TheBackShed.com

Forum Index : Microcontroller and PC projects : Programmable calculator on Pico DIY

Page 3 of 3

Author

Message

zeitfest
Guru

Joined: 31/07/2019
Location: Australia
Posts: 549

Posted: 02:29am 09 Dec 2024

Copy link to clipboard

Print this post

The "clicky button" route has not gone well

/

They are way too clicky/noisy ! Apart from the click annoying everyone including me within earshot over a large area, they also numb/damage the fingers after moderate use. The pcb seems to act as a sounding board making it worse. Adding logos/foam pads made the sound louder (??!) and attaching them was unreliable too. Also the round active button-press rotates within the switch, so a logo "+" rotates to look like a "x" which is not good. I also tried a membrane on top of the switches, it works ok but does not muffle the click.

Undeterred I have found a pushbutton that still has a definite action but is basically silent. Cheap too. Of course, the pinout is different to the usual 4 pin switch...so needs another pcb ...

The question is, is a two-pad smd pushbutton going to be fastened to the pcb robustly enough to stand up to keypad use. Dunno.

Mixtel90

Guru

Joined: 05/10/2019
Location: United Kingdom
Posts: 7504

Posted: 07:50am 09 Dec 2024

Copy link to clipboard

Print this post

Would these silent ones be of interest? AliExpress

same switch in 4-pin

They are 8x8, so a little more finger space.
.
Edited 2024-12-09 17:53 by Mixtel90

Mick

Zilog Inside! nascom.info for Nascom & Gemini
Preliminary MMBasic docs & my PCB designs

zeitfest
Guru

Joined: 31/07/2019
Location: Australia
Posts: 549

Posted: 10:06am 09 Dec 2024

Copy link to clipboard

Print this post

Great ! They look very useful, I will see if I can wrangle the existing pcb a bit to suit the 8x8 mm.

Mixtel90

Guru

Joined: 05/10/2019
Location: United Kingdom
Posts: 7504

Posted: 10:20am 09 Dec 2024

Copy link to clipboard

Print this post

You might want to try a couple first. The actuator isn't quite as rigid as a normal switch. Absolutely silent though. :)

Mick

Zilog Inside! nascom.info for Nascom & Gemini
Preliminary MMBasic docs & my PCB designs

zeitfest
Guru

Joined: 31/07/2019
Location: Australia
Posts: 549

Posted: 12:57pm 09 Dec 2024

Copy link to clipboard

Print this post

I tried these and they are good, they are smd with two pads though with less spacing for the pcb traces.

One hassle was that the 4-pin button switches included internal connections as two sets of two pins. Kicad schematics didn't really accept that so I bypassed the schematic and just did the pcb layout directly using the switches as wire links, so that is going to need rework.

Mixtel90

Guru

Joined: 05/10/2019
Location: United Kingdom
Posts: 7504

Posted: 01:42pm 09 Dec 2024

Copy link to clipboard

Print this post

I say those little switches and almost bought some, but I was saving money and took them out of the basket.

Trust KiCAD not to have a model for the most common tactile switch. :) I can't really believe that.

4-pin is probably the easiest arrangement to multiplex.

Mick

Zilog Inside! nascom.info for Nascom & Gemini
Preliminary MMBasic docs & my PCB designs

zeitfest
Guru

Joined: 31/07/2019
Location: Australia
Posts: 549

Posted: 10:06pm 23 Dec 2024

Copy link to clipboard

Print this post

Reworked the pcb to use the silent push buttons.

As before, the tca8414 module plugs in and provides i2c sockets.
The kicad view includes some pins on the front which will be left off in use.

The 6x5 area should be useful separately as a keypad too so I have created a cut-down with just that. Maybe the 'mites can use a I2C keypad ?

zeitfest
Guru

Joined: 31/07/2019
Location: Australia
Posts: 549

Posted: 01:41pm 05 Jan 2025

Copy link to clipboard

Print this post

The new pcb is working well, and is quiet

.

Tidying up the I/O (read and write) has had a benefit, as it happens the input fields now accept csv input from the keypad as data so multiple values can be input at once, it meant I had to add a "comma" button to the keypad, but much more convenient !.

Adding formatted I/O has been a real cow but worth it. Getting back into it after xmas snooze

zeitfest
Guru

Joined: 31/07/2019
Location: Australia
Posts: 549

Posted: 02:26am 12 Feb 2025

Copy link to clipboard

Print this post

Quote Adding formatted I/O

After a fair slog.. the software I/O is now fairly conventional and can read/write files well, eg the following accesses two files, reading from one and writing to the other. It looks like simultaneous operation is ok ( quibble, dodge, weave etc

) which was a design aim. That adds a very powerful processing ability. It includes csv format i/o for spreadsheets etc.

There are small bugs yet eg the (ed) float format is default still but I am well pleased that it's a goer.

>R
Running

PROGRAM testread
INTEGER*4 x
REAL a
DOUBLE b
INTEGER*1 c
INTEGER*2 d
INTEGER*4 e
CHARACTER*1 f
CHARACTER*7 rinso
75 FORMAT (F4,F4,I1,I2,I4,A1,A7,/)
85 FORMAT (A1,A1,F4,F4,I1,I2,I4,A1,A7,/)
OPEN "/Output.txt" WRITE
OPEN "/Inputfrm.txt"
READE (6,75) a, b, c, d, e, f, rinso
WRITE (8,75) a, b, c, d, e, f, rinso
READE (6,85) f, f, a, b, c, d, e, f, rinso
WRITE (8,75) a, b, c, d, e, f, rinso
END
\

open file write
open file read
readin

readin

OK
...

Input file content :

1.112.223445555MFlubber
9.998.887664444TGrungit

Output file content:

1.1100002.2200003445555MFlubber
9.9900008.8800007664444TGrungit

Edited 2025-02-12 13:06 by zeitfest

zeitfest
Guru

Joined: 31/07/2019
Location: Australia
Posts: 549

Posted: 09:42pm 12 Feb 2025

Copy link to clipboard

Print this post

After mods and the right format spec ..

75 FORMAT (F4,F4,I1,I2,I4,A1,A7,/)
85 FORMAT (F4.2,F4.2,I1,I2,I4,A1,A7,/)

OPEN "/Nuout.txt" WRITE
OPEN "/Inputfrm.txt"

READE (6,75) a, b, c, d, e, f, rinso
WRITE (8,85) a, b, c, d, e, f, rinso

READE (6,75) a, b, c, d, e, f, rinso
WRITE (8,85) a, b, c, d, e, f, rinso

END
\

Iput file
1.112.223445555MFlubber
9.998.887664444TGrungit

Output file
1.112.223445555MFlubber
9.998.887664444TGrungit

zeitfest
Guru

Joined: 31/07/2019
Location: Australia
Posts: 549

Posted: 05:34am 17 Feb 2025

Copy link to clipboard

Print this post

The csv concurrent file i/o is a goer. The files are declared as csv. The * specifies default formats for the individual fields which can be overridden by a declaration and so on.

As a test this reads a line of csv file data into variables and then writes/appends it to another file in csv format, then repeats. As a check the values are printed on screen so that I can make sure it works ok !! It also tests the implied-list handling.

PROGRAM testread
INTEGER*4 n(5), x

OPEN "/RepImp.txt" WRITE CSV
OPEN "/InImp.txt" READONLY CSV

READE (7,*) ( n(x), x = 1, 5 )
WRITE (8,*) ( n(x), x = 1, 5 )

PRINT n(1), n(2), n(3), n(4), n(5)

READE (7,*) ( n(x), x = 1, 5 )
WRITE (8,*) ( n(x), x = 1, 5 )

PRINT n(1), n(2), n(3), n(4), n(5)
END
\

>R
Running
...

open file write
open file read
readin
54321
readin
246810

OK
...

Input file on 7. InImp.txt is :

5,4,3,2,1
2,4,6,8,10

Output file on 8, RepImp.txt is :

5,4,3,2,1
2,4,6,8,10

zeitfest
Guru

Joined: 31/07/2019
Location: Australia
Posts: 549

Posted: 04:37am 03 Mar 2025

Copy link to clipboard

Print this post

Added a small graph display, just 100x200 pixel but good for a quick optic.
It is fed an array of y value integers scaled 0 to 100, assumes x 0-100 integer inclusive scale wide so 101 values - needs some tweaks yet.

Eg this calculates sin values for three cycles and displays it.

>R
Running

PROGRAM steffi
DOUBLE x, y
INTEGER*4 j, yarts(101)
x = 0.0
DO j = 1, 101
y = 50.0 + 50.0 * SIN(x)
x = x + 0.0628 * 3.0
yarts(j) = IFIX y
END DO
GRAPH yarts(101)
PRINT "finish"
END
\

Edit - replaced photo

Edited 2025-03-04 10:16 by zeitfest

zeitfest
Guru

Joined: 31/07/2019
Location: Australia
Posts: 549

Posted: 08:13am 04 Mar 2025

Copy link to clipboard

Print this post

Also added access to the analog-digital converter, just as a basic nicety ( we can't have graphs without measurements now, what in the world rant etc

)

Just using as a keyword in the first instance eg ADC0 gets the value from the A/D measuring on analog pin 0.

Running

PROGRAM sousvide
INTEGER*4 i, j(101)
DOUBLE a
DO i = 1, 101
a = DBLE( ADC0 ) / 1024.0
j(i) = IFIX ( a * 100.0 )
END DO
GRAPH j(101)
END

The screen result from above shows about one and half cycles of background humm at 50 Hz as present on my thumb so that is the loop repeating at a bit over 3kHz by default.

The A/D can be operated much faster, the loop is not intended for speed, but it is ok for the time being. I think the A/D is being switched on and off each loop which is not optimal. OK for a "calcometer" so far

Edited 2025-03-04 18:16 by zeitfest

zeitfest
Guru

Joined: 31/07/2019
Location: Australia
Posts: 549

Posted: 02:21pm 01 Apr 2025

Copy link to clipboard

Print this post

Now I realize the quiet push-buttons used are the ones suggested by Franknfurter,
excellent idea thanks !!!

A problem with DIY keys is putting a logo on each. I tried labels/glue/foam/etc but the results were not robust and didn't look very good. However I noticed a photo service shop would print photos onto a "card" which turned out to be laminate.
It has two thin aluminum sheets bonded to a plastic core, the photo is scanned and printed on it, the ink is light-cured and robust. Very easy to work with and not expensive. So I got the key logos printed onto a "card" and and stuck them on the keys with bonding adhesive, the keys destruct before the logo can be separated. Needs a suitable case but looks good so far.

Also added a "line continuation" , ie so an equation can be continued on
the following line. Otherwise the equation had to be split which was lousy.

eg the equation

C
C el = ASIN ( SIN ( dec ) * SIN ( lat * rpd ) +
C COS ( dec ) * COS ( lat * rpd ) * COS ( ha ) )
C

had to be split and reconstituted

temp = COS ( dec ) * COS ( lat * rpd ) * COS ( ha )
temp = temp + SIN ( dec ) * SIN ( lat * rpd )
el = ASIN ( temp )

but now the standard continuation process is used instead

el = ASIN ( SIN ( dec ) * SIN ( lat * rpd ) +
+ COS ( dec ) * COS ( lat * rpd ) * COS ( ha ) )

That just about finishes any additions, now just the debugging

Edited 2025-04-02 00:26 by zeitfest

Volhout
Guru

Joined: 05/03/2018
Location: Netherlands
Posts: 4854

Posted: 04:58am 02 Apr 2025

Copy link to clipboard

Print this post

Hi zeifest,

At the moment you are using I2C for keyboard and lcd. Since the development is almost complete, are you interesten in a simplified version where the pico does It all. Key scan, and direct drive the lcd. And a low power circuit, so It can sleep weeks ad month until you press the ON push button.

Volhout

PicomiteVGA PETSCII ROBOTS

zeitfest
Guru

Joined: 31/07/2019
Location: Australia
Posts: 549

Posted: 11:19am 02 Apr 2025

Copy link to clipboard

Print this post

Is there a project you have at the moment, post some info ?

The keypad i2c/TCA8418 is good and lives in the event loop ok, and also ties it in to supplier chains. But the display via i2c is pretty slow so an alternate path would be good. Maybe a second pico to drive the interface..

Driving a moderate display can be an iceberg. Updating a few fields from a program is straightforward but maintaining state (so that the data is consistent) can be tricky. ATM it is driving a 3.2" display, maintaining state of data fields etc, but it is overkill for a calculator. In fact I looked at a cut down version with a 16x2 lcd, but the lcd needed 5v though. But it would have fitted nicely on a cheap 4"" pcb ...

It is heavily based in open-source and Arduino so there is no commercial direction.

ed - A pcb with pico with 16x2 lcd /sd micro /i2c probably exists out there already (?)

More thinking to do

Edited 2025-04-02 21:35 by zeitfest

Page 3 of 3

Print this page

To reply to this topic, you need to log in.