NUCLEO F411RE timer in order to generate an interrupt

[Roman_E]

Learning on how to code 4x4 keypad with F411RE I came to need to learn 

timer in order to generate an interrupt (saying simply C function should be called lets say every 1 second) 
I created STM32 project CubeIDE+CubeMX  selected TIM3 (in Timers)  -> set Clock Source to InternalClock, set Prescaler ( PSC... ) to 8399 ( googled, found example ),
set Counter Period to 9999 and set NVIC checkmark for TIM3 (enabled)
 Saved IOC file and CubeMX generated some code for me.
In stm32f4xx_it.c file I added 

void TIM3_IRQHandler(void)
{
  /* USER CODE BEGIN TIM3_IRQn 0 */

  /* USER CODE END TIM3_IRQn 0 */
  HAL_TIM_IRQHandler(&htim3);
  /* USER CODE BEGIN TIM3_IRQn 1 */
  HAL_GPIO_TogglePin(LD2_GPIO_Port, LD2_Pin);
  /* USER CODE END TIM3_IRQn 1 */
}

and running the code I expected to see blinking LED ( at least wanted no matter 1 sec or 0.5 sec )

but nothing happens
Did I miss something ? ( I'm learning this stuff )

[Gyessine]

Hello @Roman_E 
Hope you enjoy your time learning about timers
you can follow the steps of this article. It will help you learn how to do 1 second interruptions with timers
BR
Gyessine

[Roman_E]

Thank you for your advice !
As I'm learning on how to use 4x4 keypad with F411re, I understand , that this interrupt function should be called periodically. Frequency it should be called based on Prescaler&Period ( at simple case + can be configured via CubeMX) 

htim3.Init.Prescaler = 8399;
htim3.Init.CounterMode = TIM_COUNTERMODE_UP;
htim3.Init.Period = 4499;

What would be your suggestion on "frequency to call the interrupt function" ?
I understand that in this function I need to add code "to scan 4x4 rows and columns". But how often this function should be called ? Every 10 ms ? 20ms?

[Andrew Neil]

I think your questions here - generating a timer interrupt - has now been answered?

Is so, please mark the solution on the post which provided the answer.

You already have a separate thread on 4 x 4 keypad with Nucleo 411RE and HAL, and @gbm has already given a suggestion for the scan rate there...