The sole purpose of this repository is to store notes and code helping me learn things related to Internet of Things, but also basic electronics.
I started out with reading and writing examples in assembly language to gain a basic understanding for how to program a micrcontroller and also implement the most common communication protocols, UART, SPI, I2C, and CAN.
After that I started to look at using Rust and chose to take one of the assembly examples, led-ext.s, and re-write it in Rust, first without any external crates, rust-low-level, and then using higher level crates, rust-pac, and rust-hal. This was followed by adding an example that uses Embassy, embassy-led, and then one that uses Drogue drogue-led.
The notes directory contains notes on various topics related to microcontrollers and IoT. Currently this document also contains notes but I'm in the process of extracting these notes into separate documents.
This process will most likely be an iterative process where I go back to the lower levels when learning new concepts and then working upwards from there as needed.
Is a very small computer system self contained on an integrated circuit (IC). They normally don't have any external data storage available so everything has to be able to be stored in memory (or perhaps sent off somewhere to be stored elsewhere).
Is what actually runs our code. It looks like a small black box and has a number of small metal pins around it. The pins are connected to tracers that are like wires connecting components on the board.
Microcontrollers convert analog values (continuous) that arrive on inputs (pins) into to digital values (descrete) to be processed.
ARM holding is the company behind Arm based chips do not manufacture chips but instread they design parts of the chip and they license these designs to hardware manufactures that use them in their products (sometimes enhancing them). In contrast to Intel both design chips and manufacture them.
Arm has a number of families of chip designs. For example there is the Cortex-M family which are mostly used in microcontrollers.
Cortex-M0 is designed for low cost/power usage. Cortex-M4 is a middle ground more features/performance than M0 but less expensive then M7.
Cortex-M7 is higher cost but has more features and better performance. Some additional information about Arm version can be found here.
Is the process of transmitting one bit of data at a time. Think of this as being one wire and the bits flow through it. Only requires one wire which is one reason at it is used often in microcontroller system design.
Would require only two wires:
+---------+ b0 b1 b2 b3 b4 b5 b6 b7 +---------+
| OUT|-----------------------------|IN |
| | p1 p2 p3 p4 p5 p6 p7 p8 | |
| CLK|-----------------------------|CLK |
+---------+ +---------+
p = puls of the clock
So this would send one bit per plus of the clock.
Is the process of transmitting multiple bits of data at one time. Think of this as muliple wires connected to the other side and multiple bits can be set and transmitted at once. Since it requires more wires it can be more expensive to implement. Crosstalk is an issue and it is susceptible to clock skew.
+---------+ +---------+
| OUT0|-------------b0--------------|IN0 |
| OUT1|-------------b1--------------|IN1 |
| OUT2|-------------b2--------------|IN2 |
| OUT3|-------------b3--------------|IN3 |
| OUT4|-------------b4--------------|IN4 |
| OUT5|-------------b5--------------|IN5 |
| OUT6|-------------b6--------------|IN6 |
| OUT7|-------------b7--------------|IN7 |
| | p1 | |
| CLK|-----------------------------|CLK |
+---------+ +---------+
p = pulse of the clock
So this would send 8 bit per pluse of the clock.
It's called synchronous because the sender and reciever are synchronized. They both have a clock which use the same rate.
A clock in this case is something that changes between two values, like high/low or something like that. The sender might send only when the clock is high for example. This is not much of a problem if both sender and reciever share the same clock, if they are part of the same circuit for example, but if they are not it might be difficult as they need someway of sharing the clock.
Fast transmission (when is has been sent). May have to wait until data can be sent.
Examples are SPI, and I²C.
This type of communication is not synced, but instead it uses a start bit and and stop bit
start bit stop bit
↓ [ data ] ↓
+--------+ 0 11101001 1 +--------+
| Sender | -----------------> |Reciever|
+--------+ +--------+
The extra processing of the start/stop bit can affect performance making asynchronous slower. Is cheap and easy to implement (no clock sharing). Can transmit whenever it is ready (does not have to wait for the clock).
Is a signal that repeats/oscillates between high and low states:
high 1 | | | | | |
| | | | | |
low 0---- ---- ---- ---- ...
So it starts out at zero, changes to 1 for the same amount of time and then repeats like this. The clock tells us when the receiver should read from the data line. We might say that it should receive when the clock is high but notice that the line can be high for a certain period of time. Instead we say that it should read when transitioning from low to high, and this point is called an edge:
high 1 | | | | | |
| | | | | |
low 0---- ---- ---- ---- ...
^ ^
raising edge +-- fallin edge
(from low (from high
to high) to low)
This is a single point instead of a region.
So lets say we want to send 5 (101). First we have to let the receiver know that we are going to send by setting a wire to low which is sometimes called slave select (SS_bar). After this is done we can send bits to the receiver:
Sender Receiver
-
|
|
CLK --* > CLK
101 SOUT --1-------------> RIN 1
So above when the first raising edge is encountered, remember that the sender and receiver share the same clock line so the receiver also knows when this happens. When this happens SOUT will become high for the binary digit 1 to be sent. This can then be read by the reciever from RIN above and placed into a buffer. The next time there is a raising edge encountered (clock tick) another bit will be placed in SOUT:
Sender Receiver
---- -
| | |
| | |
CLK ---- ---* > CLK
10 SOUT --------0-------> RIN 10
And so on:
Sender Receiver
---- ----
| | | |
| | | |
CLK ---- ---- ---* > CLK
10 SOUT --------------1-> RIN 101
Is a synchronous serial communication spec developed by Motorola (1980s) and used for communicating between microcontrollers and small peripheral devices. Note that this is a defacto standard but there is no official standard for this protocol.
+-----------------+ +-----------------+
| SPI SCLK|-------+-->|SCLK SPI |
| Controller COPI|------+--->|COPI Peripheral|
| CIPO|<----+-----|CIPO |
| CS|---------->|CS |
+-----------------+ ||| +-----------------+
^ |||
| ||| +-----------------+
| ||+-->|SCLK SPI |
| |+--->|COPI Peripheral|
| +-----|CIPO |
+-------------->|CS |
+-----------------+
SCLK = SPI Clock
CIPO = Controller Input Peripheral Output
COPI = Controller Output Peripheral Input
CS = Chip Select (to select among mulitiple connected peripherals like above)
- Full Duplex
- Synchronous (it uses the clock to sync)
- 4+ wires (with multiple slaves there will be more than 4 wires)
- 25 Mbps
- No ack
- Low power consumption
- 20 cm distances
If the CPOL bit is 0, then the SCK pin has a low-level idle state:
---
|
|
----
So the clock in the idle state will be low.
If the CPOL bit is 1, then the SCK pin has a high-level idle state:
----
|
|
----
And in this case the clock in the idle state will be high.
CPAH detrmines when data will go out, or when data will sampled during the clock cycle phase. If the communication should be writing data on the falling edge of the cycle or on the raising edge of the clock cycle.
CPOL = 0 (remember this means the clock is low when idle)
CPAH = 0
---- ----
| | | |
| | | |
---- ---- ----
↑
|
raising edge
CPOL = 0 (remember this means the clock is low when idle)
CPAH = 1
falling edge
|
↓
---- ----
| | | |
| | | |
---- ---- ----
CPOL = 1 (remember this means the clock is high when idle)
CPAH = 0
falling edge
|
↓
---- ---- -----
| | | |
| | | |
---- ----
CPOL = 1 (remember this means the clock is high when idle)
CPAH = 1
---- ---- -----
| | | |
| | | |
---- ----
↑
|
raising edge
Now, clock polarity and clock phase can be combined as we might be able to tell from above. We can have {CPOL=0, CPHA=0}, {CPOL=1, CPAH=0}, {CPOL=0, CPAH=1}, and {CPOL=1, CPAH=1}.
One thing to note is that we always want to sample data in the middle of a clock cycle and never close to the edges as that means that data i changing and sampling then migth cause corruption to data.
This is a serial interface where 4 data lines are used.
UART is not a communications protocol like SPI and I2C but instead a physical component/circuit in a microcontroller or a standalone integrated circuit. The standard it implements is RS-232 protocol. It is main purpose is to send and receive serial data.
An UART can look something like the following:
Data bus +------------+ +------------+ Data bus
---->|D0 P|S | | P|S D0|---->
---->|D1 A|E | | A|E D1|---->
---->|D2 R|R | | R|R D2|---->
---->|D3 A|I RX|<----------|TX A|A D3|---->
---->|D4 L|A TX|---------->|RX L|L D4|---->
---->|D5 L|L | | L| D5|---->
---->|D6 E| | | E| D6|---->
---->|D7 L| | | L| D7|---->
| | | | | |
-|R/W | | | | R/W|
-|CLK | | | | CLK|
-|INT | | | | INT|
+------------+ +------------+
An packet on the serial wire (TX->RX) will looks something like this:
+-+ +-+-+-+-+-+ +--+ +--+
|S| | | | | | | |P | |ST|
+-+ +-+-+-+-+-+ +--+ +--+
Start Data Parity Stop
The transimission line (TX) is usually held at a high voltage when not trasmitting. To start sending the trasmitting UART will pull the TX line from high to low for one clock cycle. When the receiving (RX) UART see this it will begin reading the bits in the dataframe at the frequency of the baud rate. Notice that the sender will get the data to be sent from the data bus, and likewise the reciever will place the received data on the data bus.
So there are only two wires which is a nice property. And there is no clock signal required between two UARTs. There is also parity checking which is good for reliable communication. One thing to note is that the data being sent can be a maxium of 9 bits. The is not way to support multiple peripherals.
Was first developed in 1982 by Phillips Semiconductors (which is now NXP semiconductors) and a public specification came out in 1992 which allowed speeds of 400kHz and an address of 10-bits.
There are three additional modes:
- fast-mode plus, 1MHz
- high-speed mode, 3.4MHz
- ultra-fast mode, 5MHz
Intel introduced a variant of I²C in 1995 called System Management Bus (SMBus) intended to improve communiciation predictability between ICs on PC motherboards. SMBus limits the speed of communication to 10kHz to 100kHz.
Only requires two pins/wires:
+------------------+ +-----------------+
| Controller 1 SCL|<--+------>|SCL Peripheral 1|
| SDA|-------+-->|SDA |
+------------------+ | | +-----------------+
+------------------+ | | +-----------------+
| Controller 2 SCL|<--+------>|SCL Peripheral 2|
| SDA|-------+-->|SDA |
+------------------+ +-----------------+
SCL = Serial Clock Line for the clock signal
SDA = Serial Data line for sending and recieving data
So we have the usage of a clock so this is a synchronous protocol like SPI. Each I²C consists of two signals/lines, the clock signal and the data signal. The clock signal is always generated by the current bus controller And notice that we have a single wire to both transmit and recieve.
The format of data send on the bus is done in 8-bit frames. Each frame also includes an ACK bit. The acknowledgment takes place after every bytes.
Data is transported in messages which have a specific format, and an address is used to identify the destination peripheral:
+-+-+-+-+-+-+-+ +---+ +---+ +-+-+-+-+-+-+-+-+
| | | | | | | | |R/W| |ACK| | | | | | | | | |
+-+-+-+-+-+-+-+ +---+ +---+ +-+-+-+-+-+-+-+-+
Address Frame Data Frame
(7 or 10 bits) (8 bits)
R/W = 1 Controller is requesting to read
R/W = 0 Controller is requesting to write
ACK = Is used by the peripheral that matches the address, which is pulled low
so there should be a 0 in the ACK bit if the address maches. If this is not
pulled low this is interpreted as a NACK by the controller.
Try to remember that this communication is one clock cycle at a time, so a peripheral can check each bit of the addess and if it matches it can pull its signal low. The other peripherals do nothing.
I²C uses open collector circuit for it's bus:
+Vcc
----
|
/
\ R
/
Signal->|-------------+-----------+----------+
| | | |
S / / / /
| | | |
--- GND --- --- ---
- - - -
. . . .
C1 C2 P1 P2
It is always the controller that initiates the communication. Also, controllers cannot communicate with each other, they can only communicate with peripherals.
The controller begins by pulling SDA low (remember that when all switches are open the signal is a logic one) while SCL is high:
Address Frame
SDA -+ +--+ +--+ +--+--+--+ +-
| |A6|A5|A4|A3|A2|A1|A0|RW|ACK|
+--+ +--+--+ +--+ +---+
1 0 0 1 0 1 1 1 0
And this is followed by pulling SCL low, this is referred to the START condition in the spec I think and I think this is what is meant when other documents say the start bit.
The clock is what enables the peripheral when to reads these bits
Address Frame
1 0 0 1 0 1 1 1 0
SDA -+ +--+ +--+ +--+--+--+ +-------+ --------
| |A6|A5|A4|A3|A2|A1|A0|RW|ACK| | |D7|D6|D5|D4|D3|D2|D1|D0|ACK|
+--+ +--+--+ +--+ +---+ +--+
SCL -+ + + + + + + + + + + + + + + + + + + +-------
| | | | | | | | | | | | | | | | | | | |
+----+--+--+--+--+--+--+--+--+-------------+--+--+--+--+--+--+--+--+--+
↑
(between frame the clock is low)
The clock pulses are supposed to show that they are in the middle of each bit and that the peripheral should read at the point to get the most accurate reading/sampling.
Now, each peripheral will detect this pulling low of SDA and read the address frame.
If none of the peripherals match the address, that is SDA has not been pulled low, then they all leave the SDA unchanged and the master will consider this an NACK and the controller will send the stop bit.
After data has be acknowledged then a STOP condition is sent which is the opposite of the start conditions where SCL is pulled high, while SCL is high. Instead of the STOP condition an new START condition can be sent instead which is called a repeated start.
Is a way for the peripheral to signal that it needs a little more time to gather data it is going to send.
This is performed by the peripheral by pulling the clock low.
1 0 0 1 0 1 1
SDA -+ +--+ +--+ +--+--+ +
| |D7|D6|D5|D4|D3|D2|D0| |ACK|
+--+ +--+--+ +--+ +----+---+
SCL -+ + + + + + + + +
| | | | | | | | |
+----+--+--+--+--+--+--+--+--+
i2c-c.s is the controller which currently
send an single A to the peripheral i2c-p.s.
Both of these programs are flashed onto STM32F072B-Discovery boards and
connected using PB6 (SCL) and PB7(SDA).
$ minicom --baudrate 115200 --device /dev/ttyUSB0
Welcome to minicom 2.7.1
OPTIONS: I18n
Compiled on Jan 26 2021, 00:00:00.
Port /dev/ttyUSB0, 18:49:45
Press CTRL-A Z for help on special keys
ACurrently the controller only sends on byte so to see more As being sent
we have to press the reset button. TODO: change this so that it just
continues sending bytes.
We mentioned earlier that we can have more than one controller. So what happens if two controllers start sending data at the same time. I turns out that the controllers not only send out data on the SDA but also read from it. So if a controller sends out a bit pattern it will check that that bit pattern is also on the wire. If another controller wrote the exact same bits then nothing happens but if the bits differ then a collision is detected and the controller who did not read the same bit pattern that is sent will back off.
In SPI we had a peripheral select wire (SS) which selected the peripheral we want to talk to. Instead in I2C each peripheral has an address. This is contained in the address frame which is always first frame after the start bit has been set. The controller sends this message frame onto the SDA which all peripherals connected will see. Each peripheral will compare this to their own address and do nothing if the address does not match. If the address matches a peripheral it sends a low voltage ACK bit back to the controller.
The READ/WRITE bit indicates if the controller wants to send or recieve data from the peripheral. If it wants to send then this bit is low (0) and if it wants to read it will be a high voltage.
We've already talked about USART and SPI previously so why do we need another option for communication. USART is asynchonous, there is not clock but instead start/stop bits (which reduces the data rate somewhat), and it also a little more complex with regards to the hardware implementation. USART is usually used for communicating between to components and note multiple connected to a bus like SPI. SPI requires four pins/wires to connect two components, and adding additional ones requires another wire for the chip select pin. The hardware is usually a simple shift register at either end.
Compared to SPI there is one additional bit of data being sent which can affect the data rate.
Is the flow of free electrons.
Take a conductor like a coppar wire which is good at allowing a current to flow through it. Now, coppar is made up of coppar atoms which like all atoms contain protons, neutrons, and electrons:
Cu
Protons: 29
Neutrons: 35
Electrons: 29
Shell 1: 2 electrons
Shell 2: 8 electrons
Shell 3: 18 electrons
Shell 4: 1 electrons (free electron)
A coppar wire without any current flowing will have electrons randomly moving from atom to atom, but this is not in any perticular direction. If an electron moves away from its atom that will leave the atom slightly positively charged and that would attract an electron.
So we have the sea of electrons that are moving around randomly. Now, if we connect a cell that contains one end with negative electrons and the otherside with atoms what are positivly charged (would attract electrons) the electrons will move towards the positiv end.
The conduction band is the band above the valence band.
-----------------------
Conduction band
-----------------------
Band gap
-----------------------
Valance band
-----------------------
...
----------------------- Electron shell n-1
For electrons to be able to move to an outer shell energy needs to be applied. For conductors the conduction band and the valance band/band gap/conduction band overlap allowing the free electrons to move in this space. But for insulators there is a clear band gap and energy is required for this to happen (that is if there are free electrons in the valance shell), if we are thinking of atoms where no energy is applied to it.
Bond: When an atom bonds with another they can share electrons in their valence shell, this type of bond is called covalent bond.
Take two silicon atoms for example, they would share a pair of electrons:
* * * = electron in valance shell
** = paired electrons
* Si ** Si *
* *
Many silicon atoms that are bonded like this will allow for each Si atom to be connected though sharing four electrons:
This is something that can be seen in silicon where each silicon atom is connected to four other silicon atoms.
If a conductor (metal) is heated it becomes less conductive. So metals conduct electricity better at colder temperatures.
Energy
^
| +--------------+
| | Conduction |
| +--------------+ | |
| | Conduction | +--------------+
| +--------------+ | |
| Conduction | +--------------+
|--------------|
|--------------| +--------------+ +--------------+
| Valancy | | Valancy | | Valancy |
+--------------+ +--------------+ +--------------+
Conductors Semiconductors Insulators
Notice that for conductors the conduction band and the valancy band overlap which is why the conduct so well. For semiconductors there is a gap between these two bands so energy is needed for electrons to move into the conduction band. This is possible with heat.
High reisistance to the flow of electricity, which means that their valance shell is full and also there is a large gap between the valancy and conduction bands.
These have properties that are in between conductors and insulators, hence the name semi-conductors. If a semi-conductor is exposed to heat this can improve its conductivity. But there is another way to improve the conductivity of a semi-conductor which is to add impurities to it (called doping).
If we take a look at a silicon cristal:
[* *] = covalent bond
Si [* *] Si [* *] Si [* *]
⌈*⌉ ⌈*⌉ ⌈*⌉
⌊*⌋ ⌊*⌋ ⌊*⌋
Si [* *] Si [* *] Si [* *]
⌈*⌉ ⌈*⌉ ⌈*⌉
⌊*⌋ ⌊*⌋ ⌊*⌋
Si [* *] Si [* *] Si [* *]
So we can see that we have these covelent bounds between the atoms and all of the free electrons in the valance shell are bound. There are no free electrons or "holes" (which is an atom which is slightly positively charge and it lacks an electron).
If we replace one of these atoms with an atom with only 5 electrons (like fosfor) in its valance shell what would happen:
[* *] = covalent bond
Si [* *] Si [* *] Si [* *]
⌈*⌉ ⌈*⌉ ⌈*⌉
⌊*⌋ ⌊*⌋ ⌊*⌋
*<-------------------- free electron
Si [* *] P [* *] Si [* *]
⌈*⌉ ⌈*⌉ ⌈*⌉
⌊*⌋ ⌊*⌋ ⌊*⌋
Si [* *] Si [* *] Si [* *]
It can still bond with the other 3 atoms around it but also has an additional electron left over. This free electron can now move in the conduction band. This makes this silicon cristal more conductive then before. Because we have a free electron by this process this is called an n-type (for negative) conductor now. But also notice that there is still a balance of protons and electrons sinse the introduced atom has 5 protons.
So if we hook up a battery this the free electron will be attracted to the positive terminal and the electrons from the negative terminal will be able to move in their place through the material (in/through the conduction band):
Si [* *] Si [* *] Si [* *]
⌈*⌉ ⌈*⌉ ⌈*⌉
⌊*⌋ ⌊*⌋ ⌊*⌋
* <------------------ free electron moved
+--- Si [* *] P [* *] Si [* *] <-----------+
| |
| ⌈*⌉ ⌈*⌉ ⌈*⌉ |
| ⌊*⌋ ⌊*⌋ ⌊*⌋ |
| |
| Si [* *] Si [* *] Si [* *] |
| |
| | |
| || |
+-------------> + || - -------------------------+
|
So lets try with alunminum (Ai):
[* *] = covalent bond
x = hole/missing electron
Si [* *] Si [* *] Si [* *]
⌈*⌉ ⌈*⌉ ⌈*⌉
⌊*⌋ ⌊*⌋ ⌊*⌋
Si [* *] Ai x * Si [* *]
⌈*⌉ ⌈*⌉ ⌈*⌉
⌊*⌋ ⌊*⌋ ⌊*⌋
Si [* *] Si [* *] Si [* *]
Notice that we have a missing electron so there is no covalent bond between the Ai atom in the middle and the Si atom to the right of it. This is called a p-type semi-conductor. This also conducts electricity better than pure silicon so the conductivity goes up.
In the case of N-type we know that the electrons are free to flow in the conduction band. This is not what happens for P-type, instead the electrons flow in the valance band, so if we apply a current electrons will be attracted to the positive terminal, hence they will move to holes:
Si [* *] Si [* *] Si [* *]
⌈*⌉ ⌈*⌉ ⌈*⌉
⌊*⌋ ⌊*⌋ ⌊*⌋
*
+--- Ai x * Si [* *] Si [* *] <-----------+
| |
| ⌈*⌉ ⌈*⌉ ⌈*⌉ |
| ⌊*⌋ ⌊*⌋ ⌊*⌋ |
| |
| Si [* *] Si [* *] Si [* *] |
| |
| | |
| || |
+-------------- + || - <-------------------------+
|
The holes are what allow electrons to flow and this happens in the valance band as opposed to n-type doping where the flow happens in the conduction band.
With a batteri cell attached the electrons will be attracted to the positive terminal. Therefor electrons will fill the holes:
x o x x x
<-
x x o x x
<-
x x x o x
<-
x x x x o
Now, what I think will happen is that one electron will "leave" and and one will enter:
x <- o x x x o <-x
o x x x x
And once again the process of the electrons filling the holes will continue and a electrons will be transferred/conducted accross the valance band.
Now, lets see what happens if we combine a p-type and an n-type:
[* *] = covalent bond
x = hole
* = free electron
P-type N-type
Si [* *] Si [* *] Si | | Si [* *] Si [* *] Si [* *]
| |
⌈*⌉ ⌈*⌉ ⌈*⌉ | | ⌈*⌉ ⌈*⌉ ⌈*⌉
⌊*⌋ ⌊*⌋ ⌊*⌋ | | ⌊*⌋ ⌊*⌋ ⌊*⌋
| | *
| |
Si [* *] Ai x * Si | | Si [* *] P [* *] Si [* *]
| |
⌈*⌉ ⌈*⌉ ⌈*⌉ | | ⌈*⌉ ⌈*⌉ ⌈*⌉
⌊*⌋ ⌊*⌋ ⌊*⌋ | | ⌊*⌋ ⌊*⌋ ⌊*⌋
| |
Si [* *] Si [* *] Si | | Si [* *] Si [* *] Si [* *]
The free electron would be attracted to fill the hole
Si [* *] Si [* *] Si | | Si [* *] Si [* *] Si [* *]
| |
⌈*⌉ ⌈*⌉ ⌈*⌉ | | ⌈*⌉ ⌈*⌉ ⌈*⌉
⌊*⌋ ⌊*⌋ ⌊*⌋ | | ⌊*⌋ ⌊*⌋ ⌊*⌋
+-------------------------------- *
↓ | |
Si [* *] Ai x * Si | | Si [* *] P [* *] Si [* *]
| |
⌈*⌉ ⌈*⌉ ⌈*⌉ | | ⌈*⌉ ⌈*⌉ ⌈*⌉
⌊*⌋ ⌊*⌋ ⌊*⌋ | | ⌊*⌋ ⌊*⌋ ⌊*⌋
| |
Si [* *] Si [* *] Si | | Si [* *] Si [* *] Si [* *]
If an electron moves from the n-type to the p-type this will cause an imbalance where the n-type size now has a mismatch of the P atom, there is not one proton more that there are electrons (remember that this was not the case when the free electron was in the n-type side). So the n-type side will now become slightly positively charged.
A hole can also move to the n-type side in which case the p-type area will not be imbalanced, the Ai atom will not have one more electron compared to protons and become slightly negatively charged. And remember that a charge will create an electric field. This electric field will prevent electrons and holes to move between these to areas. This process will create an area between these two reqions where there are now extra electrons, or any extra holes, so there are now charge carriers in this region. This region is called the depletion region and it like an insulator.
Depletion Region
N ↓ P
+----------------+---------------+
| * * ** *** * * | o o oo ooo |
| ** * ** * * * | o o o o o oo |
| * * ** *** * |o o o o |
| * * ** *** * | o o o o o|
+----------------+---------------+
Depletion Region
N ↓ P
+--------------+----+-------------+
| * * ** *** * | |o o oo ooo |
| ** * ** * * | | o o o o oo |
| * * ** *** * | |o o o |
| * * ** *** * | |o o o o o|
+--------------+----+-------------+
Now if we hook up a power source we will not get a current flowing initially as the depletion region prevents the flow. But if the batteri has enough voltage, like > 0.7v this will overcome the barrier.
Depletion Region
N ↓ P
+--------------+----+-------------+
| * * ** *** * | |o o oo ooo |
+--> | ** * ** * * | | o o o o oo |---------+
| | * * ** *** * | |o o o | |
| | * * ** *** * | |o o o o o| |
| +--------------+----+-------------+ |
| |
| | |
| - | | + |
+-----------------| |<---------------------------+
| |
|
Electrons will enter the n-region and there will be more electrons in that region. There will also be electrons (remember there are holes but there are also electrons in this region. Those electrons will be attracted to the positive terminal and will exit leaving more holes behind. If the batteri has a voltage
0.7 the depletion region will collapse and current can flow.
Now lets hook this up the other way, where the negative terminal is connected to the p-type region, and the positive terminal to the n-type region:
Depletion Region
N ↓ P
+--------------+----+-------------+
| * * ** *** * | |o o oo ooo |
+----| ** * ** * * | | o o o o oo |<--------+
| | * * ** *** * | |o o o | |
| | * * ** *** * | |o o o o o| |
| +--------------+----+-------------+ |
| |
| | |
| + | | - |
+---------------->| |----------------------------+
| |
|
In this case the electrons will fill the holes in the p-region, and electrons will leave the n-region, causing the depletion region to become larger:
Depletion Region
N ↓ P
+---------------------------------+
| * * ** **| |o oo ooo |
+----| ** * ** | | o o o oo |<--------+
| | * * ** * | |o o | |
| | * * ** * | | o o o o| |
| +--------------+----+-------------+ |
| |
| | |
| + | | - |
+---------------->| |----------------------------+
| |
|
As the depletion region becomes larger, meaning that there are no charge carriers in this region, this component will act like an insulator. So no current will flow.
If we stick an n-type region next to a p-type, and then another n-type we get the following:
Depletion Regions
| |
N ↓ P ↓ N
+-------------+-+--------+-+----------+
| * * ** *** *| |o o oo o| |* * * * * |
| ** * ** * | | o o o o| | * * ** |
| * * ** *** | |o o o | | * * * |
| * * ** *** *| |o o o| | * * * * |
+-------------+-+--------+-+----------+
And say we want a current to flow through this thing:
Depletion Regions
| |
N ↓ P ↓ N
+-------------+-+--------+-+----------+
| * * ** *** *| |o o oo o| |* * * * * |
+--->| ** * ** * | | o o o o| | * * ** |-----+
| | * * ** ** | |o o o | | * * * | |
| | * * ** ** * | |o o o| | * * * * | |
| +-------------+-+--------+-+----------+ |
| |
| |
| |
| |
| | |
| - | | + |
+-----------------| |<---------------------------+
| |
|
Well this will increase number of electrons in the left-most n-type region, and some electrons will leave the right-most n-type region but there will not be a current flowing.
Lets try hooking up a second power source like this:
Depletion Regions
| |
N ↓ P ↓ N
+-------------+-+--------+-+----------+
| * * ** *** *| |o o oo -| |+ * * * * |
+--->| ** * ** * | | o o o -| |+ * * ** |-----+
| | * * ** ** | |o o o -| |+* * * | |
| +--| * * ** ** *| |o o -| |+ * * * * | |
| | +-------------+-+--------+-+----------+ |
| | | | |
| | - | | + | |
| +---------------| |-----+ |
| | | >0.7v |
| | |
| |
| | |
| - | | + |
+-----------------| |<---------------------------+
| |
|
Notice that this extra connection between the left-most n-type and the p-type is the same as we hade above where we saw that current would flow through that connetion.
Now, we have electrons entering from both battries. When they make these types of components the first n-type region is very heavyly doped. So once this second current starts flowing, those extra electrons can now also move through into the p-type region. Some of these electrons are going to fill in holes in the p-type region, and some are also going to be attracted to the outgoing plus terminal of the p-type region. The base is very thin (how they are manufactured) and these holes are physically close to the left-most n-type's electrons. That depletion region was created when electrons left the n-type into the depletion region which caused the n-type region to become positivley charged (remember that the number of electrons and protons match up when neutral and removing a electron will make the atom postitive). And those electrons that move into depletion region cause the p-type region to become negatively charged. The electrons that have now started flowing through the p-type region will be attracted to the positive right-most n-type region and therefor be able to flow through it towards the positive terminal output.
This component is called an NPN Bipolar Junction Transistor:
Emitter(N) Base(P) Collector(N)
+-------------+-+--------+-+----------+
| * * ** *** *| |o o oo -| |+ * * * * |
+--->| ** * ** * | | o o o -| |+ * * ** |-----+
| | * * ** ** | |o o o -| |+* * * | |
| +--| * * ** ** *| |o o -| |+ * * * * | |
| | +-------------+-+--------+-+----------+ |
| | | | |
| | - | | + | |
| +---------------| |-----+ |
| | | >0.7v |
| | |
| |
| | |
| - | | + |
+-----------------| |<---------------------------+
| |
|
This would be drawn as:
Collector
/
+----/---+
| |/ |
Base ----| |
| |\ |
| _\| | (this is supposed to be an arrow pointing to the Emitter)
| \ |
+------|-+
|
|
Emitter
But notice that the electrons are actually flowing from the emitter to the base which is always confusing me when looking at schematics. We have the base which will have circuit with the emitter, and when there is a voltage applied to this, current will flow through the emitter to the base. And there will then also be a current flowing through the whole thing/transistor from the emitter through the first junction, through the base, through the second junction, and through the collector. The box represents the transistor and is usually circle but now I think it makes sense that it is there with the above explaination.
Alright, so that was the theory now lets see how we can create a circuit with such an NPN transistor:
For this example I used as button switch which is connected to the base of the transistor:
So without a current flowing through the base and the emitter there is no current flowing though the transistor. Pressing the button will cause a current to flow:
So that is how a NPN Bipolar Junction transistor works and how we can hook it up physically. We can use two such transistors:
Now, pressing just the left button no current will flow:
And, pressing just the right button no current will flow:
But pressing both and current will flow:
Notice that what we have created here is a AND gate:
Left button | Right button LED
0 | 0 0 (Off)
1 | 0 0 (Off)
0 | 1 0 (Off)
1 | 1 1 (On)
Before any movement has occured we have the following Formal Charge for P:
⌈*⌉
⌊*⌋ *
[* *] P [* *]
⌈*⌉
⌊*⌋
Bounds: 4
FC = valance electrons - (Bonds + dots)
FC = 5 - (4+1)
FC = 0
This is important and we can note that the charge/voltage is neutral because the phosphorus atom has 15 protons and 15 electrons, and 5 of those electrons are in its valance shell. The extra free electron does not cause an inbalance here, there are still 15 protons and 15 electrons and the other (not shown) Si atmos are also neutral.
And for Ai:
⌈*⌉
⌊*⌋
[* *]Ai x *
⌈*⌉
⌊*⌋
FC = 3 - (3 + O)
FC = 0
Ai has 13 protons and 3 electrons, and three of those electons are in its valance shell. So even though there is a hole it does not cause an imbalance at this point.
But if/when a free electron from from one side moves over to the other side things change. When the free electron from moves from the P atom to the Ai atom the P atom is now imbalanced, it now has one more proton than it has electrons and is now positively charged. A similar thing happens on the Ai side where the hole is filled with the electron. The Ai atom now has one more electron than it has protons and is therefor negatively charged.
Now, after the move of the free electron to the hole we have the following formal charge for P:
FC = 5 - (4 + 0)
= 1
This means that it has a positive formal charge.
And for Ai:
⌈*⌉
⌊*⌋
[* *]Ai [* *]
⌈*⌉
⌊*⌋
FC = 3 - (4 + O)
FC = 3 - 4
FC = -1
This means that it has a negative formal charge.
P N
+--------------------+-------------------------+
| o o o | * * * * |
| o o o | * * * |
| o o o | * * |
| o o o | * * * * * |
+--------------------+-------------------------+
The free electrons will drift to fill in the holes:
P N
+--------------------+-------------------------+
| o o o <--* * * * |
| o o o <-- * * * |
| o o o<-- * * |
| o o o <--* * * * * |
+--------------------+-------------------------+
And this will cause the sides to positively and negatively charged:
P N
+--------------------+-------------------------+
| o o o -|+* * * * |
| o o o -|+ * * * |
| o o o-|+ * * |
| o o o -|+* * * * * |
+--------------------+-------------------------+
When we have a charge(voltage difference) we also have the generation of an electric field. So for electrons to move through this electric field more energy is required (up until now we have only been thinking in terms of thermal energy like the head of room temperature).
P N
+--------------------+-------------------------+
| o o o | |* * * * |
| o o o | | * * * |
| o o o| | * * |
| o o o | |* * * * * |
+--------------------+-------------------------+
^
Electric field accross the PN junction
The electric field makes it difficult for new electrons on the N side to cross the PN junction now. The free electrons still have a force acting upon them that wants to move them to the other side, but there is also a force acting in the opposite direction.
| ↑
P ↓ | N
o o -->-<--|<--+--> * *
o o -->-<--|<--+--> * *
o o -->-<--|<--+--> *
↑ |
| ↓
Notice that the electric field is a vector, which has a magnitude and a direction and the direction is different, the opposite for the negative side.
When these forces are equal then no electrons will move over from the right side to the left side.
So applying a current to the N side that overcomes the force of the PN junction would allow electrons to move through to the P side and beyond if the P-side is connected to something. Reversing this process, that is connecting a voltage to the anode (P-side) will not cause any (or vary little current to flow). I'm taking about electron current here and not convertional current. So the flow of electrons can only happen in one direction.
P N
+--------------------+-------------------------+
| o o o | |* * * * |
| o o o | | * * * |
| o o o| | * * |
| o o o | |* * * * * |
+--------------------+-------------------------+
Anode Cathode
----->|---
<-- e⁻
+----+----+
anode-----| P | N |-----cathode
+----+----+
This is how a Diode is made.
So current will flow when electrons can flow from the n-type region to the p-type region provided that the voltage applied can over come the depletion regions electric field:
+----+----+
+---<-----| P | N |-----<----+
| +----+----+ |
| +-----------+ |
+--->----|+ -|---->----+
+-----------+
But no current will flow if we connect it the diode in the other way:
+----+----+
+---------| N | P |----------+
| +----+----+ |
| +-----------+ |
+--------|+ -|---------+
+-----------+
Notice that this is basically a Diode that can emit light.
+--+
| |
+--+
| |
| |
| |
|
Anode Cathode
When we connect these the electron current must flow through the cathode so it should be connected to the negative output of the batteri.
There are two (bi) junctions (think of the PN junctions in a diode as explained above):
+-----+-----+-----+
| N | P | N |
+-----+-----+------
^ ^
junction junction
1 2
The N is for N-Type, and P for P-Type which are the same as in the diode example above.
+-----+-----+-----+
E-----| N | P | N |-----C
+-----+-----+------
|
|
B
E = emitter
C = collector
B = base
So we would have free electrons in the left and right boxes, and holes in the middle box. And notice if we removed the first box we would be left with a normal diode:
+-----+-----+
| P | N |-----C (cathode)
+-----+------
|
|
B (anode)
And the addition of the first N box is basically the reversal of a diode
Source
|
| |-+
Gate --| |
| |-+
|
Drain
closed-circuit means that electricity is flowing from the Gate to the Source.
open-circuit means that electricity is not flowing from the Gate to the
Source but instead from the Gate to the Drain.
NMOS:
Source
|
| |-+
Gate --| |
| |-+
|
Drain
In NMOS a voltage is applied to the gate this will cause a closed circuit and current will flow between the source and the drain. When a zero voltage is applied to the gate this will cause an open circuit and no current will flow.
PMOS:
Source
|
| |-+
Gate -o| |
| |-+
|
Drain
In PMOS when a volatage is applied to the gate this will cause an open circuit
and no current will flow. And when zero voltage is applied it will cause a
closed circuit and current will flow.
The circle, o, on the PMOS gate inverts the value from the voltage. If the
gate sends a voltage that represents the value 1 the inverter will change this
to become 0.
We can combine NMOS and PMOS functions into something called complementary metal-oxide semiconductors (CMOS):
------ (Source of PMOS)
| PMOS
| +--+
|-o| |
| | +--+ (Drain of PMOS)
In ----| |--- Out
| | +--+ (Source of NMOS)
|--| |
| +--+ NMOS
|
↓ (Drain of NMOS)
So lets say that we pass a logical zero on the In wire. This will go to both the PMOS Gate, and to the NMOS gate
First lets take a look what happens with the PMOS transistor:
------ (Source of PMOS
| PMOS
1| +--+
0|-o|\
| | +--+ 1 (Drain of PMOS)
0 -----| |--- Out
| | +--+ (Source of NMOS)
0|--| |
| +--+ NMOS
|
↓
A voltage on the PMOS Gate will cause an open circuit which means that there will no be any electricity flowing from the Gate to the Source, but from the Gate to the Drain. So this logical 1 will be the Source of the NMOS transistor.
And now lets see what happens with the NMOS transistor:
------ (Source of PMOS
| PMOS
1| +--+
0|-o|\
| | +--+ 1 (Drain of PMOS)
0 -----| |--- Out
| | +--+ (Source of NMOS)
0|--| |
| +--+ NMOS
|
↓ (Drain of NMOS)
The logical 0 will cause the NMOS transistor to become an open circuit to ground (drain). So the output of this will be a 1.
Now, lets lets take a look what happens with the In is 1 (both PMOS and NMOS):
------ (Source of PMOS
| PMOS
0| +--+
1|-o| |
| | +--+ 0 (Drain of PMOS)
1 -----| |--- Out
| | +--+ (Source of NMOS)
1|--| |
| +--+ NMOS
|
↓
A zero voltage on the PMOS Gate will cause current to flow from the Source to the Drain of the PMOS. And for NMOS we are now applying a voltage to its Gate will cause a closed circuit and current will flow from the Source to the Drain, and in this case the Output will be 0.
Drain +-----------------------+
| |
+------------------+ |
| N | |
| | --- + Vds
|--+ C +--| - -
|P | H |P | |
| | A | | |
Gate -+--+-| | N | |---+ |
| | | | N | | | |
| | | | E | | | |
Vgs --- | |--+ L +--| | |
- | | | | |
| | | | | |
| | +------------------+ | |
| +-----------|------------+ |
| Source | |
+--------------+-----------------------+
N = N-Type material
P = P-Type material
Vds = Voltage Drain to Source
Vgs = Voltagae Gate to Source
Source is the source of electrons and drain is the output of electrons. By increasing the voltage of Vgs we can cause a depletion zone between the two P-type materials which will cause the channel to become smaller, there will be more resistance between the source and the drain. One can think of this as having a garden hose (the channel) where water is flowing through the source to the drain and the gate is like a wheel that can be turned to pinch the hose which reduces the flow or water. How much pinching is done is determined by the voltage beween the gate and the source voltage.
FET
+-----------------+------------------------+
| |
Junction FET (JFET) Metal Oxide FET (MOSFET)
| | |
Depletion Mode Depletion Mode Enhancement Mode
| | |
+-----------------+ +---------- +-----------+
| | | | |
N-Channel P-Channel N-Channel N-Channel P-Channel
D D D
D D | | |
| | |-+ |-+ |-+
|-+ |-+ || || ||
G->| G-<| G--||<+ G--||<+ G--||>+
|-+ |-+ || | || | || |
| | |-+ |-+ |-+
S S | | |
S S S
Depletion mode requires the Gate-Source (Vgs) voltage to switch OFF the current. Enhancement mode requires the Gate-Source (Vgs) voltage to switch ON the current.
+----------------------+
| |
| P-Type Substrate |
| |
| |
| |
+----------------------+
+----------------------+
|+---+ +---+ |
|| N | | N | |
|| | | | | S = Source
|| S | | D | | D = Drain
|| | | | |
|+---+ +---+ |
+----------------------+
So at this stage we have a PN junction between the substrate and the two N-type regions. Next an oxide insulator is added between the two N-types. And on top of that a metal layer is added. And this is the origin of the first part of the name, Metal for the metal plate, oxide for the insulator, and semiconductor for the PN.
----+ | +----
| | |
| ---------------- |
| ---------------- |
+----------------------------------+
|| N | | N ||
|+----------+ +--------+|
| ^ ^ |
| +---- pn junction -----+ |
| |
+----------------------------------+
FC = Valance electrons - (Bonds + dots)
Contains functions pointers to handlers of exceptions (and perhaps the ResetHandler in entry 0 but that is not clear to me yet).
This is a condition that changes the normal flow of control in a program.
Exceptions have a number associated with them and this is used as an index into the Vector Table which contains function pointer to Exception Handlers or Interrupt Service Routine (IRS). The ARM hardware will look up and call the function when an exception is triggered.
1 Reset
2 NMI
3 HardFault
4 MemoryManagement
5 BusFault
6 UsageFault
7 Reserved
8 Reserved
9 Reserved
10 Reserved
11 SVCall
12 DebugMonitor
13 Reserved
14 PendSV
15 SysTick
16 External interrupt 0
...
Each Exception also has a priority number.
All Cortex-M variants support 6 exceptions:
-
Reset This is the function called when the chip comes out of reset, like power on, or the reset button is pressed (can this be called programatically also?).
-
Non Maskable Interrupt (NMI) If an error happens in another exception handler this function will be called to handle it. It cannot be masked to be be ignore.
-
HardFault This is used for various system failures. There are also more fine grained exceptions handlers for MemoryManagement, BusFault, UsageFault.
-
SVCall This is the exception handler that will take care of supervisor call (svc) instruction is called.
-
PendSV/SysTick System level interrupts triggered by software and seem to be used mostly for RTOS.
If we take a look at the symbols we should be able to see the above handlers:
$ cargo objdump --release -- -t
Finished release [optimized] target(s) in 0.05s
app: file format elf32-littlearm
SYMBOL TABLE:
...
0000055a g F .text 00000000 DefaultHandler
00000040 g O .vector_table 000003c0 __INTERRUPTS
0000055a g F .text 00000000 BusFault
0000055a g F .text 00000000 DebugMonitor
0000055a g F .text 00000002 DefaultHandler_
0000055c g F .text 00000002 DefaultPreInit
0000068e g F .text 00000002 HardFault_
0000055a g F .text 00000000 MemoryManagement
0000055a g F .text 00000000 NonMaskableInt
0000055a g F .text 00000000 PendSV
00000400 g F .text 0000007c Reset
0000055a g F .text 00000000 SVCall
0000055a g F .text 00000000 SysTick
0000055a g F .text 00000000 UsageFault
0000047c g F .text 0000000a main
0000068e g F .text 00000000 HardFaultRemember that when an interupt occurs the thread state (registers etc) will be saved and execution will be paused. The interrupt handler is then executed and when it yields the state previously saved thread state is restored and execution continues.
So we can have one thread of execution, which can get interrupted, and then restored again. In embedded devices we can use this to have an interrupt wake up the processor from low power mode. There are processor instructions like ARMs Wait For Event (WFE) and Wait For Interrupt (WFI) which could be used to put a processor in low power mode. And when a interrupt happens the Set EVent (SEV) instruction can be issued to signal all other processors and wakes them up from low-power stand-by mode. A usage of this can be in an async executor were if there is not progress to be made by a future, the executor set up a waker function, which in this case would call SEV, and then use WFE to place itself into low-power mode to save on resources.
An atom is composed of a necleus which consists of a core of tightly bound subatomic particles called protons (positive charge) and neutrons (neither positive of negative). Rotating around the necleus are electrons. These have orbits that are referred to rings or shells and an electron has a negative charge. The number of electrons in orbit equal the number of protons in the necleus and the atom is electrically balanced. Electrons can be manipulated, like storing or moving, and be used to produce electricity.
This is where one object has an excess of electrons and the other objects has a shortage of electrons. The object with the excess is negatively charge as it now has more negative electrons that positivly charged protons. There is an invisible force field called an electric field between two charged objects. The object with a shortage of electrons attracts the object with the excess electrons. When we have such a situation, where we have two objects with opposite charge we say there is an electrical potential, or a difference of potential between them. This difference is called voltage. If these two charged objects come to close to one another the electrons jump the gap between them and create a spark. This is how lightning occurs for example. The earth is positively charged and the clouds negatively
This is the flow of electrons. The electrons flowing from one place to another is called current flow. Voltage, the difference in the charge between to object is what causes the flow. Electronics is about controlling the electrons with special components and circuits.
This is movement of electrons.
Voltage is what pushes electrons around a circuit. Without Voltage the electrons will move randomly in any direction.
(Spänning in Swedish) is the difference in charge between two points. This is
measured in volt (V) and the symbol used is U from the German word unterschied
that means difference.
Electrons flow from the negative terminal of a voltage source around the curcuit
as they are attracted by the positive terminal.
In the beginning voltage was known as Electromotive Force (EMF) and is the
reason for using the Ohm's lay uses E as the symbol for voltage.
Is the rate at which charge is flowing. Is measured in ampere (A) and the symbol
used is I which comes from the French word intensite de courant which means
current strength.
The number of electrons that move past a point in a conductor during a specific period of time is measured in coulombs (C). One coulomb of charge is equal to 1 Coulomb = 6.242x10^18 electrons
If 1 coulomb moves past a point in 1s, we say that the current is 1 ampere (A) 1 Ampere = 1 coulomb/s
Is a materials tendency to resist the flow of charge (current). Is measured in ohm using the symbol used is capital omega Ω and resistance uses the symbol R.
One thing that I got wrong initially was that if we look at the following circuit:
(Resistor)
+------/\/\/\\-------+
| |
+ | |
----- ---
- -- \ / (LED)
| _v_
| |
+--------------------+
Now, my understanding was that electrons flow out of the negative terminal of the voltage source through the LED where they cause the LED to shine. So the resistor would not be of any use in this case, it would have to be placed before the LED in my way of thinking. This is not how electricity works. Instead what happens is it is the voltage difference that pushes the electrons through the LED. To find out the current (I) that flows through this circuit we use current = voltage/resistence. So the pressure, the pushing of electrons, through the circuit will be limited by the resistor. Think of this like a pipe that is narrower than the pipe of the circuit, this causes the flow of water to slow down. It does not matter if the resistor comes before or after the LED.
Ohm's lay is more important that I initially thought when thinking about current.
voltage
current = ----------
resistance
voltage = current x resistance
voltage
resistance = -------
current
Notice that current is the voltage difference divided by the resistance of the circuit. TODO: expand on this a little more.
In a ciruit with one battery we refer to the negative terminal as ground. It is the point that has the lowest potential in a circuit. This type of ground is sometimes called reference ground, common ground, or floating ground.
Floating ground is a type of ground in which the ground doesn't have a physical connection to the earth; it simply serves as a type of 0V reference line that serves as a return path for current back to the negative side of the power supply.
Floating ground really means a 0V reference line. What is meant by a reference point is similar to when we measure our own hight, we measure from a certain point, most often the ground or floor. Same with voltage, we measure a voltage at a point in a circuit from the ground reference.
And in schematics instead of drawing lines that should be connected to the negative terminal we simply use the ground symbol. For example:
+----------/\/\/--------+
| |
+ | |
----- |
--- |
| |
| |
--- ---
- -
. .
Which is the same as writing:
+----------/\/\/--------+
| |
+ | |
----- |
--- |
| |
| |
+-----------------------+
Ground is a place in a circuit that has 0V and is used as a reference point when talking about other voltages in the circuit.
+--------------(A)
|
+ | 9V
-----
---
|
|
+--------------(B)
|
|
+ | 9V
-----
---
|
|
+--------------(C)
So anyone of the points A, B, or C could be selected to be ground, yes even A which I found strange at first as it not connected to the negative terminal but in this case if A is ground then B is -9V and C is -18V. And if we choose B to be ground the A will be 9V and C -9V. And if we choose C to be ground the A will be 18V and B 9V.
Protective earth, on the other hand, is used in high-voltage AC circuits to provide a safe path for unexpected current, protecting people and property from electrocution, failure, and fire. Earth ground is where there is a physical connection to the earth. This will cause electrons to pass through. Only devices that are connected to AC mains have an earth ground.
voltage = Resistance * Current U = R * I I = U/R R = U/I
Is the rate, per unit time, at which electrical energy is transferred by an electric circuit. The unit of power is watt which is one joule per second.
Is open (broken) by default so no current flows
The opposite of normally open.
All batteries have a voltage of 1.5 V. 9V batteries are simply 6 such batteries that are connected in a series.
Do electro magnetic waves carry energy?
Yes, think about what happens when you sit in the sun, you get warm and if you
are like me your skin will become red and burn. We absorb that energy from the
sun.
Recommended Standard 232.
Is a protocol for inspecting/testing microcontrollers. This is a standard that goes back to the 1980 where manufacturues ran into problems when components were becoming smaller and it was not as easy to access pins on their devices for testing. What was used was a bed of nails test system which I think were pins that needed to be in contact with the CI for testing. This is the reason for the Test in the name JTAG, it was for testing pins on a chip that has JTAG built into it.
There was also an issue with the higher performace of signals which I think lead to issues where having test probes which disturbed the signals on the device. So these companies came together to come up with a standard piece of hardware embedded on the chips to enable this kind of testing. This is called the JTAG Port which is the test access port.
+----+ +-----------------------------------------+
|TMS |----------------------→| [Boundry scan ]←+ |
|TCK |----------------------→|+----+ [B |
|TDI |---------------------->||JTAG| +----------------+ o |
|TDO |<----------------------||CTRL|←-------→|Flash Controller| u |
|TRST| || | +----------------+ n |
+----+ || |←---+ d |
|+----+ | +----------------+ r |
|| +---→|Debug Controller| y |
|| +----------------+ |
|| s |
|| c |
|| a |
|| n |
|| n |
|| ]|
|| ↑|
|+→[Boundry scan ]-+|
+-----------------------------------------+
Notice that the JTAG can talk to the flash controller which enables writing to flash storage memory.
The connection to the Debug Controller was not part of the original standard but something that was added afterwards. This enables us to monitor and debug the internals of the processor.
This is additional circuitry that is in between the I/O pins and connects all the pins. By default these cells do nothing and just pass the signal on the pin through.
This allows the the pins to be read/written. So we can read all the values from the pins and send it out, or we can write a value to the pins. This can be used for checking the that the pins are connected(soldered?) properly.
+---+
|TMS |------------------+-----------+
|TCK |----------------+-|---------+ |
|TDI |-------------->+-----+ +-----+
|TDO |<----+ |TMS | |TMS |
|TRST| | |TCK | |TSK |
+----+ | |TDI |---->|TDI |-----+
| +-----+ +-----+ |
+---------------------------------+
TMS = Test Mode Select
TCK = Test Clock
TDI = Test Data Input
TDO = Test Data Output
TRST = Test Reset (optional)
Notice that there can be multiple microprocessors connected and debugged.
Is a protocol for inspecting microcontrollers and is propriatary to ARM.
+-----+ +-----+
|SWDIO|<----->|SWDIO|
|SWCLK| |SWCLK|
+-----+ +-----+
In this case we can only debug a single microcontroller.
Notice that SWD only requires two pins where as JTAG required 4 pins. Another difference is that while both support programming and debugging only JTAG supports Boundry scanning.
SWD like I mentioned above is only for ARM, where as JTAG is supported for other devices as well. But also remember that ARM uses a licening model and there are a lot of implementations out there so it wil be available on a lot of devices.
SWD has a but called Debug Access Port (DAP) and there is a master, the debug port (DP) and then one or more access ports (AP). The debug port communicates with a specific access port by specifying the access ports address in the packet it sends.
+-----+
+-------+ | DAP | +--------+ JTAG
|Laptop |---------->| |----->| JTAG-AP|--------->
+-------+ SWJ_DP | | +--------+
JTAG-DP | |
SW-DP | | +--------+ AHB
| |----->|AHB-AP |--------->
| | +--------+
| |
| | +--------+ APB
| |----->| APB-AP |--------->
| | +--------+
+-----+
SWJ-DP = Serial Wire/JTAG Debug Port
This uses the standard JTAG interface to access the DAP.
JTAG-DP = JTAG Debug Port
This port can use either JTAG or SWD to access the DAP.
SW-DP = Serial Wire Debug Port
Uses the SWD protocol to access the DAP.
Multiple access ports can be added to the debug access port and ARM provides spec for two:
- MEM-AP provides access to core memory and registers.
- JTAG-AP allows for a JTAG chaing to be added to the DAP.
Are small hardware modules which provide the right kind of signaling (JTAG and/or SWD like discussed above). So this would be connected to the device/board with a USB to the host computer doing the debugging.
This is a protocol spec and an implemenation firmware that supports access to Debug Access Port. So you connect an USB to the board and can then use a debugger with the device. TODO:
$ git clone https://github.com/stlink-org/stlink
$ cd stlink
$ cmake .
$ makeIs a computer arch where there are separate buses for data and instructions. This is in contrast to a Von Neumann arch which uses the same bus for both instruction and data.
All microprocessors have clocks in them or use one that is outside of the microcontroller itself. The microprocessor needs the clock to execute our program, to the rythm of the clock. So the microprocessor could execute one instruction for every tick of the clock. We also have timers and counters in the system.
A counter is a device that records the number of occurences of a particular event. A timer is used to generate a delay.
But we know that the microprocessor is capable of a huge number of clock ticks per second, so if we have a counter based on that it would generate a very large number. For this reason the microcontroller provides a feature called prescaling which is a way for the counter to skip a certain number of microcontroller clock ticks. So one could set the prescaler value to something like 256 and that could cause the counter to only count/increment every time the clock ticks 256 times. So if we have a clock that tick 1000000 times per second we would have 1000000/256 = 3906 counts of the counter per second. So if we used this counter in a program and checked that it was 3906 then we could perform an action every second.
A timer most often has a control register, and a register for the count number itself.
Registers are temporary storage locations inside the CPU that hold data and and addresses and I know how they are used. But what is not clear to me is how they are actually implemented. This section attempts to explain this.
Now, the registers are contains in what is called a register file which is something that always confused my, like what is mean by a file in this case?
Tri-state buffer:
a
|
+\↓
| \
[0 1 1 0 ] --->| > ----> y
| /
+/
a = 0 do nothing
a = 1 pass through the vector of binary values
y = binary values or the binary values will not flow through (unchanged)
A GPIO pin looks something like this:
enable line
-----------------+--------+
| |\↓(inverter)
Output buffer | | \ +------+
-----------------|------| /-------+------| pin |
↓/| |/ | +------+
input buffer / | |
-----------------\ |--------------+
\|
The enable line controlls if this pin will acts as an output of input buffer.
This is used in many circuits so that the same output line can be shared.
A GPIO pin can assume one of three values:
- Logical 0 (connected to ground (0 volatage, on))
- Logical 1 (connected to VCC (positive voltage, off))
- High-impedance (floating, Hi-Z, or 'tri-stated' (disconnected)
* (enable)
|\↓
| \
*----| /------* (output)
(input) |/
This acts as a switch where is "enable" is on, then the switch is closed, and if "enable" is off it is open. Notice that when the switch is closed the input is not affecting the output so this removes this input from the output. If we have multiple inputs connected to the same output we can have it such that only one of the inputs affects the output at a time.
When a signal floats is means that it is neither connected to ground or to VCC, and the signals volatage is indeterminate. For example, say we have a pin on a microcontroller which we hook up with a button. When we press the button, lets say we have set things up so that the pin will go low (0V). So when we press it will be 0V and we can read the voltage from our program and act when the pin is 0V. But what is the value when we are not pressing. It is actually going to be random 0 or 1. This is a bad thing.
Is really just a resistor, but it is the way it is used that give it its name:
5V
|
\
/ R
\
| +------+
| | MCU |
+- \------------*| |
| +------+
Gnd
So when this circuit is open the pin will be read as 5V (on). When it is closed it will be read as Gnd (off). But it will not be random on/off.
pull-up.s is an example of what can happen if we have GPIO pin in input mode and the type as open-drain with out any pull-up or pull-down resistors
$ minicom --baudrate 115200 --device /dev/ttyUSB0 -H
Welcome to minicom 2.7.1
OPTIONS: I18n
Compiled on Jan 26 2021, 00:00:00.
Port /dev/ttyUSB0, 11:08:09
Press CTRL-A Z for help on special keys
01 01 01 01 01 01 01 01 01 01 00 00 00 00 01 01 ...Notice that we are randomly reading 01/00.
Now if we connect a pull-up resistor:
And run the example we get:
$ minicom --baudrate 115200 --device /dev/ttyUSB0 -H
Welcome to minicom 2.7.1
OPTIONS: I18n
Compiled on Jan 26 2021, 00:00:00.
Port /dev/ttyUSB0, 11:08:09
Press CTRL-A Z for help on special keys
01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01 01Now we are reading 01 consistently.
Is pretty much the inverse of a pull-up resistor:
+-------+
| MCU |
5V -----/ ---*---| |
| +-------+
\
/
\
/
↓
Gnd
In this case when the ciruit is open the pin will be read as Gnd (off) and when it is closed it will be read as 5V(on). So again it will not be random on/off.
Pull-up and pull-down are mostly used in interfaces that have unidirectional lines like SPI and UART (there is a one-to-one connection, compare this with I²C which can one controller can be connected to multiple peripherals).
Below is an example which can be used with pull-up.s to show what using a pull-down resistor might look like:
And if we run this example the output will be:
$ minicom --baudrate 115200 --device /dev/ttyUSB0 -H
Welcome to minicom 2.7.1
OPTIONS: I18n
Compiled on Jan 26 2021, 00:00:00.
Port /dev/ttyUSB0, 11:08:09
Press CTRL-A Z for help on special keys
00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00Now we are reading 00 consistently.
Each pin, which are physical points on the IC, have circuitry which is called a pad. This circuitry can be configured to interface with different types of digital and analog circuits.
This is an output configuration of a Pad which has two output levels.
When the output goes low, the signal is actively pulled to ground, and when
the output goes high it is actively pushed to VCC.
Push phase:
------ Vdd
| PMOS
| +--+
|-o| |
| | +--+
Vin ----| |--- Vout
| | +--+
|--| |
| +--+ NMOS
|
↓
gnd
Vin = Voltage In (Gate in put voltage)
VDD = Voltage applied to Drain
If Vin is 0, that will get inverted by the PMOS, and from what we've learned
(TODO: link of move this closer to the MOSFET notes), this will cause the value
out Vout to be taken from Vdd (what every voltage level a logical 1 has). The
value of Vout is pushed.
If Vin is 1, that will get inverted by the PMOS, and from what we've learned
the value will be take from ground, or pulled to ground. The value of Vout is
pulled from ground.
In this case the pin only has two states, GND or floating, which does not sound very useful, but it can be combined with a pull-resistor. So I think this allows for multiple components connected to the same line, like in I²C.
Compared to Push-Pull Open Drain does not contain an PMOS transistor. Instead if only has an NMOS transistor:
----- Vout
|
| +--+
Vin ----| |
| +--+ NMOS
|
↓
gnd
In NMOS if a voltage is applied to the Vin this will cause a closed circuit and current will flow between the source (Vout) and the drain (gnd). When a zero voltage is applied to the gate this will cause an open circuit and no current will flow.
So if we apply 1 to a pin that is in open drain this will cause current to flow
between the source and the drain. This this case we say it drains the current
and is where the Drain part in Open Drain comes from.
And if we apply 0 this will cause an open-circuit which means that electricity
is not flowing from the Gate to the Source but instead from the Gate to the
Drain. This is where open comes from in the name Open Drain.
NMOS -> ON (1) -> Vout -> Logic 0
NMOS -> OFF (0) -> Vout -> Floating
Floating here means that we don't know what value might be seen on the pin at
this stage, reading this pin could be 0 or could be 1 and change over time.
Why would we use something like this?
Well lets take a look at what happens if we connect multiple of this with each
other:
Vdd
/
\ (Pull Pull Resistor)
/
|
+-----Vout+---------+
| |
| +--+ +--+ |
Vin ----| | | | ---- Vin
| +--+ +--+ |
| |
↓ ↓
gnd gnd
In the stm32 we have the GPIOx_OTYPER register which allow us to configure pins:
Bits 31:16 Reserved, must be kept at reset value.
Bits 15:0 OTy: Port x configuration bits (y = 0..15)
These bits are written by software to configure the I/O output type.
0: Output push-pull (reset state)
1: Output open-drain
So by default the pins will be in push-pull state and perhaps we will need to set it to open-drain when using I²C.
If we chose push-pull the pin can be further configured using GPIOx_PUPDR:
Bits 2y+1:2y PUPDRy[1:0]: Port x configuration bits (y = 0..15)
These bits are written by software to configure the I/O pull-up or pull-down
00: No pull-up, pull-down
01: Pull-up
10: Pull-down
11: Reserved
Does this mean that if we don't configure our pin in GPIOx-OTYPER and don't configure it in GPIOx-PUPDR that it will floating?
Means that current is flowing into the pin. Opposite or current source. Think kitchen sink.
Means that current if flowing out of the pin. Opposite of current sink. Think kitchen tap.
This is the rate at which a signal can change between high and low. In the stm32 we have GPIOx_OSPEEDR which has the following configuration options:
Bits 2y+1:2y OSPEEDRy[1:0]: Port x configuration bits (y = 0..15)
These bits are written by software to configure the I/O output speed.
x0: Low speed
01: Medium speed
11: High speed
Is really just a switch:
VCC Voltage Collector Collector
___
|
\
\ Resistor
/
|
*-----------------> Binary output signal
|
/
/
|
|
↓
When the switch is open the output signal will be binary 1. And when the switch is closed it will be zero.
These are circuits that take two voltages as input and compare them, and output
|\
Vin |+\
--------| \
| / --------> output
--------|-/
Vref |/
The comparator will compare Vin against the Vref:
Vin > Vref = output 1
Vin < Vref = output 0
The output is the same as the output of the open collector switch.
This input to this circuit is an analog signal which is constantly changing.
analog signal +--------------+ fixed analog value
------------------->| Sample/Hold |--------------------->
+--------------+
The output is what would be read by another component, for example an analog to digtial converter which needs to be able to sample the analog signal and that signal needs to be fixed for that time period.
So this is about converting an analog signal, which remember is a continous signal into a digital signal (two states off/on/high/low).
analog +-----------+ +-----------+ +----------+ digital
-------->|Sample&Hold|---->|Quantize |---->| Encoder |----------->
signal +-----------+ +-----------+ +----------+ signal
The sample stage is about sampling the analog signal at a specific time interval. So this would be reading the amplitude of the wave at a specific time and storing it.
The hold block just stores the value read in the sample stage. So the sample stage has read an amplitude, a voltage level, and this has been stored. This is now passed/read to the quantize stage.
ADCs are of specific bit sizes which has an impact on the resolution of the converter. If we only have one bit to describe the analog signal as a digital signal then we would only have two states. Imagine we have a sine wave that has a maxium amplitude of say 5V and minimum 0V. With just one bit we would get something like:
|
|
|
|
1 |______ ______
| |_____|
0 +------------------
We can see the ups and downs but if we wanted to turn this back into the analog signal it would not represent it very well.
If we have 2-bits instead then:
4 | _
3 | _| |_
2 | _| |_
1 |_| |
|
0 +------------------
The above diagrams are not really to scale but they help me to visualize how more bits help to represent the analog wave more closely. Just with two bits the digital representation of the wave looks more like the original. Adding more bits will improve this even further.
Resolution
2¹ = 0-1
2² = 0-3
2³ = 0-7
...
2⁸ = 0-255
This is called the resolution of the ADC.
Lets say that we have sampled and stored a value of 2.5V, and that we have a 3-bit resolution:
2.5V
----------------------+ |\ C₇
+--| \
Vref-------- 3.5V | | /------------------+
4.0V |------------|/ |
R₁ \ | |
/ +--|\ C₆ | +-----------+
\ 3.0V | | \ +---| |
|------------| /----------------------| Encoder |
| | |/ +---| |
R₁ \ | | | |
/ +--|\ C₅ | | |
\ 2.5V | | \ | | |
|------------| /------------------+ | |---------b₂
| | |/ +---| |
\ | | | |
R₁ / +--|\ C₄ | | |
\ 2.0V | | \ | | |---------b₁
|------------| /------------------+ | |
| | |/ +---| |
\ | | | |
R₁ / +--|\ C₃ | | |---------b₀
\ 1.5V | | \ | | |
|------------| /------------------+ | |
| | |/ +---| |
\ | | | |
R₁ / +--|\ C₂ | | |
\ 1.0V | | \ | | |
|------------| /------------------+ | |
| | |/ +---| |
\ | | +-| |
R₁ / +--|\ C₁ | | | |
\ 0.5V | | \ | | +-----------+
|------------| /------------------+ | |
| | |/ +-------+
\ |
/ |
\ ------ GND
| --
GND -----
--
So our 2.5V analog signal is read or passed in and we have a reference voltage of 4.0V for ciruit. This is the voltage used for the comparators and it is divided among the 7 comparators in this circuit. Recall that this is a 3-bit ADC so we have 0-7 values that can be represented. The comparators work in this way:
Vin > Vref = output 1
Vin < Vref = output 0
C₇ = 2.5 > 3.5 = 0
C₆ = 2.5 > 3.0 = 0
C₅ = 2.5 > 2.5 = 0
C₄ = 2.5 > 2.0 = 1
C₃ = 2.5 > 1.5 = 1
C₂ = 2.5 > 1.0 = 1
C₁ = 2.5 > 0.5 = 1
C₇ C₆ C₅ C₄ C₃ C₂ C₁ 0 (From GND on the Encoder)
0 0 0 1 1 1 1 0
I think the encoder will look at the inputs and search for the highest 1 value and that position will be the output binary number. So above we have C₄ that contains the first/hightest 1, 4 in binary is 100 so that would be what the endoder would output for 2.5V. The above type of ADC is called a Flash ADC.
This type of ADC is not as fast as the Flash ADC but is a lot cheaper. This is like a binary search used to approximate the target value.
In this case we also have a comparator but only one. Vin is the sampled and held analog signal to be converted:
2.5V
----------------------+ |\
CLK +->| \
↓ | /------------+
+------+ +------+ +->|/ |
|Logic | | DAC | | |
+>| |-->| |-+ |
| | |-->| | |
| | |-->| | |
| +------+ +------+ |
| |
+---------------------------------------+
The digital to analog (DAC) converter will create an analog signal that can be used as Vref for the comparator. This will be compared with Vin and the output will be passed by to the logic unit to modify the value of the next output input value to the DAC. Being a binary search it will either increase or decrease the value to be checked by half.
The output of the DAC will initially be 100:
Vref < Vin = 111
/
Vref < Vin = 110
/ \
/ Vref > Vin = 101
100
\ Vref < Vin = 011
\ /
Vref > Vin = 010
\
Vref > Vin = 001
This will take 3 clock pulses to complete and more with more bits.
Vcc and Vdd are the positive supply voltage to an IC or circuit.
Vss and Vee are the negative supply voltage to an IC or electronic circuit.
Is a circuit that converts from one form of enery into an electrical signal. The engery may be heat, light, motion, position, or sound (and perhaps other sources as well).
A sensor is a transducer that recieves and responds to a signal from a physical system.
An actuator is a divice responsible for controlling something.
Is when an outside force acts on an object, then that object will either immediately spring back to its orignal state, or it will somehow change. An example I read about was thinking of a paper clip which if you bend it enough it will stay in the bent shape, but if you only bend it a little it will snap back into its original state.
TODO:
TODO:
Is where the sender and receiver switch between narrowband frequencies. Both the receiver and sender know the squence to switch between. So the sender will send for a short burst on the current channel, and then both switch to the next frequency for the next. Each channel is used for a short time and the odds that another sender/receiver using the same channel is low. Data signals are difficult to intercept as the frequency hopping pattern is not known.
This was first invented in the second world war to get around the problem of radio jamming of torpedos and their controllers. Radio jamming is when too many competing signals are being transmitted over the same frequency. Actress Hedy Lamarr came up with the idea when a wavelegth was jammed, enough information could still be sent by changing the frequence (hop) up or down. After the was it was realised the frequency jamming is not just for military usages but lso radio, tv and other wireless communication also experience jamming. And its not just jamming but also makes the communiction more robost to noise and interference.
Related to IoT is that very little computation power is requrired to send or receive signals over the spread spectrum. And also in IoT deployments one migth have a large number of sensors all outputting conflicting radion signals.
There are variations of FHSS, for example LoRa uses a system known as Chirp Spread Spectrum (CSS).
Is a component that is capable of storing electric charge, like a battery but it works in an different way and cannot hold the charge for the same amount of time as a battery. Also where a battery stores charge using chemicals, a capacitator stores enery in an electric field. The capacitor can be charged quickly and also release the energy quickly (much faster than a battery).
One can think of the setup where we have a water pipe that has a valve which can be used to turn off and on the flow of water. If we add a tank to this pipe, the valve coming before the tank, we can allow the tank to fill, and even if we close the valve water will still flow out of the tank for a while (until the tank is empty). If we open the valve before the tank is empty we can make sure that we have a steady flow of water.
A real capacitor has a container and inside it we have anode foil and a cathode foil with a separator between them called dielectric insulator.
+----------->------------->--+
| +-----------+ | +---------------+
^ | Capacitor | +->| Battery |
| |* * * | | |
+----<+-----------+ | |
| Dielectric| | |
+---->+-----------+ +-<| |
^ | ** | | +---------------+
| +-----------+ |
+-----------<-------------<--+
* = electrons
We have a few properties that are important here and these properties decide the capacity of the capitor (how much charge it can hold):
1) Area of the conductive plates
2) Distance between the plates
3) ε₀ the permittitvity of the free space (vacuum between the plates)
4) εᵣ the dielectric constant (of the insulator material between the plates)
A
C = ε₀ * εᵣ * -
d
So the size of the plates and the distance between them, and also the type of insulator material between them affect the amount of charge that a capacitor can store. I was not obvious to me what the use of dielectric value was but this is the ratio of charge that is permitted on the suface of the conductor (where the electrons and "holes" gather). When a dielectric is present the permittitivity becomes Kε₀, so we are increasing the capacity by factor kappa. Kappa can be influenced by "influencing" the electric field in the capacitor. For examples a fingers interaction with the capacitor's electric field represents an increase in the dielectric constant and this causes an increase in the amount of charge that the capacitor can store. This is way of making a touch sensor if we then have something that monitors the capacity, when it increases something is influencing the electric field and this could then send a signal to cause and action to happen.
Notice that there is a non-conductive layer in between the conducive plates preventing the flow of electrons. Hooking up a battery will eventually lead to the capacitor having the same voltage as the battery, the electrons in the lower compartment in the diagram above and no more current will flow. So we have a build up of electrons on one side:
+----------->------------->--+
| +-----------+ | +---------------+
^ | Capacitor | +->| Battery |
| | | (+) | |
+----<+-----------+ | |
| Dielectric| | |
+---->+-----------+ +-<| |
^ |***********| (-) | +---------------+
| +-----------+ |
+-----------<-------------<--+
* = electrons
We have a difference in potential between the positive and negative which is a potential voltage. The positively charged particles (electron holes) attract the negatively charge electrons and it is the electric field that keeps the electrons in place. The material used for the dielectric (insulator) has a "dielectric constant" k (kappa, that is a symbol no the letter k).
Now if we connect an LED (and it should have a resistor in this but skipping that for now) electrons will have a way to flow:
+--+----------------------------+
| ↓ +-----------+ |
^ | | Capacitor | \
| | | * * | (+)
+---+-->---+ +---->+-----------+ |
|LED| +----<+-----------+ |
+---+--<---+ ↓ |***********| (-) |
| | +-----------+ |
+<-+----------------------------+
* = electrons
And they will flow until they fill up the positive side of the capacitor. When the build up is equalized then no electrons will flow as the voltage is zero. When we connect the battery again electrons will start to build up in the negative side of the capacitor and holes will be created in the positive side as electrons flow out (attracted to the positive terminal of the battery):
+--+------>------------->-------+
| ^ +-----------+ | +----------+
^ | | Capacitor | +->| Battery |
| | |***********| (+) | |
+---+-->---+ +---->+-----------+ | |
|LED| +---->+-----------+ +-<| |
+---+--<---+ ^ |* | (-) | +----------+
| | +-----------+ |
+<-+----------------------<-----+
* = electrons
This will again charge the capacitator. So the power supply can be interrupted without effecting the LED (in this example).
Is a component that stores energy in it's magnetic field. Recall that a capacitor also stores energy but it stores electrons so there differ.
⌒⌒ ⌒ ⌒ ⌒ ⌒
-------/\/\/\/\/\-------------
⌒ ⌒ ⌒ ⌒ ⌒ ⌒
⌒ = magnetic field
/\ = inductor
Think of a water pump which pumps water through a pipe. The pipe branches like this:
+------------+
+-----| Water Wheel|---->---+
^ +------------+ |
+------+ | |
| Pump |--->----+ ↓
+------+ | |
| ↓ +--------------+ |
^ +----| Reducer |--->---+
| +--------------+ |
| ↓
+-------<---------------<----------------+
Reducer = resistance in an electrical circuit (like an LED)
Wheel = inductor
The water wheel takes a certain amount of pressure to start moving so initially water will start flowing and it will take the path of the water wheel, but then as it take alot of pressure to flow that way, water will start to flow through the reducer path as it ccan flow straight throw this back to the pump to close the loop (think circuit here). But when the wheel starts to move faster there will be a point when that path provides no resistance and water will flow only through that path and not through the reducer path. Now, if we turn off the pump the water wheel will continue to turn for a while and water can still flow through and continue to flow in a loop through the reducer until the wheel slows down (remember that the reducer could be a LED in an electric circuit so it would continue to shine for this period of time):
+------------+
+-----| Water Wheel|---->---+
^ +------------+ |
+------+ | |
| Pump |---/ | ↓
+------+ ^ |
| | +--------------+ |
| +-<--| Reducer |---<---+
| +--------------+ |
| |
+----------------------------------------+
Reducer = resistance in an electrical circuit (like an LED)
Wheel = inductor
So, similar to how we could disconnect the battery of a capacitor and current would continue to flow, the same is true in the case of an inductor.
+--+------>------------->-------+
| \ | +----------+
^ / +->| Battery |
| \ | |
+---+-->---+ / | |
|LED| \ +-<| |
+---+--<---+ / | +----------+
| \ |
+--<----------------------<-----+
Initially electrons are going to flow through the circuit (the outer loop) and turn on the LED as that it is the path of least resistance. But once the inductors resistance reduces it will be come the path of least resistance and the electrons will prefer that path (inner loop) meaning that the LED will turn off. Now, if we disconnect the battery in some way (using a switch perhaps) the electrons will no flow in a loop through the inductor to the LED and back until the inductor disapates the energy (like the water wheel slowing down and eventually stopping);
+--<------<---------------------+
| \ | +----------+
↓ / +--| Battery |
| \ | |
+---+--<---+ / | |
|LED| \ +--| |
+---+-->---+ / | +----------+
| \ |
+-->------↑------------- \-----+
Now the magnatic field that is generated by the inductor will collapse into electic energy when the current stops flowing. It will reach its max magnatic field and once the current stops flowing start to collapse which will release electrons (I think) into the wire which allows current continue to flow.
This section will introduce touch sensors and try to explain how they work. Recall from the section on Capacitors these circuits are made up of two conductive plates (with a certain area) separated with an insulating/non-conducting material between them (with a specific distance) called a dielectric. And the electric field between these plates is what keeps the electrons in place.
For touch sensors the amount of charge the capacitor is able to store is not really important, it is the change in capacitiy that is of interest.
Normally if a capacitor is used one choses it having a fixed amount of charge that is can hold. This value is determined by the area of two conducting plates, the distance between the plates, and also the dielectric constant of the material between the plates. The area and the distance is not something that can be changes with easy (if possible at all) but we can influence the dielectric constant. So imaging the electric field between the two plates which are not just in the region in between them but also above/below. The dielectric material could then be air for part of this region. And air has a certain dielectric constant. Now what if we replace the air with something with a higher dielectric constant. That would increase the capacity of the capacitor! It turns out that water has a higher dielectric constant than air and the human body contains water, so a finger move in this area will cause this change to happen.
A finger is also conductive and while the circuit is insulated there is still an effect introduced by this. The finger acts like a second conductive plate (connect to a virtual ground) of an additional capacitor. This capacitor is in parallel with the first capacitor and they therefor add their capatitance. So both of these factors play a part, the dielectric constant is increased by the finger which increases the capacity of the capacitor in the ciruit, and the finger also acts as a second capacitor plate which also increase the capacitance.
So we have this change in the amount of charge that the capacitor can store but we need to be able to detect this change in some way.
+---------+
|MCU |
|I/O Port ---------*-------------+
| | | |
| | R₀ \ ----- C₁
| | /
| | \ -----
| | | |
| | +------*------+
| | |
+---------+ -----
-- GND
Now if first configure the I/O port as output we can charge the capacitor C₁ to its logic high voltage. Now, we want the capacitor to discharge through the resistor R₀. But if we simply set the output pin to 0V that will provide an output with a low-impedance connection to the ground node, so current would flow through it instread of through the resistor. This would cause the capacitor to discharge too quickly and the MCU would not be able to detect it. Instead we need a high-impedance pin that will force most of the current to discharge through the resistor. This can be done by configuring the pin as input. Upon switching this we also start a counter whose purpose is to measure the time to discharge. A normal discharge will take a certain amount of time, but when the capacity of in increased there will be slightly more charge to be discharged. This will cause the discharge when the sensor is "touched" to take a little more time. The clock used will be able to show this difference with a different in the number of clock cycles it takes to discharge.
So we repeatedly charge and discharge the capacitor while monitoring the discharge time and if that time exceeds a predefined threshold the MCU takes that as though the sensor has been "pressed". When the discharging voltage crosses the pins logic-low threshold, for example 0.6V, the timer value will be sent to the pin and stored in a register (not 100 clear how this works yet). If the time saved in this register exceeds a pre-determined value the microcontroller assumes that the sensor was "pressed".
We know about the two states in a digital circuit as ON/OFF and in binary as 1 or 0 and these sometimes called HIGH and LOW as well.
The strenght of a signal is measured by its voltage level. So what voltage is a ON/HIGH/1 and what voltage is OFF/LOW/0. Well these are defined by the makers and we have to refer the specs.
5V ----- Vcc
|
|
|
|
|
|
2.7V ----- Vₒₕ Min OUTPUT voltage for a HIGH signal
|
|
2V ----- Vᵢₕ Min INPUT voltage to be considered a HIGH signal
|
|
|
0.8V ----- Vᵢₗ Maximum INPUT voltage level to still be considered LOW
|
0.4V ----- V₀ₗ Maximum OUTPUT voltage level to still be considered LOW
|
0V ----- GND
INPUT level: OUTPUT Levels:
HIGH: 5V-2V HIGH: 5V-2.7V
LOW: 0.8V-0.V LOW: 0.4V-OV
If we have a voltage between 0.8V and 2V we don't know if it is high or low, and this is called a floating voltage and can be read by a microcontroller as alternating between 0 and 1.
3.3V ----- Vcc
|
|
|
2.4V ----- Vₒₕ Min OUTPUT voltage for a HIGH signal
|
2V ----- Vᵢₕ Min INPUT voltage to be considered a HIGH signal
|
|
0.8V ----- Vᵢₗ Maximum INPUT voltage level to still be considered LOW
|
0.5V ----- V₀ₗ Maximum OUTPUT voltage level to still be considered LOW
|
0V ----- GND
INPUT level: OUTPUT Levels:
HIGH: 3.3V-2V HIGH: 3.3V-2.4V
LOW: 0.8V-0.V LOW: 0.5V-OV
Is a component that can take an input voltage and divide it into smaller voltages. For this two resistors can be used.
+5V
---
|
/
\ R₁ 1Ω ----+
/ | I = V/R
| |
V₁->* +---> 1 + 4 = 5Ω I = 5/5 = 1 Amp
/ |
\ R₂ 4Ω ----+
/
|
|
--- GND
-
.
So we have 1 Amp flowing through both resistors. Now, if we take a look at the voltage a point V₁ compared to ground we use Ohm's law and get:
V₁ = I * R
V₁ = 1Amp * 4Ω
V₁ = 4V
If we measure point V₁ and ground we will get a voltage of 4V. Now, we can use this to figure/divide the voltage range into values. For example we can say that our min represents all zeros, and our max voltage all ones:
max voltage -> 11111111
...
10000000
...
min voltage -> 00000000
This can be used for analog to digital convertes.
+5V
--- ---+
| |
0 A | / | 0V (V = 0 * 10kΩ)
| \ 10kΩ |
↓ / |
V₁* ---+
/ (open switch) Reistance of an open switch is infinity
| ---+
| | 5V (V = I * R)
--- GND ---+
-
.
Notice that the resistance of the open switch is not defined as there is no way for current to actually flow and nothing to resist. If we think of this circuit as a voltage divider it would be like having only two states, 0000 and 1111. So the second part will get the full voltage of 5V.
Now, lets look at a closed switch:
+5V
---
|
0.5A | /
| \ 10kΩ
↓ /
V₁*
| (closed switch) Reistance of a closed switch is 0
|
|
--- GND
- 5V/10kΩ = 0.5mA
.
If we measure the voltage of V₁ we get V₁ = 0.5A * 0 = 0. So notice that with the resistor we get a fairly low current when the switch is closed.
+Vcc
----
|
/
\ R
/
V₁ |----
|
S /
|
--- GND
-
.
V₁ is going to be equal to either the voltate of the pull up resistor if the switch is open, and equal to ground if the switch is closed. So either a logic one or logic 0. If the switch is open it is called that it is pulled up to the Vcc, and if it is closed it is pulled down to GND.
Now, take a look at what we can do if we place multiple of these next to each other in parallel:
+Vcc
----
|
/
\ R
/
Signal->|-------------+-----------+
| | |
S / / /
| | |
--- GND --- ---
- - -
. . .
So the only way for the signal, which is a voltage, to represent a logic 1 is if one of the switches is close. We don't know which one is closed just that one is closed. This acts like AND, if one is closed then the signal is 1 and if all are 0 then we know they are all open. This can be used to allow devices/component to communicate in half-duplex, only one can communicate at a time but the communication can be bidirectional. This is something that is used by the I²C protocol.
Was designed by Borsch and originally for automobile systems back in 1986. A single CAN package can move up to 8 bytes of data.
While this is also a bus like SPI and I²C there is no concept of a single controller (I²C can have multiple controllers but is seems to be a little messy to set up and CAN is designed for having this).
CAN is not a device to device (or circuit to circuit) communication system like SPI and I²C are. Instead CAN uses broadscasting so any node on the bus can read the data.
The CAN bus has two wires and looks something like this:
CAN Bus
High signal +---------------------------+--------------+
/ | /
120Ω \ | \ 120Ω
/ | /
Low signal +---------------------------|---+-----------
| |
+------------+
|Transciever |
| Tx Rx |
+------------+
| |
+------------+
| Tx Rx |
| |
| MCI |
+------------+
The resistors at both ends are terminating resistors. To connect a microcontroller to the bus a transciever is needed. So to connect two boards we will need two 120Ω resistors and two transceivers. I've got two MCP-2551 CAN modules.
There is no chip select wire like in SPI, and also there is no address like we find in I²C.
Signaling:
5.0V +
High | ----- ----- <--- Dominant state (Active)
| / \ /
| / \ /
| / \ /
|----- ------ <--- Recessive state (Idle)
2.5V |-----\ /------\
Low | \ / \
| 1 \ 0 / 1 \ 0
| ----- -----
0V +--------------------------------------------
Dominant state is when the lines are pulled apart, and recessive is when the lines are pulled together. Recessive state is the state when the bus is idle and notice that the state is a logic 1 in this case and a dominant state would be logic 0. This means that only on CAN Mode (transciever + mcu in our case) can force CAN- HI to 0 by using a logic AND:
Node1 Node2 Dominant
0 & 0 0 (ok, recessive state)
1 & 0 0 (ok, dominant state)
0 & 1 0 (ok, dominant state)
1 & 1 1 (two nodes cannot request the dominant state)
Sending a 0 is will always `win" which is why this is called dominant.
Message format:
+---+----------+---+--+----+---+---+---+---+
|SOF|Identifier|RTR|R₀|Data|CRC|ACK|EOF|IFS|
+---+----------+---+--+----+---+---+---+---+
Bits: 1 11 1 0-64 16 2 7 7
SOF = Start of Frame, 0 signals a new packet (the dominant state)
Identifier = Priority of the message, lower value = higher priority so
00000000000 is the highest priority and 11111111111 the lowest.
RTR = Remote Transmission Request.
IDE = Single identifier extension, 1 = standard extension.
R₀ = Reserved bit.
DLC = Data Length Code contains the number of bytes of data being transmitted.
CRC = Cyclic Redundance Check contains the checksum.
ACK = One bit is the ack and one is the delimiter.
EOF = End Of Frame, marks the end of a CAN frame.
IFS = Inter-Frame Space, is the time required to move a received frame into its
proper message buffer area.
The Identifier is more like an tag/topic for the content in the message which
receivers can use to determine if they are interested in those types of
messages. Notice that messages can have a 0 byte payload (Data field) which
can act a signaling message.
The STM32 have packet buffers for incoming and outgoing messages which are
called mailboxes. There are three for outgoing messages and 6 for incoming.
For incoming messages filters can be configured to specify which messages the device is interested in and the other messages are ignored. The messages that match a filter are passed to a FIFO queue.
Is a way of transmitting an electrical signal from sender to receiver. The electrical signal is transmitted by a voltage and we can think of this type of signaling of the circuits discussed so far, we have two conductors
TODO:
The program we write is stored in non-volatile (volatile means likely to change but in this case it is negated so is not likely to change) memory, that means that even if the power goes down the program still remains in this memory chip.
The program is stored in Flash memory (non-volatile). Data memory uses SRAM (Volatile) or EEPROM(non-volatile).
Is a one time programmable memory chips. So it programmed during production and after that it cannot change. After this point the program cannot be changed.
Is also a one-time programmable memory like ROM but using a device called a programmer it can be written to again (programmed). So I think this means that instead of having to write to the chip during production it can be done at a later stage. But it can only be programmed once.
In this case the contents can be erased using UV rays and the chip can be re-programmed/re-written.
Similar to EPROM but uses electrical voltage to erase. With this technology it is possible to read, write, or erase a particular byte or word of data at a time (or 8 or 16 bits at a time). This is what makes it a little slow.
Is a type of EEPROM which is programmed and erased in large blocks in stead of bytes/words.
Headers on a PCB (Printed Circuit Board) board are internal ports. So this is nothing to do with software headers.
This is a feature that allows a microcontroller to execute code straight from external flash memory (without having to copy it first). So if the code size is to big for the on-chip storage a memory mapping of the flash drive can be done and if it can use QSPI to improve the speed can be close to that of the on-chip storage so it can be used to execute code as well (and not just for storage of data).
Is a program used to load and run other programs. These programs run as one of the first things upon booting and load the code to be run.
Is a standard interface for industrial applications between a master/controller and slave/perhipheral. Is a point-to-point connection and all devices use a single CPU bus to share data and clock. Since this is synchronous there is a clock signal/line:
+----------+ +----------+
|Controller| |Peripheral|
| Clk |--------|Clk |
| Datq |--------|Data |
+----------+ +----------+