LGT8F328P clone I Would like to use the SPI interface.
as indicated on the datasheet:
It has everything necessary but I don't know how to configure <SPI.h>.
and how to make it works for example with W5500?
I found out the problem, using LGT8F328P-SSOP20 there are two pins identical!
the MOSI on 11 and SS on 10.
in this way I cannot communicate with SPI protocol.
This may help you:
https://github.com/dbuezas/lgt8fx/issues/54
Somebody else got SPI working on the SSOP20 package by overriding the default pin IO direction of PB0
For W5500 chip, you can using ethernet2 library. Edit the ethernet2.h file for replacing w5500_cspin to 9.
EthernetClass() { _dhcp = NULL; w5500_cspin = 9; } //default 10
void init(uint8_t _cspin = 9) { w5500_cspin = _cspin; }
This is specially for LGT8F328P-SSOP20
Related
I want to Initialize and send a single int using UART in blue_pill (STM32F10C8). Manual ask to set GPIO mode on ALTRN_PULL_PUSH in blue_pill. But Low level HALL library dosn't have such a option. Here is my code for initializing the UART:
void uart2_init()
{
LL_APB1_GRP1_EnableClock(LL_APB2_GRP1_PERIPH_GPIOA);
LL_APB1_GRP1_EnableClock(LL_APB2_GRP1_PERIPH_USART1);
LL_GPIO_SetPinMode(GPIOA, LL_GPIO_PIN_9, LL_GPIO_MODE_ALTERNATE); // ***this line should be corrected***
LL_GPIO_SetPinSpeed(GPIOA, LL_GPIO_PIN_9, LL_GPIO_MODE_OUTPUT_50MHz);
LL_USART_SetTransferDirection(USART1, LL_USART_DIRECTION_TX);
LL_USART_ConfigCharacter(USART1, LL_USART_DATAWIDTH_8B, LL_USART_PARITY_NONE, LL_USART_STOPBITS_1);
LL_USART_SetBaudRate (USART1, 8000000, 9600);
LL_USART_Enable(USART1);
}
I need to set pin mode to "LL_GPIO_MODE_ALTERNATE_PUSH_PULL" but the library dosn't provide
it. Can anyone correct me?
I think you should use LL_GPIO_SetPinOutputType(GPIO_TypeDef *GPIOx, uint32_t PinMask, uint32_t OutputType) function.
Which allows You to set output type of pin as :
LL_GPIO_OUTPUT_PUSHPULL
LL_GPIO_OUTPUT_OPENDRAIN
I am highly recommend You to search this kind of information in - UM1850, where You can find a lot of information, for e.g. there is a information about above function on page 807.
I'm trying to write code in order to generate software trigger. Datasheets seem confusing to me and I cannot find any example on NET.
In the "Cortex-M3 devices Generic User Guide" in 4.2.8 I read to write in NVIC->STIR register to generate a software interrupt:
[STIR register][1]
But, if I understood correctly USERSETMPEND must be set to 1 for unprivileged software (I think it's my case).
But If I want to access to SCR I have to be "privileged"
[SCR register][2]
So, how can I enter in privileged mode? Is it really necessary enter in this mode only for interrupt configuration?
What I've done in code is
SCB->CCR |= 0x02; // USERSETMPEND = 1. But I must be "privileged"
HAL_NVIC_EnableIRQ(EXTI0_IRQn); // enable interrupt line
.... some code....
... ...
if(my_software_trigger)
{
my_software_trigger = 0;
NVIC->STIR = EXTI0_IRQn; // triggers interrupt??
}
Is it the correct approach? Anyone has experience in this matter? Any help is really appreciatedd..
P.s. Sorry for my english
Marco.
Code below physically breaks nucleo boards. 2 so far.
ST links is unable to connect and boards are not detected.
first dead one was xncleo stm32f411re, another one is nucleo stm32f446re.
I'm a complete beginner and that is pretty much my first code.
and 2 boards dead;) The question is why it may happen? or can one reset the board harder that normal jumper reset
int main()
{
RCC->AHB1ENR = (1<<0);
GPIOA->MODER = (1<<5);
while(1)
{
GPIOA->ODR |= (0<<5);
}
}
This is similar to this question ->
https://electronics.stackexchange.com/questions/204996/stm32-st-link-cannot-connect-to-mcu-after-successful-programming
, answer to which helped me to restore the board, however answers to my problem in this thread are also very helpfull.
some pins used for debugging need to have some settings. if you change those settings the debugger cannot connect anymore. you need to set the debugger to connect under reset when the debug pins are in the initial corrct state. stm32 processors cannot be bricked this simple way.
so not listen to the advices that you need to use crappy STM libraries. just only set the pins you use, not the whole port.
Yes, your problem is with the line "RCC->AHB1ENR = (1<<0);". That enables power to GPIOA, but disables power to the other GPIOs (B,C,D,E, and H on the F411), which includes the SWD pins.
Edit:Reseting problem solved here: https://electronics.stackexchange.com/questions/204996/stm32-st-link-cannot-connect-to-mcu-after-successful-programming
the reason problem occured solved below
Ok, it doesn't physically break the board. Needed hard reset - shorting rst and sb11 pins on nucleo stm32f446re. didn't try on xnucleo yet.
User P__J__ got it right in the comment. I'll cite:
just do not assign to the moder. use |= or &= instead – P__J__
Done.
And the fixed code:
int main()
{
RCC->AHB1ENR |= RCC_AHB1ENR_GPIOAEN;
GPIOA->MODER |= GPIO_MODER_MODE5_0;
while(1)
{
GPIOA->ODR |= GPIO_ODR_OD5;
}
}
Hi I am developed a board based on the Discovery L476 board (STM32L476VGT6) using MBED and after porting it to VisualGDB everything works great. The only thing that doesn't work is LPUART1. I hooked it to PB10(LPUART1_RX), PB11(LPUART1_TX) but whenever I declare the port in my code and download it, the program hangs and doesn't even start:
Serial RS232(PB_11, PB_10);
If I remove this line, the code works great (but I can't use this port)
I changed the pin definitions in PeripheralPins.c so PB10 and PB11 will function as the LPUART TX and RX pins: (I added the lines)
const PinMap PinMap_UART_RX[] = { {PB_10, LPUART_1, STM_PIN_DATA(STM_MODE_AF_PP, GPIO_PULLUP, GPIO_AF8_LPUART1)},
//
const PinMap PinMap_UART_TX[] = { {PB_11, LPUART_1, STM_PIN_DATA(STM_MODE_AF_PP, GPIO_PULLUP, GPIO_AF8_LPUART1)},
but it still doesn't work. Any ideas?
See https://github.com/ARMmbed/mbed-os/issues/5389, the baud rate needs to be set at [sys_clk / 4096 ... sys_clk / 3]. Sys clock on this device is running at 80MHz. You could fix it in the HAL for this board until a real fix is deployed.
(I am really unsure of which tags to apply, so apologies in advance if I chose the wrong ones.)
I hope that this is not a too stupid question, but I'm really lost here.
A client lent me an Atmel SAMA5D2 Xplained board with Linux4SAM to play around with. I'm trying to interface the SPI interface on it, but I have no clue where to start.
To be clear, I've used similar boards (not this particular one) bare-metal. I've also used Linux for many years and wrote a few simple devices drivers some years ago. And still I'm lost.
dmesg | grep spi gives me the following output:
[ 1.840000] atmel_spi f8000000.spi: version: 0x311
[ 1.840000] atmel_spi f8000000.spi: Using dma0chan0 (tx) and dma0chan1 (rx) for DMA transfers
[ 1.850000] atmel_spi f8000000.spi: Using FIFO (16 data)
[ 1.860000] atmel_spi f8000000.spi: Atmel SPI Controller at 0xf8000000 (irq 32)
[ 1.860000] m25p80 spi32766.0: at25df321a (4096 Kbytes)
From this I infer that a driver is loaded and that it is configured to use DMA. Yet, looking in /dev/ there is nothing that looks like a SPI device (I was expecting to find something like /dev/spidev or /dev/spi32766.0 or similar.)
Does this mean that there is no actual device driver loaded? Do I have to write one in order to use the SPI?
If I look at the Makefile in the Linux4SAM source tree, I see around line 1171 that the kernel does not support loading of modules. Does this imply I have to recompile the kernel to include my new driver? This seems to be a silly approach; why providing a Linux distribution if I can't access the hardware with it?
What am I missing here?
(I feel rather stupid...)
EDIT To be clear: I want to access the external SPI interface that will be connected to some external device. I think the m25p80 is some internal Flash memory; I'm not interested to read/write there.
spidev is a standard Linux device driver which just exports a low level API to userspace via /dev interface
if you want to access specific SPI client(slave) you should write your driver
according to Linux SPI driver model:
static const struct of_device_id myspi_dt_ids[] = {
{ .compatible = "xxxx,yyyyy" },
{},
static struct spi_driver xxx_spi_driver = {
.driver = {
.name = "myspi",
.owner = THIS_MODULE,
.of_match_table = of_match_ptr(myspi_dt_ids),
},
.probe = myspi_probe,
.remove = myspi_remove,
e.g. define spi_driver struct, of_device_id struct, callbacks(hooks), and register it.
You should bind it to your DT node with compatible property string.
myspi#0 {
compatible = "xxxx,yyyyy";
spi-max-frequency = <5000000>;
reg = <0>;
};
second,
compatible = "linux,spidev";
Linux discourages using this compatible property because nodes should
describe real HW devices not userspace abstractions
#0andriy put me on the right track. I had to add a SPI resource to the Device Tree and flash the compile Device Tree Blob to the board. (Since I didn't know about Device Trees at all, this information is really hard to find...).
I now have a /dev/spidev32765.0. I added this to the device tree:
spi1: spi#fc000000 {
pinctrl-names = "default";
pinctrl-0 = <&pinctrl_spi1_default>;
status = "okay";
spidev#0 {
compatible = "linux,spidev";
spi-max-frequency = <83000000>;
reg = <0>;
};
};
pinctrl#fc038000 {
pinctrl_spi1_default: spi1_default {
pinmux = <PIN_PD25__SPI1_SPCK>,
<PIN_PD26__SPI1_MOSI>,
<PIN_PD27__SPI1_MISO>,
<PIN_PD28__SPI1_NPCS0>;
bias-disable;
};
};
Although I read that adding the spidev#0 is not really the right thing to do (I indeed see in dmesg output "spidev spi32765.0: buggy DT: spidev listed directly in DT").
Now, if I run spidev_test it still doesn't work (it times out), but I guess that's for another question.