STM32F4xx alternate configuration SPI1 issues

[gvigelet2]

Posted on May 19, 2014 at 14:04

Hello Everyone,

I was just wondering if anyone has seen this, I have seen this only referenced on one website.  I have SPI1 on a STM32F415 configured as 

  PA7   ------> SPI1_MOSI

  PB3   ------> SPI1_SCK

  PB4   ------> SPI1_MISO 

Which works fine until i configure PE3, as soon as I do that the SPI1 stops working if I comment the following out it will work again.

  /*Configure GPIO pins : PE3  */

//  GPIO_InitStruct.Pin = GPIO_PIN_3;

//  GPIO_InitStruct.Mode = GPIO_MODE_OUTPUT_PP;

//  GPIO_InitStruct.Pull = GPIO_NOPULL;

//  GPIO_InitStruct.Speed = GPIO_SPEED_FAST;

//  HAL_GPIO_Init(GPIOE, &GPIO_InitStruct);

I haven't seen no errata about it, am I the only one seeing this?

Thanks,

George

[waclawek.jan]

Posted on May 19, 2014 at 14:24

Post a minimal, but complete compilable example which exhibits the problem.

You might also want to post the link to the mentioned website.

JW

[gvigelet2]

Posted on May 19, 2014 at 15:11

Hi the following is a link to the minimal project, to exhibit the behavior un-comment the PE3 Configuration code in main::MX_GPIO_Init()

 

https://drive.google.com/file/d/0B9AbQf-Nlf0KQ0ZJOWFaVjlfUlE/edit?usp=sharing

Thank You,

George

[waclawek.jan]

Posted on May 19, 2014 at 15:50

Code looks reasonable.

What exactly are the symptoms of ''not working''? Did you check the signals with an oscilloscope?

Can't there be a hardware issue, e.g. PE3 shorted onto some of the SPI lines?

JW

[Tesla DeLorean]

Posted on May 19, 2014 at 16:02

I haven't seen no errata about it, am I the only one seeing this?

Assume you are, and provide details about the board or anything else unique to your situation that you would need to communicate to a third-party for them to be able to independently analyzer and verify the issue.

Most Cube/HAL related support is occurring on the other STM32 Forum page.

[waclawek.jan]

Posted on May 19, 2014 at 16:27

What is connected to PE3?

Don't you use a DISCOVERY board, by chance? There is an error in the table indicating the pins' connections, related to PE3.

JW

[waclawek.jan]

Posted on May 19, 2014 at 16:28

Related thread, but abandoned without resolution: https://my.st.com/public/STe2ecommunities/mcu/Lists/cortex_mx_stm32/Flat.aspx?RootFolder=%2Fpublic%2FSTe2ecommunities%2Fmcu%2FLists%2Fcortex_mx_stm32%2FSTM32F4%20Bug%20with%20SPI1&FolderCTID=0x01200200770978C69A1141439FE559EB459D7580009C4E14902C3CDE46A77F0FFD06506F5B&currentviews=205

[Tesla DeLorean]

Posted on May 19, 2014 at 16:47

Related thread, but abandoned without resolution

Jan thanks for the link. Indicates we're not alone, but still really not enough meat to suggest it's a chip errata and not a circuit board level issue.

[gvigelet2]

Posted on May 19, 2014 at 17:59

Thanks for all the responses, I will check with my STM324xG-EVAL board when I get home, the board that I am using is a custom board and I was able to reproduce it on the stm32f4discovery board (I will try that board again also when I get home).  I will look for a possible short.

Thanks,

George

[gvigelet2]

Posted on May 19, 2014 at 18:11

I have an Accelerometer CS line hooked to PE3, the issue is if I am not using the pin for anything and when I initialize the pin in the test program i set it as output with nopull (which should leave it floating).  I did verify that there is no shorts, I will need to hook up the scope again and see what it looks like in both states.

[waclawek.jan]

Posted on May 19, 2014 at 18:37

> I have an Accelerometer CS line hooked to PE3, the issue is if I am not using the pin for anything

I don't understand, so do you have the accelerometer's CS line hooked to PE3, or not?

If so, simply set PE3 to 1 when you set it to output.

JW