Program won't run after upload

QwyntexAuthor

Explorer

I don't know how I missed that, thank you. I will try and solder it on and when it works I'll mark your answer as the solution

M

MM..1

Graduate II

Is your added VCAP electrolytic? If yes is you connection polarity right or better change to ...?

QwyntexAuthor

Explorer

I have soldered the capacitor and now the program gets past the SystemClock_Config function. The blink script still doesn't work though. After inspecting it with the debugger I narrowed it down to the HAL_Delay function, specifically inside it in the HAL_GetTick function that returns uwTick, which I am assuming should change but for me it stays at zero. I wanted to try the internal clock but the IDE keeps crashing.

Graduate II

Make sure HSE and PLL are set with the anticipation that it's using 25 MHz, check HSE_VALUE define. Make sure the clock starts, or use the initial HSI clock to run the part, at least until you've established things are working.

Have Error_Handler() and HardFault_Handler() indicate if you've died there.

PC13, PC14, PC15 are assumed to be Low Current, and in the backup/low power domain.

SysTick not working, double check the base address of the firmware, check vector table, and that you haven't disabled interrupts. Often SysTick set up in HAL_Init(), Vectors in SystemInit()

I

INaee.1

Graduate

Hi

according to your schematic , NRST has no pullup resistor.

connect like 4.7k pullup resistor.

I.N

P

Peter BENSCH

Technical Moderator

It is NOT recommended to use a pull-up at NRST, because it is already built-in.

You will find details in the reference manual.

[edit] Even though such pull-ups are familiar from MCUs of the last millennium and can be found in some horribly constructed Blue, Black or other Pills, this does not mean that they are actually recommended or useful. [/edit]

I

INaee.1

Graduate

Hi Peter ,,

Thanks for information.

regards

I.N

QwyntexAuthor

Explorer

HSE and PLL should both be expecting a 25MHz clock. I have probed the oscillator with an oscilloscope and it works fine. I don't know if that means the clock actually started nor how to check for that.
When I change the clock source to internal my IDE crashes whenever I try to have it automatically resolve the clock and manually changing it doesn't work either because it insists that something is wrong even if there isn't. Manually changing it in the code would probably work but I have not done that yet and don't know what parameters I would need to change where, I would probably introduce more issues (but I want to ditch HAL sometime anyways, so any good resources will help me later).

The error handlers don't trigger as it is just stuck in a while loop waiting for uwTick to increment.

I am not sure of the relevance of the pins being assumed to be low current. Do you mean that drawing too much current on them could introduce odd behavior? I have only one red LED and a 120ohm resistor connected to PC13.

For the last part I assume they are correct because if they weren't it shouldn't work on the same chip model on a different board, right?

This is what puzzles me the most, I can upload the same sketch to a different f411ceu6 chip and it works just fine. But it doesn't work on my custom PCB.

C

Chris21

Graduate

As Tesla mentioned PC13 is not a good choice to drive an LED.

Datasheet footnote:

PC13, PC14 and PC15 are supplied through the power switch. Since the switch only sinks a limited amount of current (3mA), the use of GPIOs PC13 to PC15 in output mode is limited:
- The speed should not exceed 2 MHz with a maximum load of 30 pF.
- These I/Os must not be used as a current source (e.g. to drive an LED).

Graduate II

>>IDE keeps crashing

What IDE, you keep mentioning "Sketches" which makes me think Arduino. Do you have some loader between you and the code you're running?

Is it compiling / linking for 0x08000000 operation?

QwyntexAuthor

Explorer

The IDE is the STM32CubeIDE. I guess I just internalized using the word sketches from Arudino and didn't want to write 'code' all the time. I'll just write 'script' or 'code' from now on.

As for the current drawing problem, I didn't know that. I've put the LED there because the dev board I have did, too. I'll change that in the next version but for now, is it actually the problem? Even when the LED should be off and isn't drawing any current the HAL_Delay gets stuck in its loop.

> Is it compiling / linking for 0x08000000 operation?

yes

QwyntexAuthor

Explorer

That I couldn't use the internal clock was because I had still enabled the controller as a USB device which wouldn't allow me to use the internal clock. After I changed that I could manually resolve the clock issues and because it was the dialog box that crashes the IDE I was able to generate the code.
It still doesn't work though, the HAL_Delay gets stuck because the uwTick doesn't increase. Uploading the same code without changes works without a problem on the dev board.

QwyntexAuthor

Explorer

Should I open a new post for this? It might make it easier to find for other people with the same issue and the original question was answered in the first reply by Peter.

Graduate II

You could, I suppose, but already a bunch of detail here, I'd press on..

The SysTick not working is a bit odd. Suggests some issue with NVIC or VTOR configuration, or simply not started. Was a different TIM selected for timing purposes? SysTick is a "System Handler" rather than and interrupt in the broader sense.

Is this example something you could share on GitHub?

Look at HAL_Init() code

F411, no issues are jumping to top-of-mind about odd failings..

QwyntexAuthor

Explorer

Interrupts for the System Tick timer are enabled in the NVIC and I don't think the issue lies in the vector tables either because it runs just fine on the F411 on my dev board (the MCU is exactly the same model, I chose it to not have to modify my code after receiving the PCB). Correct me if I am wrong but the vector tables shouldn't change from one physical chip to another only for different types, right?

There was, at the beginning, a different timer used but after I figured out the clock configuration issue had something to do with just enabling the USB I created a new project from scratch only enabling the HSE and PC13.
Same result. The code runs successfully on the dev board but not on my custom PCB.

That kind of also answers whether I can put it on github. I guess, I could but it really is just a bare bones blinking script.

The only other thing I can currently think of besides another design mistake in my schematic is that there is maybe some firmware or boot configuration flashed to the dev board that isn't on the MCU of my PCB, but I really don't know anything about that.

Graduate II

Well by default SCB->VTOR resets to ZERO, and then what BOOT0 pin maps/shadows there gets to determine what runs, ie FLASH, ROM, RAM,etc

The actual address then usually gets established in SystemInit() with a bunch of defines getting in the way of SCB->VTOR = 0x08000000;

You could perhaps re-gen the project, or use one of the NUCLEO or DISCO examples, watching to address the PLL and HSE_VALUE differences.

When I make a template project for a new board I tend to get a UART working with STDIO so printf() works, and get Error_Handler() and HardFault_Handler() to output actionable output via that channel. Also print out the MCU, AHB, and APB clocks so I know what the machine thinks it's doing.

Check that the Hard Fault works, good way to confirm the Vector Table is in order.

QwyntexAuthor

Explorer

I can't find any errors with the debugger and the Hard Fault seems to work. I didn't have time right now to get UART working, I will do that and print out the clocks if you think it's worth it

I just have a few other things to do right now