
19174 00-0
There ar more memory registers in 15.2C mode. 174 uncommited registers (initial status) vs 78, total 195 vs 99.

J-F Garnier · (This post was last modified: 08-04-2023 03:44 PM by J-F Garnier.)

(08-03-2023 06:14 PM)J-F Garnier Wrote: Try to reset the calculator after switching to the [15.2] mode, and check again MEM.

(08-03-2023 06:24 PM)Divasson Wrote: AFAIK the move from 15 to 15.2 keeps the whole memory as it was. The reverse is not true, obviously. However, resetting is good procedure just in case.

The memory moves between the modes (storage registers are at the bottom, that moves from A0h to 40h) but you're probably right, José.
I quickly checked in my emulator, indeed moving to 15.2 mode ("192 registers") without reset seems safe, the system pointers are automatically updated by the firmware.
The reverse, going back to 15 mode ("96 registers") can be safe too, provided there are at least 96 uncommitted registers.

It's amazing that the original firmware, with just a few patches, can manage this situation so well, it was only supposed to handle the memory adjustments when complex mode/integration/solve are activated/deactivated.
That's the virtues of a robust coding!

J-F

Pekis · 08-04-2023, 03:20 PM

(08-04-2023 12:40 PM)Voldemar Wrote: So far, I don't see any difference in performance between 15C mode and 15.2C mode. The only thing I've noticed that is different is the memory ( [g] [RCL]).
The initial status
15C:
19 78 00-0
15.2C:
19174 00-0
There ar more memory registers in 15.2C mode. 174 uncommited registers (initial status) vs 78, total 195 vs 99.

Mine had 94 99 00-0 as initial MEM in 15.2 mode. After setting 19 DIM (i), MEM became 19174 00-0, so with max 174*7=1218 theorical programming steps ! (but still only 25 labels)

**Gene** · 08-04-2023, 03:41 PM

Beware entering programs > 999 steps. Line numbers do not function properly when 4 digits are needed. You are on your own in that territory. :-)

Also memory is properly conserved from 15 to 15.2 mode. No problem.

From 15.2 to 15 mode, it is conserved as long as no ram was used in the area abound the 15 mode. Example, if you have a 15.2 program using 800 steps and switch to 15 mode, it will not properly conserve memory. Same with data registers.