SD Card will not mount

54f37a1928a96a1c

I appear to have everything working apart from the SD card, the OLED screen just says ‘Insert SDCard’
MIOS Studio also says card not present.
The SD card is a class 10 16Gig SanDisk Ultra, brand new out of the pack and it works when inserted into the laptop, files written to it and read back OK.
I have had the core board out and checked for solder faults, none found, there are no shorts between pins and the headers to the MCU appear OK as does the MCU around the removed components. Any ideas where to start on this one?
Side question — should there be any config file or anything on it?

e8abac78cdf65c56

well, try a other SD-Card…. i plugged in a brandnew card from reichelt.com 8GB size into my matrix, and it worked…
booth worked:
https://www.reichelt.de/at/de/microsdhc-speicherkarte-8gb-intenso-class-10-intenso-3413460-p126586.html?&trstct=pol_1&nbc=1
https://www.reichelt.de/at/de/microsdhc-speicherkarte-8gb-transcend-class-10-ts8gusdc10-p261369.html?&trstct=pol_14&nbc=1

what i know from my Midibox-past is that everything should be cleaned on the PCB (isoprophyl alkohol) sometimes (or more often) the Flux in the solder wire has a impedance, specially the SD-Card-Connection dont like this. - so clean all MCU-Headers-Points, and the SD-Card Socket, and everything that has to do with the SD-Card-Slot.

while some things dont care Zero about Impedances, and other do care extrem > sd-card is such a canditate.

17703227d9536da1

You could also check connections to the MCU board as there might be a cold joint for the header socket or a slightly bent pin.

(viewed from the bottom)

54f37a1928a96a1c

Well, I have checked all solder joints on the SD socket, re-washed with IPA, checked continuity from SD socket pins to top side of MCU board (so proving PCB tracks as well as joints, I can find no fault.
I have tried a second SD card (unfortunately, same make and type as the first) with no joy.

I am now left wondering if the socket is working!
I will try a different make of SD card as soon as I can get one to rule out some sort of compatibility issue but the two I have tried both work OK on my laptop … 🤣

17703227d9536da1

Strange! It should simply mount and write the necessary config files on first boot.

Have you tried to format the card in Windows/FAT? Sometimes that can help. You could also see if you can access the card through the MIOS browser.

54f37a1928a96a1c

Config files? did I miss a step somewhere?
I loaded MatriX-project-v100b3.hex via MIOS Studio and see “Matrix v1.00b1” reported in MIOS Studio, all LED, slider and button tests are good, MIOS Studio shows a MIDI message every 5 seconds or so.
BUT
At no point have I had the SD card working on the MatriX.
I have just formatted the SD card as FAT32 with allocation unit size of 32k (the default settings) but to no avail.

MIOS Studio still simply says “SD Card not mounted!” and Matrix says “Insert SDcard”, reinsert card, reset MatriX, restart MIOS - same results.

0fc5608d995b44ff

@Mechie
A config file is written after first boot, automatically by MatriX - it contains all configuration settings (MatriX can be fully configured via its menu system - no text editor necessary) - this e.g. stores the configured LED brightness and so on.

An empty SDHC card (no SDXC) with a FAT32 filesystem is all that is needed for the first start.

Two ideas:

a) cold solder joint on the core board, just as Andy described - please measure the connections from every SD card pin to the top of the Waveshare board. I’ve seen plenty of interconnection problems to the waveshare MCU, so this is a valid potential problem.

b) (my main suspect): a potentially “burnt” or damaged sd card slot - it happened to me before - then an insertion event of a SD card is not detected anymore - it might be necessary to replace the sd card holder.

Before going into a replacement operation (a hot air reflow station highly recommended, e.g. an 858D clone for around 30€ is absolutely sufficient), please check with more/different SDHC cards, that are formatted with FAT32 - it might be that a particular card does not work, but to me it sounds like a broken SD card slot.

You’ll find it! Good luck and many greets!
Peter

54f37a1928a96a1c

Well, today I tried a new SD slot, I hoped to dissect the old one after removal but it fell apart during removal 🙁
New slot, four different cards, same result - MIOS studio reports “SD Card: not connected”.
I have checked for shorts - none found
Continuity from SD slot pin to core pin is good
Reflash MatriX-project.hex (in the hope it might help) - same results!
There is minor solder mask damage from the removal but the tracks are still healthy.

EDIT core board PB2 shorted to 3v3 - that can’t be right can it?
There is a solder bridge on the underside of the core labelled “BOOT1” linking centre pin to 3.3v is that correct?

0fc5608d995b44ff

@Mechie - wow i think you’re onto something here!
BOOT1 is shared with PB1, which is used by the SD card - i just checked our stock of STM32F4s and the BOOT1 solderbridge is always between GND and the center pin, not 3.3V and the center pin like shown in your second picture.

Could you resolder it to bridge between GND and the center pin and try again? Sorry for the inconvenience! This is the first time we see this - i will examine all STM32F4s when i flash them manually from now on for the correct solder bridge (the bridge is soldered by Waveshare).

Best regards and have a good evening!
Peter

54f37a1928a96a1c

And MOIS Studio says… … …

Reading Root Directory

Retrieving SD Sard information - please wait!
SD Card: '': 3038 of 3038 MB free

and a second (used) card gives a list of files stored on it!

… … … OK, I think I’m back in the game here 😀
Just one question - PB2 is now permanently shorted to 0v, why not just connect the SD card’s CS/SS to 0v at the SD slot, remove the solder bridge on BOOT1 and allow PB2 to be used for anything that may be a future feature?

17703227d9536da1

😌 Very glad to hear it, sorry about the extra effort! I would not have expected that and I congratulate you for looking closer into it!

As to why, well, we tried to make everything compatible to standard settings, just in case there would be an occasion to drive a different peripheral (it is like an SPI port). If you’ve ever looked at the back of a LoopA Core PCB it’s almost fully wired in. In the case of MatriX, the same basic Core circuit is in use, so it was just a carry-over from other variations.