MBHP Core STM32F4 Debugging

[alfadeo]

  I have assembled my SeqV4 modules (mbhp_core_stm32f4, wilba control surface, 2x MIDI I/O, Quad IIc R2, AOUT_NG R1, DOUT R5, PCBs from SmashTV).

 

The STM32F4DISCOVERY board seems to work fine, no problem with the bootloader and I have successfully uploaded MIDIbox SEQ V4.088 via MIOS Studio.

 

After connecting the STM32F4DISCOVERY board with the MBHP Core the LCDs and the SD card work as expected.

 

Troubles begin as soon as I try to connect other modules (wilba control surface, MIDI I/O, Quad IIc tested, wilba/MBSEQ_HW.V4 on SD card root):

 

Wilba control surface (J8/9): I get the LCD message asking to create a new session, but the core does not react to the control surface. No LED response on the surface (L31 is mounted in reverse).

 

MIDI I/O (J11E): No response to MIDI in/output, if I start the sequencer in MIOS Studio I get timeout error messages:

 

  [MIOS32_MIDI_Receive_Handler] Timeout on port 0x20

  [MIOS32_MIDI_Receive_Handler] Timeout on port 0x21

 

Quad IIc: The core does not boot (no LEDs on, no MIDI ports at computer)

 

It is also strange that the CPU always runs at ~ 230% (obtained by sending the "system" command from MIOS Studio). And I have the impression that the whole board gets rather hot (I have no comparisons though).

 

I have tried different ribbon cable orientation (very confusing for newbe), but the effect was the same. So I concluded that I have messed up the MPHP core and resoldered suspicious looking joints (having messed up 4 modules is quite possible but less likely). In particular I have tried to remove a suspected short at the left pins (see attached image) of the SD card connectors. I am also not sure, if they should be soldered to the PCB or not).

 

The only effect was that LCDs stopped working when I connected the control surface. 

 

Searching for tests I found the troubleshooting apps in the mios32 source code. After a few adjustments I am stuck at (I am on 4.0.4-2-ARCH Linux, MIOS Studio compiled without problems):

 

  Creating object file for check.c

  check.c: In function 'CHECK_Pins':

  check.c:160:34: error: 'GPIO_Mode_IPU' undeclared (first use in this function)

     GPIO_InitStructure.GPIO_Mode = GPIO_Mode_IPU;

                                    ^

  check.c:160:34: note: each undeclared identifier is reported only once for each function it appears in

  check.c:180:36: error: 'GPIO_Mode_Out_PP' undeclared (first use in this function)

       GPIO_InitStructure.GPIO_Mode = GPIO_Mode_Out_PP;

                                      ^

  check.c:229:36: error: 'GPIO_Mode_IPD' undeclared (first use in this function)

       GPIO_InitStructure.GPIO_Mode = GPIO_Mode_IPD;

                                      ^

  /home/ch/src/mios32/trunk/include/makefile/common.mk:160: recipe for target 'project_build/check.o' failed

  make: *** [project_build/check.o] Error 1

 

Indeed GPIO_Mode_IPU, GPIO_Mode_Out_PP and GPIO_Mode_IPD are defined in drivers/STM32F10x/v3.3.0/STM32F10x_StdPeriph_Driver/inc/stm32f10x_gpio.h, but not in drivers/STM32F4xx/v1.1.0/STM32F4xx_StdPeriph_Driver/inc/stm32f4xx_gpio.h.

 

Sorry for the massive amount of probably unrelated details. I am pretty much stuck at the moment and would appreciate any sort of help. 

 

I am attaching images of the (i) MBHP solder joints, (ii) SD card joints, (iii+iv) two ribbon cable connections, I would consider (iii) to be correct but most images I have seen would point to (iv).

 

Best,

Christoph

 

[TK.]

Hi,

 

Wilba control surface (J8/9): I get the LCD message asking to create a new session, but the core does not react to the control surface. No LED response on the surface (L31 is mounted in reverse).

 

did you also upload the hwcfg/wilba/MBSEQ_HW.V4 file into the root directory of the SD Card?

Without this special configuration neither buttons, nor LEDs will work.

 

 

MIDI I/O (J11E): No response to MIDI in/output, if I start the sequencer in MIOS Studio I get timeout error messages:

 

  [MIOS32_MIDI_Receive_Handler] Timeout on port 0x20

  [MIOS32_MIDI_Receive_Handler] Timeout on port 0x21

 

strange! The timeout indicates that something is received, but that the incoming MIDI stream is corrupted.

Does this even happen if no MIDI device is connected to the board?

 

 

Quad IIc: The core does not boot (no LEDs on, no MIDI ports at computer)

 

This board has to be connected to J4A - are you sure that the ribbon cable is connected correctly?

 

 

Indeed GPIO_Mode_IPU, GPIO_Mode_Out_PP and GPIO_Mode_IPD are defined in drivers/STM32F10x/v3.3.0/STM32F10x_StdPeriph_Driver/inc/stm32f10x_gpio.h, but not in drivers/STM32F4xx/v1.1.0/STM32F4xx_StdPeriph_Driver/inc/stm32f4xx_gpio.h.

 

Which app are you trying exactly?

It seems that it hasn't been prepared for STM32F4

 

Best Regards, Thorsten.

[alfadeo]

Hi,

 

 

did you also upload the hwcfg/wilba/MBSEQ_HW.V4 file into the root directory of the SD Card?

Without this special configuration neither buttons, nor LEDs will work.

Yes.

strange! The timeout indicates that something is received, but that the incoming MIDI stream is corrupted.

Does this even happen if no MIDI device is connected to the board?

Yes this happens without a MIDI device. If I connect an external MIDI device no events are sent or received.

This board has to be connected to J4A - are you sure that the ribbon cable is connected correctly?

J4A is correct, but I am unsure about the ribbon cables (for all boards). Do you know good instructions how to assemble them correctly?

Which app are you trying exactly?

It seems that it hasn't been prepared for STM32F4

I was trying to compile  apps/troubleshooting/mbhp_core_stm32_module in order to check the core module, but now I realize that this is not the same as  STM32F4. My fault. Are there any other tools that could help to debug my core module.

 

Thanks a lot for your help!

Christoph