I don't get the delay at all
#Arduino Help
coral geyser
I don't get the delay at all
We wait for 100 ms to make sure the key is actually not pressed. instead of just a bounce in the switch
wary cloak
Since all methods have this now, shouldn't get key also?
coral geyser
because when you press a mechanical switch the waveform looks like
wary cloak
pub fn get_key(&mut self) -> Option<char> {
for col in 0..self.columns.len() {
for row in 0..self.rows.len() {
if self.check_row_col(row, col) == true {
return Some(self.matrix[row][col]);
}
}
}
None
}
we can now simplify get_char_array
coral geyser
```rust
pub fn get_key(&mut self) -> Option<char> {
for col in 0..self.c...
get_key is fine because we just ask for the current state
wary cloak
we can now simplify `get_char_array`
How?
coral geyser
How?
for index in 0..buffer.len() {
match self.wait_for_Key() {
'a' => { return index; }
key => {
// put key in the buffer
buffer[index] = key;
}
}
}
buffer.len()
Because we get rid of all the asking if there's nothing pressed?
coral geyser
Because we get rid of all the asking if there's nothing pressed?
yep, now wait_for_key handles that
wary cloak
And with the for loop I guess we no longer need the if and so..
Ok done
So know we finally write code to main?
wary cloak
Yes unless you think of another cool stuff we could do or need but I don't think so
main is like this now
#![no_std]
#![no_main]
mod Keypad;
use panic_halt as _;
#[arduino_hal::entry]
fn main() -> ! {
let dp = arduino_hal::Peripherals::take().unwrap();
let pins = arduino_hal::pins!(dp);
let serial = arduino_hal::Usart::new(
dp.USART0,
pins.d0,
pins.d1.into_output(),
arduino_hal::usart::Baudrate::new(57600),
);
loop {
}
}
I came back from dinner
How tasty
So we should test the three methods and it would be done
Any ideas?
I guess
First I would need to make a variable that stores the matrix
But about the pins
coral geyser
in the loop wait for a string then when you get one print it to the serial?
wary cloak
But about the pins
How do I do it?
First let me check for the pins.
Which ones were
coral geyser
How do I do it?
let rows = [
pins.xx.into_pull_up_input().downgrade(),
pins.xx.into_pull_up_input().downgrade(),
pins.xx.into_pull_up_input().downgrade(),
pins.xx.into_pull_up_input().downgrade(),
];
const char keys[FILAS][COLUMNAS] = {
{ '1', '2', '3', 'A' },
{ '4', '5', '6', 'B' },
{ '7', '8', '9', 'C' },
{ '*', '0', '#', 'D' }
};
const byte pinColumnas[COLUMNAS] = {22, 23, 24, 25};
const byte pinFilas[FILAS] = {26, 27, 28, 29};
Here is the code from the Arduino
wary cloak
```rust
let rows = [
pins.xx.into_pull_up_input().downgrade(),
pins.xx.into_...
Oh ok
Can I copy the declaration of the matrix as is? (Changing filas and columnas for 4)
coral geyser
Can I copy the declaration of the matrix as is? (Changing filas and columnas for...
yes, and replace the {} with []
wary cloak
```rust
let rows = [
pins.xx.into_pull_up_input().downgrade(),
pins.xx.into_...
Should we anotate the type alias row?
coral geyser
Should we anotate the type alias row?
we don't need to
coral geyser
Do I make it constant also?
No just a let keys = ...;
wary cloak
At the end or separating the arrays?
coral geyser
let keys = [
[ '1', '2', '3', 'A' ],
[ '4', '5', '6', 'B' ],
[ '7', '8', '9', 'C' ],
[ '*', '0', '#', 'D' ],
];
Ok now the pins
let rows = [
pins.d26.into_pull_up_input().downgrade(),
pins.d27.into_pull_up_input().downgrade(),
pins.d28.into_pull_up_input().downgrade(),
pins.d29.into_pull_up_input().downgrade(),
];
Why downgrade?
coral geyser
Why downgrade?
that makes them use Dynamic
so you can put them all in an array together
coral geyser
And how do I make the others open drain outputs?
into_opendrain() it's in the docs
wary cloak
let columns = [
pins.d22.into_opendrain().downgrade(),
pins.d23.into_opendrain().downgrade(),
pins.d24.into_opendrain().downgrade(),
pins.d25.into_opendrain().downgrade(),
];
And we can finally call new
coral geyser
```rust
let columns = [
pins.d22.into_opendrain().downgrade(),
p...
should be able to yeah
coral geyser
run cargo check
wary cloak
Checking keypad v0.1.0 (C:\Users\fuchi\OneDrive\Documentos\Codigos_Rust\keypad)
error[E0425]: cannot find function `new` in module `Keypad`
--> src\main.rs:42:26
|
42 | let keypad = Keypad::new(rows, columns, keys);
| ^^^ not found in `Keypad`
For more information about this error, try `rustc --explain E0425`.
error: could not compile `keypad` due to previous error
impl Keypad {
pub fn new(rows: RowPins, columns: ColumnPins, matrix: Matrix) -> Self {
Self {
rows,
columns,
matrix,
}
}
It doesn't make sense to me
coral geyser
oh, it thinks Keypad is a module
wary cloak
Isn't it?
coral geyser
keypad::Keypad::new()
wary cloak
Wait now that I think about it
I didn't make a use ::crate::Keypad::*
Or idk how it was it
Let's go one by one
wary cloak
oh, it thinks Keypad is a module
Is it not?
coral geyser
Is it not?
Keypad isn't keypad is
wary cloak
Yes
wary cloak
But without let to declare a variable to store it then..?
wary cloak
I didn't make a use ::crate::Keypad::*
Couldn't I use this?
Or I mean
Why does this happen
Because it's named the same the struct and module?
coral geyser
Couldn't I use this?
you can do that also
though... it has a name conflict then
which is why modules are usually snake case
wary cloak
Yeah but it is one word(?
I could call it Matricial_Keypad
And the use the use statement above
coral geyser
How do I make the use statement I tried to refer to?
use matricial_keypad::*; ?
wary cloak
Yeah ig
Idk why but it moved places
Pls
How do I get the module about main
Above*
coral geyser
vscode orders them alphabetically
wary cloak
Oh that's why
coral geyser
it helps a lot when you have like 20-100 files in one folder
wary cloak
Ok finally
We can move to loop
And test
I guess what I would do
Is create a variable that stores the result from calling get key
And then print it to serial port
Same for the rest basically
coral geyser
If you use the get string one, you can test all 3 at the same time
coral geyser
If you use the get string one, you can test all 3 at the same time
since they call each other
coral geyser
Initialize it?
yeah, it has to have something in it
wary cloak
The second one is the size right?
coral geyser
yep
wary cloak
Ok time to get testing
Ok now I just have to remember how to print things I will see the last project
How did I solve this?
coral geyser
How did I solve this?
you will need to print each char at a time
because it's not a &str
wary cloak
But what about the second that is a usize?
coral geyser
But what about the second that is a usize?
actually you can do
uwriteln!(&mut serial, "{:?}", buffer).void_unwrap();
uwriteln!(&mut serial, "{}", buffer_size).void_unwrap();
I think that works
coral geyser
what does cargo check say for the second error?
wary cloak
It wrote so many stuff..
coral geyser
wow they forgot the char impl 
wary cloak
Checking keypad v0.1.0 (C:\Users\fuchi\OneDrive\Documentos\Codigos_Rust\keypad)
error[E0277]: the trait bound `char: uDebug` is not satisfied
--> src\main.rs:50:9
|
50 | uwriteln!(&mut serial, "{:?}", buffer).void_unwrap();
| ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ the trait `uDebug` is not implemented for `char`
|
= help: the following other types implement trait `uDebug`:
&T
&mut T
()
(A, B)
(A, B, C)
(A, B, C, D)
(A, B, C, D, E)
(A, B, C, D, E, F)
and 76 others
= note: required because of the requirements on the impl of `uDebug` for `[char; 16]`
= note: this error originates in the macro `proc_macro_call` (in Nightly builds, run with -Z macro-backtrace for more info)
error[E0599]: no method named `void_unwrap` found for enum `Result` in the current scope
--> src\main.rs:50:48
|
50 | uwriteln!(&mut serial, "{:?}", buffer).void_unwrap();
| ^^^^^^^^^^^
|
::: C:\Users\fuchi\.cargo\registry\src\github.com-1ecc6299db9ec823\void-1.0.2\src\lib.rs:87:8
|
87 | fn void_unwrap(self) -> T;
| ----------- the method is available for `Result<(), void::Void>` here
|
= help: items from traits can only be used if the trait is in scope
help: the following trait is implemented but not in scope; perhaps add a `use` for it:
|
4 | use arduino_hal::prelude::*; // trait ResultVoidExt
|
help: there is an associated function with a similar name
|
50 | uwriteln!(&mut serial, "{:?}", buffer).unwrap();
| ~~~~~~
error[E0599]: no method named `void_unwrap` found for enum `Result` in the current scope
--> src\main.rs:52:51
|
52 | uwriteln!(&mut serial, "{}", buffer_size).void_unwrap();
| ^^^^^^^^^^^
|
::: C:\Users\fuchi\.cargo\registry\src\github.com-1ecc6299db9ec823\void-1.0.2\src\lib.rs:87:8
|
87 | fn void_unwrap(self) -> T;
| ----------- the method is available for `Result<(), void::Void>` here
|
= help: items from traits can only be used if the trait is in scope
help: the following trait is implemented but not in scope; perhaps add a `use` for it:
|
4 | use arduino_hal::prelude::*; // trait ResultVoidExt
|
help: there is an associated function with a similar name
|
52 | uwriteln!(&mut serial, "{}", buffer_size).unwrap();
| ~~~~~~
Some errors have detailed explanations: E0277, E0599.
For more information about an error, try `rustc --explain E0277`.
error: could not compile `keypad` due to 3 previous errors
add use arduino_hal::prelude::*; ?
coral geyser
you will need to print each char at a time
alright we are back to this
coral geyser
you need a use 
wary cloak
That solve it for the second one
coral geyser
alright we are back to this
^
coral geyser
for c in &buffer {
uwrite!(&mut serial, "{}", c).void_unwrap();
}
Ok seems to be done
Also seems 0 warnings
So time to finally actually test
Mmm..
It finally executed so I pressed
From 1 to 9
Then # which is the one I choose to end
And it seems as nothing happened
Oh
I didn't change your 'a'
In the new version
It still didn't do anything
Now I'm going to press each one just in case
Nothing
I'm gonna execute it again with my change
coral geyser
try pressing them 20+ times so that it triggers the auto stop
wary cloak
Because maybe pressing A gives an 'A' and not an 'a'
wary cloak
try pressing them 20+ times so that it triggers the auto stop
Ok now I'm going to press a lot
coral geyser
Because maybe pressing A gives an 'A' and not an 'a'
yeah that is an issue
wary cloak
yeah that is an issue
That's the reason I changed to #
But forgot too when we simplified
I pressed a lot of times and nothing happened
So I'm gonna re run with my change just in case
Ok let's see
No nothing happened
I pressed 0 to 9
And then #
Nothing
wary cloak
And pressing a lot of times also doesn't work
Ok
let key = keypad.get_key();
uwriteln!(&mut serial, "{}", key).void_unwrap();
it's going to not like it
coral geyser
if let Some(key) = keypad.get_key() {
uwriteln!(&mut serial, "{}", key).void_unwrap();
} else {
uwriteln!(&mut serial, "none").void_unwrap();
}
Ok let's see
thread '<unnamed>' panicked at 'called `Result::unwrap()` on an `Err` value: Error { kind: InvalidData, message: "Windows stdio in console mode does not support writing non-UTF-8 byte sequences" }', C:\Users\fuchi\.cargo\registry\src\github.com-1ecc6299db9ec823\ravedude-0.1.4\src\console.rs:27:45
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
Didn't touch anything
coral geyser
```rust
thread '<unnamed>' panicked at 'called `Result::unwrap()` on an `Err` va...
um well then
coral geyser
```rust
thread '<unnamed>' panicked at 'called `Result::unwrap()` on an `Err` va...
if let Some(key) = keypad.get_key() {
uwriteln!(&mut serial, "{}", key as u8).void_unwrap();
} else {
uwriteln!(&mut serial, "none").void_unwrap();
}
maybe
wary cloak
Oh I thought the problem was with the function and all
coral geyser
apparently the arduino sent back invalid utf8
wary cloak
That last part I didn't get
coral geyser
as it's writing so fast
wary cloak
Oh
coral geyser
The serial port on windows can only hold so much data
try adding a delay_ms(500); in the loop
wary cloak
Which one because there are 2 for loops
pub fn get_key(&mut self) -> Option<char> {
for col in 0..self.columns.len() {
for row in 0..self.rows.len() {
if self.check_row_col(row, col) == true {
return Some(self.matrix[row][col]);
}
}
}
None
}
in main.rs
wary cloak
Oh
coral geyser
between calls to get_key
because right now it's running at like 1 million attempts per second
coral geyser
just a long list of none?
wary cloak
Wouldn't it tell me if it paniced?
coral geyser
Wouldn't it tell me if it paniced?
not on the arduino
wary cloak
```rust
pub fn get_key(&mut self) -> Option<char> {
for col in 0..self.c...
Well here is get key to see
coral geyser
it just locks up if that happens
coral geyser
Well here is get key to see
add a uwriteln!(&mut serial, "hi").unwrap(); just before you call get key
wary cloak
I finished the program but the rx led is blinking
coral geyser
I finished the program but the rx led is blinking
that means it's getting data
from the computer
tx would be sending to the computer
wary cloak
it is so annoying to go copy this each time cargo run -- --port COM6 -c
coral geyser
it is so annoying to go copy this each time cargo run -- --port COM6 -c
you can press the up key
to get your history of commands
coral geyser
is the tx led blinking?
coral geyser
in any case lets simplify even more
loop {
delay_ms(500);
uwriteln!(&mut serial, "hi").unwrap();
}
an comment out the making of the keypad and the pin setup
wary cloak
Wait
Now it printed
Maybe it's because of how slow my computer is? Remember when I said the the lcd took it's time to answer to my commands? Tho not this long
hi
none
hi
none
hi
none
hi
none
hi
none
hi
Goes like this for a while
coral geyser
what happens if you press a key?
wary cloak
Tho I don't see it you know
Keep going
It just appeard out of knowhere
And it's doing nothing again
wary cloak
what happens if you press a key?
Let's see
coral geyser
maybe we overloaded the windows serial driver
wary cloak
I pressed all keys but I didn't see the tx turn on and it didn't do anything
It's likely because it has a million more none's to print
Do we maybe try wait for key?
coral geyser
in any case lets simplify even more
```rust
loop {
delay_ms(500);
uwriteln!(...
try doing this ^
wary cloak
Ok
coral geyser
we need to get to a known working state
wary cloak
Mmm..
It doesn't work
Or again it's going to take like 5 minutes and then a lot appear
I have a suggestion
Mmm ..
I mean
I could call get_key, bind it to a variable and then do a match expression maybe where either result makes a break, a catch all that says I received none and another one for a specific key idk..
Thing is
We are printint to much to the serial port ig
Printing*
Too fast*
I'm gonna try again with a delay like 5000000000
Or smth
To see if it prints one time at least
Well it allows 50000 only more is out of bounds
And it also didn't work
Btw now rx is no longer blinking-
Oh there it is
One hi
Oh ofc
The delay is before
So it waited 50000 miliseconds
And then printed hi
And it did again :o
Weird this time it took like way less
Oh when I sent this
It's now another one
As I thought the problem is just how fast we are printing ig?
Ig last time
Since it printed it once and then entered another loop
It didn't happen
So knowing this and having tested that..
What do we do now?
Even know it really takes it's time
And I can't use a much higher delay anyways
Actually
It haven't printed another one
Is rather inconsistent
Oh
It printed so more and I didn't notice
13 times
Now it printed more
So yeah
It seems to be doing fine
With almost the highest delay I can perform.
So now that we know and tested this.. What now?
wary cloak
You know I think that it may work for what we need it, it likely works but no clue how to test it
Could maybe test driven development do it? Like calling wait_for_key in a asserteq! And compare it to an specific key I know I will press idk
Because it seems like for the non-waiting nature of get_key it seems useless or impossible to test
I think testing wait_for_key which is the one we will indeed use should do and if it works then get key too
wary cloak
Or maybe we could try them in setup so it's just once
coral geyser
Sort of
wary cloak
Noooo ai wasn't heree
wary cloak
Sort of
Well what do you think of my results yesterday ?
Just a little bit after you left with it working with the highest delay possible
And either the test driven development solution or try it in setup solution
wary cloak
You know I think that it may work for what we need it, it likely works but no c...
And this conclusion
coral geyser
Well what do you think of my results yesterday ?
I still have no idea why the serial doesn't working properly. What about this, go back to a 500 ms delay, program the mega, then open the Arduino ide and open it's console viewer
wary cloak
I was using serial just now
For the normal arduino version of this project
Trying to figure it why some code related to the rfid isn't working, I'm frustrated-
coral geyser
Yeah, so I am wondering if the ravedude console is messed up on windows
wary cloak
Hmm
That's why I said maybe we should try it in setup or smth because maybe we are losing time for something that already works
coral geyser
You can do that also. It's up to you as it's your project. I just want to know why it doesn't work properly
wary cloak
We can do all test you want if you have another idea
coral geyser
I still have no idea why the serial doesn't working properly. What about this, g...
^ would be one
wary cloak
What but
Wait*
Would that work I mean
I just open an arduino IDEA
Ide*
With no project
And open the console?
coral geyser
Yep
wary cloak
While the program is running
It works if it has a project?
I don't want to create a new one
coral geyser
The Arduino IDE should auto connect
coral geyser
I don't want to create a new one
You can just leave it on the blank one it makes or whatever you have open
wary cloak
If it's just for opening the ide
Ok now I open vs
arduino_hal::delay_ms(500);
uwriteln!(&mut serial, "hi").unwrap();
this one?
coral geyser
Yep
coral geyser
Did you ctrl-c to stop the ravedude console?
coral geyser
Cool, so it's just the ravedude console
And I think I know why
Go back to the rust code and add a \r to the end of hi
Then program it again
wary cloak
Do we test the original with buffer buffer size the for c in buffer and uwriteln?
wary cloak
Go back to the rust code and add a \r to the end of hi
Om
Ok*
It's still the same
coral geyser
It's still the same
Did it actually program? It can't while the Arduino window is open
wary cloak
I should close my arduino IDE windows?
coral geyser
I should close my arduino IDE windows?
Just the console window
wary cloak
I did that
coral geyser
And check that the command actually programmed (scroll up)
wary cloak
I mean I looked at it trigger the warnings for unused code and all but then I see this
coral geyser
Ok, so it's not going to be a simple fix 
wary cloak
I mean so we need to?
coral geyser
We will just use the Arduino console
wary cloak
let mut buffer = ['x'; 16];
let buffer_size = keypad.get_char_array(&mut buffer);
for c in &buffer {
uwriteln!(&mut serial, "{}", c).void_unwrap();
}
uwriteln!(&mut serial, "{}", buffer_size).void_unwrap();
this one
Well it didn't work
Ok then let's test
if let Some(key) = keypad.get_key() {
uwriteln!(&mut serial, "{}", key as u8).void_unwrap();
} else {
uwriteln!(&mut serial, "none").void_unwrap();
}
arduino_hal::delay_ms(500);
This one
coral geyser
```rust
if let Some(key) = keypad.get_key() {
uwriteln!(&mut serial,...
Don't forget to hold the key for at least 500 Ms for it to read it again
wary cloak
It keeps printing none
No matter how much times I press
Or keep it presses
Pressed*
wary cloak
Yeah because it works in the arsuino version I tested it last Wednesday
And the code worked with that pins
wary cloak
```rust
let columns = [
pins.d22.into_opendrain().downgrade(),
p...
.
There you have in any case
coral geyser
Hmm, maybe we test check_row_col
wary cloak
For now I need to leave, eat something and bath after that I continue
And also charge the computer while all of that happens
But could you leave me what we are going to do?
Like
The code to test because
Idk how we could test that function
Btw I have a suggestion
Even tho my c program works and it's good reference to see how to detect and print a key
Maybe we could also go to
The git repository
Of either the keypad.h Library I used for arduino
To see even tho in c++ how he did it
Or the one from the 4x3 one because it should be similar
wary cloak
@coral geyser aaaand I'm finally back, we can continue if you are
coral geyser
Please continue
wary cloak
Hmm, maybe we test check_row_col
This
Which is our most primitive function
I mean maybe we could print the result of calling our function which is a boolean and so start pushing first other keys to see if it's always false and then push the one I have to and see if it's true or smth like that
Just for refresh imma send it again
fn check_row_col(&mut self, row: usize, col: usize) -> bool {
for (i, pin) in self.columns.iter_mut().enumerate() {
if i == col {
pin.set_low();
} else {
pin.set_high();
}
}
self.rows[row].is_low()
}
Sounds good
wary cloak
How do you that AP?
Btw I will have to make the function public for now
uwriteln!(&mut serial, "{}", keypad.check_row_col);```Why is this wrong?
coral geyser
```rust
uwriteln!(&mut serial, "{}", keypad.check_row_col);```
Check_row_col needs it's params
wary cloak
```rust
fn check_row_col(&mut self, row: usize, col: usize) -> bool {
fo...
It takes as parameters a position of the array right?
So
0 to 3
I guess
Let's make A the one, so row 1 column 3
coral geyser
Yep
wary cloak
uwriteln!(&mut serial, "{}", keypad.check_row_col(1,3)).void_unwrap();
here it is
Ok let's see
thread '<unnamed>' panicked at 'called `Result::unwrap()` on an `Err` value: Error { kind: InvalidData, message: "Windows stdio in console mode does not support writing non-UTF-8 byte sequences" }', C:\Users\fuchi\.cargo\registry\src\github.com-1ecc6299db9ec823\ravedude-0.1.4\src\console.rs:27:45
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
Does this mean I don't try the arduino ide?
wary cloak
It's printing to fast
coral geyser
Delay_ms(...) 
wary cloak
:0
wary cloak
Delay_ms(...) <:ferrisGlasses:451162415850323999>
Yeah I forgot but
It works
Tho it seems I miss calculates
Calculated*
If I press B, not A
I see true
If nothing false
Let me see if nothing else triggers true
coral geyser
So we can detect buttons
wary cloak
```rust
pub fn get_key(&mut self) -> Option<char> {
for col in 0..self.c...
.
Thing is.. What's wrong?
coral geyser
Oh I think I know
wary cloak
The fact that is an optional? It makes sense for it to be, how we print it?
Oh tell me
coral geyser
We didn't take into account the speed of electricity
It's switching the pins to fast
So they haven't finished being set to 0v before it reads the input pin
coral geyser
So if we add a small delay, something like delay_us(100); between the setting low in check_row_col and reading the input it should work
wary cloak
That it changes state really fast and by the time we read it is no longer low..?
coral geyser
I didn't understand
The mega runs at 16 MHz which allows it to set the pin low, then read the input pin so fast that the electrical state of the wire hasn't changed yet
wary cloak
I get when talking about of state of the pin but, state of the wire?
coral geyser
That it changes state really fast and by the time we read it is no longer low..?
The lowness hasn't reached the input pin
coral geyser
I get when talking about of state of the pin but, state of the wire?
The voltage change flows across the wire (to the keypad, button, back to the mega) in some amount of time
Happens that that time is to slow for what we are doing
wary cloak
I really can't understand at all I just get that arduino too fast, keypad not so much?
It's a
Keypad thing
Or
coral geyser
It's a physics thing
wary cloak
Arduino is too fast for us
coral geyser
Arduino is too fast for us
We can go with that
coral geyser
Just that?
Yep, though I would put it after the for loop
I think that will behave better
wary cloak
So before the is low question?
coral geyser
So before the is low question?
Yep
wary cloak
pub fn check_row_col(&mut self, row: usize, col: usize) -> bool {
for (i, pin) in self.columns.iter_mut().enumerate() {
if i == col {
pin.set_low();
} else {
pin.set_high();
}
}
arduino_hal::delay_ms(500);
self.rows[row].is_low()
}
So now we try
wary cloak
```rust
if let Some(key) = keypad.get_key() {
uwriteln!(&mut serial,...
This again?
coral geyser
Basically what is happening is we set the column to 0v, but by the time the input is read it's at like 4.2v and dropping
coral geyser
This again?
Yep
wary cloak
Oh so..
In check row column
We ask for a column and if it's not we set it to 0
Really fast
The problem being
That then we immediately say
Hey it's 0?
And all that before it could physically be 0
Right?
wary cloak
Well now we now check_row_col works now it's just a matter of get key
coral geyser
And all that before it could physically be 0
(as a side note, this has to do with the capacitance of the wire and the internal parts of the microcontroller, it takes time to drain the small amount of charge the capacitance holds)
wary cloak
Ok so for now is printing none slower than before
coral geyser
Well now we now check_row_col works now it's just a matter of get key
And the reason we didn't see it with check_row_col is because then we weren't changing which column was set to 0 so it had time
coral geyser
Ok so for now is printing none slower than before
Any chars 
wary cloak
Let the program speak for it itself
"none"
But we got a lower none printing speed rate(?
coral geyser
```rust
pub fn check_row_col(&mut self, row: usize, col: usize) -> bool {
...
Oh, you set it to 500 ms
wary cloak
?
wary cloak
That slow is that many mili(?
coral geyser
That slow is that many mili(?
0.1 milli seconds
coral geyser
Why such a reduction in time
Because we need enough time for the physics to happen but be fast enough you don't notice the delay
coral geyser
Still none
Hmm, but faster none printing?
wary cloak
Yes
coral geyser
Try delay_ms(10);
wary cloak
I'm not sure how this is going to make it work-
coral geyser
If it doesn't I have more ideas for things to check
I would normally measure the time delay we need, but ... I am not there 🤷♂️
wary cloak
Sure I was just curious on the logic behind that being the issue for learning purposws
Still none
God damn
coral geyser
K, now setup a test with 2 check_row_col at different spots
wary cloak
What
wary cloak
Oh
uwriteln!(&mut serial, "{}", keypad.check_row_col(1,3)).void_unwrap();
uwriteln!(&mut serial, "{}", keypad.check_row_col(3,2)).void_unwrap();
Like that?
Now it doesn't work
Or it's too fast for me to see the true
uwriteln!(&mut serial, "{}", keypad.check_row_col(1,3)).void_unwrap();
arduino_hal::delay_ms(500);
uwriteln!(&mut serial, "{}", keypad.check_row_col(3,2)).void_unwrap();
arduino_hal::delay_ms(500);
Hmm
wary cloak
Like that?
coral geyser
Yeah
wary cloak
Wtf doesn't it work now-
It doesn't make sense -
Aha
Mm
I re opened and pressed really fast and saw only one true
coral geyser
Hold the key
wary cloak
I did the same for 8 which should be the other unless I'm wrong again
And it didn't work
Now I'm going to see if what I did to B works again first
I did some tries that didn't work
Now I made it again
If I hold it
I still only see false
Let's see if so that of opening already pressing but holding
It works
coral geyser
I wish we had an oscilloscope
wary cloak
If I hold the button
And enter the console
It says true
But only once
It doesn't make any sense to me-
And also it works with B not 8
coral geyser
It doesn't make any sense to me-
Yeah, same here
wary cloak
So it do seems that the problem still is in check row col
coral geyser
Go back to 1 check_row_col
wary cloak
But it's just
With 1 it works, with 2 not, but if I hold press is true only before entering and then never more but with the other it doesn't work
wary cloak
Go back to 1 check_row_col
Ok
It works perfectly fine
Holding or not
Mmm
Sometimes I pressed and it didn't
Now I pressed three times
And then it started working fine with each press
But those before didn't work
Now I pressed a lot of times
It's kind of inconsistent to start detecting
coral geyser
That's expected with the delay
wary cloak
But holding seems to he consistent
coral geyser
It only reads it every 0.5 s
wary cloak
It works with one at a time tho isn't that how it normally works anyway?
coral geyser
Set that delay to like delay_ms(100) and you should see more consistent results
Because it will read fast and report more often
wary cloak
Sure but what do we do now
coral geyser
But holding seems to he consistent
The only thing I can think of is that the file with check_row_col didn't get saved. That's it
coral geyser
Yeah
wary cloak
And also I had the slower none
coral geyser
But like I have done this exact thing many times and never seen this
wary cloak
I'm sending the function again in any case
coral geyser
And I even double checked the rust keypad crate and the Arduino keypad library
wary cloak
pub fn check_row_col(&mut self, row: usize, col: usize) -> bool {
for (i, pin) in self.columns.iter_mut().enumerate() {
if i == col {
pin.set_low();
} else {
pin.set_high();
}
}
arduino_hal::delay_ms(10);
self.rows[row].is_low()
}
I am thinking
If get key gives back wrong the key pressed but that isn't it because even if it did it would show the key just the incorrect one
But I'm gonna express it anyways
pub fn get_key(&mut self) -> Option<char> {
for col in 0..self.columns.len() {
for row in 0..self.rows.len() {
if self.check_row_col(row, col) == true {
return Some(self.matrix[row][col]);
}
}
}
None
}
let keys = [
[ '1', '2', '3', 'A' ],
[ '4', '5', '6', 'B' ],
[ '7', '8', '9', 'C' ],
[ '*', '0', '#', 'D' ],
];
Say I pressed 2
So col would 1 at it's second iteration
And row would be 0 at it's first iteration
So it would give back
Some(self.matrix[0][1])
...
Why when I thought about it
It seemed wrong
But now I see it's fine no way-
Well forget all you just saw
Anyways if you needed it there are the two criminals check and get
coral geyser
I am tempted to pull out my oscilloscope to check this all
coral geyser
It will take me a minute to wire everything up for it
wary cloak
It's not like I'm gonna be using it while chatting on discord and stuff
wary cloak
It will take me a minute to wire everything up for it
Sure I wait
Man so many thing I gave for granted with arduino
Btw..should I upload something else to the board? Because idk if it's good to keep communication for so long, seeing that led blinking non stop get's me worried
coral geyser
could you zip up the entire project and send it to me?
wary cloak
How do I?
coral geyser
in windows explorer, right click the folder, then send to, compressed (zip) folder
without the target/ folder preferably
wary cloak
alright I can replicate the behavior
wdym?
coral geyser
wdym?
I setup my mega with a 4x4 keypad to match yours and it behaves the same way (only false with 2 check row col)
wary cloak
so lucky to meet a real embedded developer with stuff at home(?
coral geyser
yep
wary cloak
in our case always false?
coral geyser
yep
coral geyser
I think I found the issue
wary cloak
Who was the agent of evil?
coral geyser
There is a small difference with mega chips, their pins default to high when you switch them to output
The hal doesn't take this into account so when we use open drain instead of grounding it goes to 5v
wary cloak
wh-
Why the hell
Is that a thing
Why when they were doing the mega they thought
Yeah
Output defaults to 5 volts
wary cloak
The hal doesn't take this into account so when we use open drain instead of grou...
Tho I didn't understand the issue at all
So let's see
Normally
coral geyser
this is what we want
wary cloak
A microcontroller would have output a 0v by default unless I do something because it's idk what it makes sense
So open drain
Disconnects them completely
And when low they are nothing when high they are 0
But in this case
coral geyser
this is what we get
wary cloak
Since MEGA is a special boy
He decided that outputs are by default 5volts unless we tell them otherwise
But.
How does this affect open drain?
I told until the part that made sense to me
Or I thought I did understand
coral geyser
how open drain is implemented on the mega is it switches from output to input
output for the ground
input for floating
coral geyser
It's hard to understand because of how it doesn't make sense
yeah
coral geyser
so when it sets it up, it makes it a output set to 0v. So then everything worked, but as soon as we set it to high (open) it becomes an input that just floats. But setting back to low makes it an output at 5v. Hence why as soon as we add another check_row_col everything breaks
wary cloak
Wait so..
I get mixed all up
So low instead of disconnecting
Set the output to 0 volt?
wary cloak
Why didn't I get a short circuit? Luck?
coral geyser
Why didn't I get a short circuit? Luck?
Its all right, but now the outputs don't pull the inputs down
everything just stays at 5v
coral geyser
yep
wary cloak
Which if I remember correctly could be whatever because of noise
coral geyser
// bitMap stores ALL the keys that are being pressed.
for (byte c=0; c<sizeKpd.columns; c++) {
pin_mode(columnPins[c],OUTPUT);
pin_write(columnPins[c], LOW); // Begin column pulse output.
for (byte r=0; r<sizeKpd.rows; r++) {
bitWrite(bitMap[r], c, !pin_read(rowPins[r])); // keypress is active low so invert to high.
}
// Set pin to high impedance input. Effectively ends column pulse.
pin_write(columnPins[c],HIGH);
pin_mode(columnPins[c],INPUT);
}
this is the keypad library you are using in the arduino code
see the pin_mod and pin_writes
they do this logic
wary cloak
And when set back to low is again an output of 5 volt because of the strange switch and no longer 0 volt? Because of the input output casting?
coral geyser
And when set back to low is again an output of 5 volt because of the strange swi...
yep
coral geyser
Why one worked perfectly
because with 1 we never change the output to high
wary cloak
Oh that input remains low
In get key
All get high
And then we are done
And with two well one of them would set to high and there's no going back
Ok I think I understood the logic of our problem
Now the thing is the solution
Being aware to tell it
Hey be an output 0 in each iteration?
Or smth
coral geyser
I think the simplest thing is to clone the hal repo and make the edit to fix it
wary cloak
As long as it works and I understand the code and why it works
wary cloak
```rust
// bitMap stores ALL the keys that are being pressed.
for (byte c=0; c<s...
Keypad.h?
coral geyser
coral geyser
rename the avr-hal-.... folder to just avr-hal
coral geyser
Now?
in cargo.toml edit it to have
[dependencies.arduino-hal]
path = "avr-hal"
features = ["arduino-mega2560"]
then try a cargo check
wary cloak
in cargo.toml edit it to have
```rust
[dependencies.arduino-hal]
path = "avr-hal...
I delete all the rest?
coral geyser
I delete all the rest?
delete the old [dependencies.arduino-hal] section
coral geyser
oh, no your cargo.toml
coral geyser
Go back to your project
coral geyser
yep there
replace that
this will use your local copy of the crate instead of the one on github
wary cloak
Ok done
coral geyser
cargo check
wary cloak
Error
PS C:\Users\fuchi\OneDrive\Documentos\Codigos_Rust\keypad> cargo check
error: failed to get `arduino-hal` as a dependency of package `keypad v0.1.0 (C:\Users\fuchi\OneDrive\Documentos\Codigos_Rust\keypad)`
Caused by:
failed to load source for dependency `arduino-hal`
Caused by:
Unable to update C:\Users\fuchi\OneDrive\Documentos\Codigos_Rust\keypad\avr-hal
Caused by:
found a virtual manifest at `C:\Users\fuchi\OneDrive\Documentos\Codigos_Rust\keypad\avr-hal\Cargo.toml` instead of a package manifest.
Blocking waiting for file lock on build directory
waiting
Still waiting
Any moment now
Finally god
4 minutes
I see a green finished so
It should be fine
Now that we finally got this
What do we do?
And what does this repo has btw? What do I have know?
coral geyser
And what does this repo has btw? What do I have know?
the avr-hal contains all the library code for the mega
wary cloak
Oh-
coral geyser
What do we do?
now to fix the bug
wary cloak
That seems quite a lot-
coral geyser
go to avr-hal/avr-hal-generic/src
wary cloak
Done
coral geyser
scroll to line 347
coral geyser
/// Set pin low (pull it to GND, Output to low).
#[inline]
pub fn set_low(&mut self) {
unsafe { self.pin.make_output() }
unsafe { self.pin.out_clear() }
}
is what it should be
wary cloak
So in the end we use unsafe
wary cloak
```rust
/// Set pin low (pull it to GND, Output to low).
#[inline]
pub fn set_l...
I don't see that
But almost
coral geyser
yeah add the unsafe { self.pin.out_clear() } line to the function
wary cloak
Done
coral geyser
which ever
coral geyser
Ok then get char array because it's the "all works" Scenario
it should also say it's rebuilding avr-hal-generic
wary cloak
What did we do, I mean ik change the behavior of the function to dealt with I/O bullshit but how did we solve it
coral geyser
What did we do, I mean ik change the behavior of the function to dealt with I/O ...
we just made it force the pin to be low after it makes it an output
wary cloak
So the behavior is
The same as before
But when it comes back from input floating to output
The clear forces it to always be 0v?
coral geyser
yep
wary cloak
It works better but still not quite
I pressed
0 to 9
Then # to exit
This got printed
0
1
3
5
6
7
8
9
x
x
x
x
x
x
x
x
8
coral geyser
we probably need to change the delay in check_row_col
do be less
I have it at delay_us(1); in my version
wary cloak
Btw I don't get still how it works I mean, to match a key we needed 0v as a high, how does having a floating input that could be whatever because of a noise match?
And didn't we want to disconnect it instead of having 0 volts when low? Because now we have 0 volts when low, why no short circuit?
coral geyser
Btw I don't get still how it works I mean, to match a key we needed 0v as a high...
because the inputs are set to default to 5v
coral geyser
And didn't we want to disconnect it instead of having 0 volts when low? Because ...
because we only ever force 0v, never both 0v and 5v
wary cloak
So it's the traditional high low now?
coral geyser
So it's the traditional high low now?
no it's still inverted (0v means the button was pressed) (5v is not pressed)
wary cloak
Yeah that I get
That's rows
But columns
High is 5 volts because is a floating pin input
And low 0volt output
coral geyser
this is what we get
^ hence the weird wiggle around 5v
wary cloak
avrdude: stk500v2_ReceiveMessage(): timeout
avrdude: ser_send(): write error: sorry no info avail
avrdude: stk500_send(): failed to send command to serial port
avrdude: ser_recv(): read error: Controlador no vßlido.
avrdude: stk500v2_ReceiveMessage(): timeout
avrdude: stk500v2_getsync(): timeout communicating with programmer
avrdude done. Thank you.
Error: avrdude failed
That but more times
wary cloak
^ hence the weird wiggle around 5v
But I mean a it's supposed to work with columns being low disconnected high 0 volts to match the floating input and to avoid short circuit, now it's low 0volt and high 5 volts so I don't know how that still works
coral geyser
But I mean a it's supposed to work with columns being low disconnected high 0 vo...
the high isn't actually 5v, in that the arduino isn't providing really any current. The pin drifts high because of how the mega's circuits work
wary cloak
the high isn't actually 5v, in that the arduino isn't providing really any curre...
You said that high was a floating input right? Isn't a floating input unstable because of noise? But anyways without taking noise into account, isn't a high floating input 5 volts?
coral geyser
You said that high was a floating input right? Isn't a floating input unstable b...
a floating input is some voltage, but we don't know what without looking at the circuit it's connected to
coral geyser
Finally <:ferrisHot:920768833982169118> <:ferrisHot:920768833982169118>
sorry about that
wary cloak
Wdym sorry
coral geyser
didn't even think to question the hal library
wary cloak
It was so goddamn hard to figure this out
Because it make so much no sense
It's titanic you solved it
coral geyser
It was so goddamn hard to figure this out
Do you want to make the github issue or should I
wary cloak
a floating input is some voltage, but we don't know what without looking at the ...
Right I guess it could be whatever
So in our case it's 0?
coral geyser
So in our case it's 0?
in our case it will go to 5v, but as soon as it gets connected to 0v it will go to 0v and basically no current will flow
wary cloak
Do you want to make the github issue or should I <:ferrisHmm:857440844171575326>
You found it, know more, and know how to use GitHub so, I only created one and never used it to have my vs code linked there
wary cloak
But isn't low also 0?
coral geyser
But isn't low also 0?
yep, just a different type
coral geyser
the high one can get dragged around by the input's pullup
wary cloak
Clear makes it nothing like the drain behavior?
coral geyser
the actual low 0v can drag the input's pullup to 0v
coral geyser
Clear makes it nothing like the drain behavior?
clear sets it to a hard 0v
wary cloak
What's the difference between hard 0 volt and the usual 0 volt of our high?
coral geyser
What's the difference between hard 0 volt and the usual 0 volt of our high?
I will make a table one second
wary cloak
Ok
Tho I still don't get why our high is 0 but just thought it would be easier to understand our low
Well I'll be turning off my computer so after understanding how this works we go to sleep (?
coral geyser
| Set Output | Pressed | Input | Real Output |
|------------|---------|-------|----------------------|
| Low | No | High | Low |
| High | No | High | Between Low and High |
| Low | Yes | Low | Low |
| High | Yes | High | High |
Let me see it on the pc because cell phone doesn't make it justice
coral geyser
"Set Output" is what the code sets the output pin to be
wary cloak
Which code, ours now?
coral geyser
"Pressed" is whether the button connecting output to input is pressed
"Input" is the input pin
coral geyser
Which code, ours now?
yeah, thats what your code sets
"Real Output" is what you would measure with an oscilloscope on the output pin
wary cloak
Oh ok so when we set out output to low and there's no button press we have a high input because it turns it into a floating input, that's correct?
Wait
Join
No*
That's when high
Let's start again.
So we set our column pin which is an output to low
And no button is pressed
Why is high when low, weren't we setting it to hard 0?
coral geyser
because the button isn't pressed the output isn't connected to the input pin
like electrically
wary cloak
Yes I get that
coral geyser
and the input defaults to high from being pulled up
wary cloak
Oh
We are seeing
Rows and column behavior
Ok now I may understand
Ok so the input defaults to high because it is pull up
So in the first to cases I know it's high for that and I'm ignoring it
No button pressed yet
When my output it's low it's and absolute 0 that doesn't match with 5 volts so the real output is now
Low*
I get that
Then when we set it to high, there's a high-high match?
I mean..
Input pull up defaults to 5 volts
Now.
When we
Have our column pin high
It turn into a floating input which can have whatever voltage, right?
coral geyser
wary cloak
..
coral geyser
It turn into a floating input which can have whatever voltage, right?
yep
coral geyser
What voltages has in our case when high?
5v, but weakly
wary cloak
Why 5 volts?
coral geyser
high is 5v
wary cloak
Oh for that it's a normal pin
Ok I get that
So
We would have a 5 volts matching with another 5 volts
What happens?
coral geyser
nothing, we happy because the voltages match so no current flow
wary cloak
But since they match isn't it like a key pressed or our code it's like no if it's not two low matches it's not a key?
coral geyser
the code will consider a high as nothing being pressed
wary cloak
Ok
coral geyser
which is good when column is set to high because that's a "disabled" column
wary cloak
But what about that between low and high? Does that matter?
coral geyser
that we aren't checking right now
coral geyser
But what about that between low and high? Does that matter?
so now if a low column gets connected to a high column, the low column will force everything to 0v
But the only thing that is giving 5v (the inputs)
can only provide a tiny amount of current
so nothing bad happens
where as if the columns also provided 5v as an output they would provide enough current to be an issue
wary cloak
I was speaking about what it said in the real output but since you mention it
wary cloak
so now if a low column gets connected to a high column, the low column will forc...
How-?
coral geyser
But what about that between low and high? Does that matter?
thats just we dont know the voltage
coral geyser
No, how does it set it all to 0
ah, because the output can pull a lot of current in to ground (0v)
so it overrides the weak 5v pullup in the inputs
coral geyser
ah, because the output can pull a lot of current in to ground (0v)
(also this "a lot" is like 100mA max)
wary cloak
Our column is low so 0 volts to the 0 volts of the input that it's now low
Why the real output it's low? Isn't this the key pressed scenario?
coral geyser
Why the real output it's low? Isn't this the key pressed scenario?
you would measure 0v if you used an oscilloscope to read the voltage of the output pin
coral geyser
But the other cases where key not pressed and this one it is?
in that case the output pin just floats around
it will be something like 4.5v in reality
not actually 0v or 5v
wary cloak
So when the pin is high which is the case we don't want it to match it's not 0 so it wouldn't match the 0-0 as before so it's fine, and when it's low it would match the 0-0 and well the in our behavies the same way
So.
Now instead of asking for the column
With the pin on high
That was 0 volts, we do it with low our now 0 volts?