Systematically powerDown() is not working for me (more precisely it returns a false)
powerUp() and other functions are working properly. The chip is a Winbond W25Q32BV

Is there support for MXL25L12835F?

Is there any reason that SPIFlash::writePage() uses a non-const parameter for its buffer?

_addressCheck() needs to modify a global variable. It currently only returns true/false. Make currentAddress a global variable.

---------------------------- DO NOT DELETE OR EDIT anything above this line ----------------------------

Add the ability for the library to clone from one chip to another - i.e. add the ability to use two chips at once.

Examples don't compile with IDE 1.0.5 and even using a Makefile and calling avr-g++ directly (and enabling gnu++11 and so on) it won't build.

If I add the #define ARDUINO_ARCH_AVR 1 to readWriteString.ino (should really be __AVR__ to do it automatically) I get a bit further but also need to change:

(void)xfer(PAGEPROG); at line 290 etc. to void xfer(PAGEPROG);

But then I get the error:

error: expected initializer before ‘.’ token
     #define xfer(n)   SPI.transfer(n)

Which I assume is because SPI.transfer() expects two arguments - a pointer and a value, or lack of trailing semi-colons in various places?

Nothing major and you'll definitely want to verify the choices I made for return value for the three functions missing a return.

spiflash.txt

Does this library support the Arduino Micro? I tried to test it with the Micro and a W25Q128FW chip using the TestFlash example and it seemed to work, but its not listed in the supported boards list.

I was wondering if it were possible to detect the following:

1. If the SPI chip is wired and initialized correctly.
2. If SPIFlash::begin() was called and completed successfully.

I'm modifying a FAT library that uses this library as a backend and it would be useful to know whether the chip and class have been initialized appropriately before calling read/write operations.

---------------------------- DO NOT DELETE OR EDIT anything above this line ----------------------------

STM32 Boards have more than one SPI port. The library has been reported to work with port 1. Would it be possible to let the user code pick which port to use through the library?

When using the Arduino-Makefile to build the example I get the following errors:

libs/SPIFlash/examples/getAddressEx/getAddressEx.ino: In function 'void setup()':
libs/SPIFlash/examples/getAddressEx/getAddressEx.ino:49:16: error: 'getAddresses' was not declared in this scope
   getAddresses();
                ^
libs/SPIFlash/examples/getAddressEx/getAddressEx.ino:50:13: error: 'writeData' was not declared in this scope
   writeData();
             ^
make: *** [build-micro/libs/SPIFlash/examples/getAddressEx/getAddressEx.ino.o] Error 1

This is due to missing function prototypes for getAddresses() and writeData()

That is a idea for enhancement, eliminate all shifts from code, that will increase average speed, just use those macros:

#define Lo(param) ((char *)&param)[0] //0x000y
#define Hi(param) ((char *)&param)[1] //0x00y0
#define Higher(param) ((char *)&param)[2] //0x0y00
#define Highest(param) ((char *)&param)[3] //0xy000
#define Low(param) ((int *)&param)[0] //0x00yy
#define Top(param) ((int *)&param)[1] //0xyy00

In my tests those macros reduce memory access in a 2000 loop from 200+mS to 2mS.

Able to communicate with Microchip flash (as of commit 25efc62 on the multiflash-compat-w.i.p branch). However, any data larger than a byte fails to read properly - (except structs and arrays). Needs more work!

All page_buffer variables in the library are uint8_t type. Need to allow for signed 8bit data.

SPIFlash::writeByteArray modify data buffer (storing data from SPI MISO into the buffer) and fails it's read after write verification. Write with error check always fails because of this.

void SPIFlash::_nextBuf(uint8_t opcode, uint8_t *data_buffer, uint32_t size) {
// snipped
#elif defined (ARDUINO_ARCH_SAMD)
#ifdef ENABLEZERODMA
spi_write(&(*data_buffer), size);
#else
_spi->transfer(&(*data_buffer), size); // this call destroy the contents of the buffer
#endif
#elif defined (ARDUINO_ARCH_AVR)
SPI.transfer(&(*data_buffer), size);

Allow library to overflow data across pages. Provide an option to let users decide if page overflow is not to be permitted.

Hi can you confirm this is how i would set this up.

#include<SPIFlash.h>

int ss = 10;
SPIFlash flash;

void setup() {
 Serial.begin(115200);
flash.begin(MB32); // is this correct
Serial.println("erasing chip");
flash.eraseChip();
delay(0);
flash.begin();
Serial.println(flash.readChar(22,44,false));// is this correct
}

void loop() {
 
}

The very first line of the readByte function should not have the braces and should return false.

The first line is the check for busy, out of range page/offset and should return false right after it.

If you compare the first two lines of the readByte and readBytes you will see the bug.

As it is today, the readByte functon only runs if the chip is busy or if you pass in an out of range page/offset.

#include<SPIFlash.h>
#include<SPI.h>
#define arrayLen(x) sizeof(x)/sizeof(x[0])
uint32_t strAddr[3], floatAddr[2], byteAddr[4];
String testStr[] = {
  "Test String 0",
  "Test String 1",
  "Test String 2"
};
float testFloat[] = {
  3.1415, 6.283
};
byte testByte[] = {
  3, 245, 84, 100
};

SPIFlash flash;

void getAddresses();
void writeData();

void setup() {
  Serial.begin(9600);
  Serial.print(F("Initialising Flash memory"));
  for (int i = 0; i < 10; ++i)
  {
    Serial.print(F("."));
  }
  Serial.println();
  flash.begin();
  
  Serial.println();
  Serial.println();

  getAddresses();
 
  writeData();
     Serial.printf("\n getMaxPage  = %d\n", flash.getMaxPage());
     
     Serial.printf("\n getCapacity= %d\n", flash.getCapacity());
      
      Serial.printf("\n getManID = %d\n", flash.getManID());
      
 
  //flash.eraseChip();
}

void loop() {

}
// Function to get adresses for various variables
void getAddresses() {
  for (uint8_t i = 0; i < arrayLen(byteAddr); i++) {
    byteAddr[i] = flash.getAddress(sizeof(byte));
    Serial.printf("error= %d\n", flash.error());
    Serial.print("Byte Address ");
    Serial.print(i);
    Serial.print(" : ");
    Serial.println(byteAddr[i]);
  }

  for (uint8_t i = 0; i < arrayLen(floatAddr); i++) {
    floatAddr[i] = flash.getAddress(sizeof(float));
     Serial.printf("error= %d\n", flash.error());
    Serial.print("Float Address ");
    Serial.print(i);
    Serial.print(" : ");
    Serial.println(floatAddr[i]);
  }

  for (uint8_t i = 0; i < arrayLen(strAddr); i++) {
    strAddr[i] = flash.getAddress(flash.sizeofStr(testStr[i]));
  Serial.printf("error= %d\n", flash.error());
    Serial.print("String Address ");
    Serial.print(i);
    Serial.print(" : ");
    Serial.println(strAddr[i]);
  }
}

// Function to write data
void writeData() {
  for (uint8_t i = 0; i < arrayLen(byteAddr); i++) {
    if (flash.writeByte(byteAddr[i], testByte[i])) {
      Serial.printf("error= %d\n", flash.error());
      Serial.print(testByte[i]);
      Serial.print(" written to ");
      Serial.println(byteAddr[i]);
    }
  }

  for (uint8_t i = 0; i < arrayLen(floatAddr); i++) {
    if (flash.writeFloat(floatAddr[i], testFloat[i])) {
     Serial.printf("error= %d\n", flash.error());
      Serial.print(testFloat[i]);
      Serial.print(" written to ");
      Serial.println(floatAddr[i]);
    }
  }

  for (uint8_t i = 0; i < arrayLen(strAddr); i++) {
    if (flash.writeStr(strAddr[i], testStr[i])) {
     Serial.printf("error= %d\n", flash.error());
      Serial.print(testStr[i]);
      Serial.print(" written to ");
      Serial.println(strAddr[i]);
    }
  }
}

========================================
out put

�cüÿInitialising Flash memory..........
error= 5
Byte Address 0 : 0
error= 5
Byte Address 1 : 0
error= 5
Byte Address 2 : 0
error= 5
Byte Address 3 : 0
error= 5
Float Address 0 : 0
error= 5
Float Address 1 : 0
error= 5
String Address 0 : 0
error= 5
String Address 1 : 0
error= 5
String Address 2 : 0

getMaxPage = 0

getCapacity= 0

getManID = 0

in the file SPIFlash/examples/TestFlash/TestFlash.ino
after doing the following...
//Uncomment this code block if your code doesn't run to check for Flash error code
i get the error below

command_list.ino: In function 'void setup()':

command_list:96: error: 'flashCode' was not declared in this scope

exit status 1
'flashCode' was not declared in this scope

for both uno & esp8266

Hi,

I'm using an Arduino Uno with a Winbond 32mB chip, W25Q256FV. I've been trying to get it to run with no success. I have already swapped the chip out and that changed nothing. I've been running the Flash Diagnostics page and it only works on special cases. I have to comment out the GetID and diagnose functions and upload. Then I can uncomment GetID and upload. Then finally I can uncomment diagnose and upload, but I have to have the powerFuncDiag commented out. If I don't do all those steps I get:

Error Code 0x04
Error Code 0x04
Error Code 0x04.

Which says could be a wiring problem, but I'm able to get it to work if I use all the steps above (besides PowerDown, that has never passed). When I do all of that stuff first I get the screenshot that I attached.

I'm using the sparkfun logic level converter you recommended in the arduino forums and I've checked my wiring 50 times at least. Would appreciate any help/wondering if there is a bug in the library. Thank you very much!

I get the following error when compiling for the ESP8266 in the Arduino IDE.

fatal error: util/delay.h: No such file or directory #include <util/delay.h>

Any thoughts on getting the code to compile?

So the constructor is SPIFlash flash(csPin); but how to create two constructors for two flash chips with different csPins and start them with .begin() in setup?

---------------------------- DO NOT DELETE OR EDIT anything above this line ----------------------------

Once a file system is setup on the Flash chip, the end of every file could possibly have a 16/32bit checksum to test the integrity of the file - without having to re-read the whole thing?

Add support for SPI.Transactions() so this library plays well with other libraries that use the SPI interface.

Hello,

With a Salea Logic 4 analyzer I was able to determine that I was missing the clock pulses when trying to read\write to my Windbond SPI memory chip.

I have this connected to the SPI bus of my ATMEGA644A microcontroller.

Through further testing I figure out that if I swapped my crystal from 8MHz to 16MHz I can get the clock pulses to generate and I am able to work with the memory chip.

According to my chips datasheet it seems like 8MHZ would be just fine all the way up to 100MHz.

What could be reasons why an 8MHz crystal does not work with the demo program?

Thanks,

Stephen

Have ATmega2560 wired up to W25Q64FV winbond chip. Using Hardware SPI pins and SS pin (53) for chipselect
Running XP.

If I use the SPIflash library from RinkyDink website the above setup is reported as working fine.

If I erase the chip using RinkyDink and then switch to your library and run your testflash sketch the only menu item that works is number 1 and it reports the correct chip ID for W25Q64FV.
All other menu items fail or appear to do nothing. ie EraseSector

EraseChip returns very quickly, the RinkyDink erasechip function takes quite a while.
WriteByte reports data failed. ReadPage reports 256 X 0xFF as you would expect but none of the commands that perform writes to the flash chip are working.

This is most strange given the large amount of positive feedback you have received from users on the internet. Is it me ?? Appreciate if anything obvious springs to mind.
BTW. Have also tried 2.4.0,wip download but results are the same.
Best Regards

I'm trying to compile the library on an ATTiny85 (adafruit trinket), and I'm getting quite some errors.

for example:

SPIFlash/defines.h:81:12: error: 'SS' was not declared in this scope
#define CS SS

I can tell there are some checks for AVR_ATtiny85 in the library, but I can't make the standard examples work out of the box. Am I missing something? Or has ATTiny support not been tested in a while?

Would it be possible to get this chip supported?
Looks like some of it works, although it might might be a false positive.

http://i.imgur.com/Oprl2li.png

Hi mate, I try using your library for my Cypress S25FL512S f.memory using Arduino MEGA and somehow it display error code 0x03. I edit some common instructions from defines.h code that match for my f.memory but still does not work. Can you give me some advice what to edit if wanna to use another f.memory (i.e Cypress f.memory for my case)? Really appreciate it if you can help me because I am still learning about this. Thank you.

Hello. I have an SST26VF064B flash chip from Microchip and am desperately trying to find a way to control it from an arduino UNO. No success so far, tried with multiple libraries including yours. Can you please add support for my chip in your library or perhaps point me to another library that supports it?

Hello!
I have plugged a Winbond W25Q256FV and run the diagnostic sketch (FlashDiagnostics.ino in the library example sketch folder).
The output says that reading chip is OK but all writings have failed...
The pins WP, HOLD and RESET are HIGH so the chip should be working.
I have uncommented the line 38 in SPIFlash.h :
#define RUNDIAGNOSTIC

Serial output :

Initialising Flash memory..........

--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
                                                                    SPIFlash Library version: 2.7.0
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
                                                                                Get ID
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
			JEDEC ID: ef4019h
			Manufacturer ID: efh
			Memory Type: 40h
			Capacity: 33554432 bytes
			Maximum pages: 131072
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
                                                                               Data Check
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
	Data Written	||	Data Read	||	Result		||	Write Time	||	Read Time	||	Write Time	||	Fast Read Time
			||			||			||			||			||	(No Error Chk)	||
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
	35		||	4		||	Fail		||	144 us		||	96 us		||	104 us		||	104 us
	-110		||	0		||	Fail		||	112 us		||	96 us		||	112 us		||	104 us
	4520		||	3072		||	Fail		||	120 us		||	104 us		||	120 us		||	104 us
	-1250		||	0		||	Fail		||	120 us		||	104 us		||	120 us		||	112 us
	876532		||	0		||	Fail		||	128 us		||	120 us		||	128 us		||	120 us
	-10959		||	16711935		||	Fail		||	144 us		||	120 us		||	128 us		||	120 us
	3.14		||	0.00		||	Fail		||	2008.9382 s		||	120 us		||	2008.9382 s		||	128 us
	123 Test !@#	||		||	Fail		||	2890.3596 s		||	264 us		||	2890.3596 s		||	3416 us
	inputStruct	||	outputStruct	||	Fail		||	144 us		||	184 us		||	144 us		||	184 us

I'm having issues with this library while working with other SPI devices.
There are circumstances where it isn't closing the SPI transaction after completing a task, which causes havoc while talking to other devices on the SPI bus that use a different SPI_MODE.
I did some digging and notices several instances where _endSPI has been commented out within the library. Uncommenting the _endSPI() in the _readStat1() function fixed my issues, but I thought ya'll might want to check it out.
-thanks.

Dear All,

I work with w25q64 module and ESP32-WROOM-32 via arduino IDE. From few days w25q64 can't start work. I deleted almost all code and I want only begin the module, but all time i recieve error code 0x04. I chech all connection and tried different combination of connection with different pins.
My connections:
CS -> IO5
CLK ->IO18
DI -> IO19
DO -> IO23

Please note that if I connect csPin to IO2 (as was said in extras in this repository) in ESP32 boatloader tell me

flash read err, 1000
Falling back to built-in command interpreter.
OK

And I should choose another pin, e.g. no 5. My code is below.

#include<SPIFlash.h>
#include<SPI.h>

#define csPin 5

SPIFlash flash(csPin); 

void setup() {
Serial.begin(115200);
while (!flash.begin()){
  if (flash.error()) {
          Serial.print(flash.error());
     }
}
}

void loop() {
}

If someone know what i do wrong please let me know.

Hello, I'm looking to get the uniqueID out of the chip to use for ID purposes. The specific chip I'm using is the W25Q16JV which I can successfully read/write to.

The dataheet http://www.winbond.com/resource-files/w25q16jv%20spi%20revf%2005092017.pdf

page 46:

The Read Unique ID instruction is initiated by driving the /CS pin
low and shifting the instruction code “4Bh” followed by a four bytes of dummy clocks.

But I have no idea how to add this! Any suggestions?

thanks!

Hi,
what I know from ur library support for 32-bit flash memory right?
Is there any possible u make library for 24-bit too?
Thanks mate.

@Marzogh
The function getAddress() is expected to return the next available address for use.

Say I have addresses 0 to 40 written with data already. Then I expect getAddress() to return 41. But it is returning 0.

Output of getAddressEx.ino:
Please note that I have data written till address 63
Just added an else condition with a print statement if any write in writeData() doesn't return 1

The output is as follows:

Initialising Flash memory..........

Byte Address 0 : 0
Byte Address 1 : 1
Byte Address 2 : 2
Byte Address 3 : 3
Float Address 0 : 4
Float Address 1 : 8
String Address 0 : 12
String Address 1 : 30
String Address 2 : 48
Data write failed
Data write failed
Data write failed
Data write failed
Data write failed
Data write failed
Data write failed
Data write failed
Data write failed

The write wouldn't happen because the addresses returned already have data and this is expected.
I am using v2.7.0 of the library with a Winbond external flashchip.

---------------------------- DO NOT DELETE OR EDIT anything above this line ----------------------------

library works fine with 25Q64 Winbond chip but I'm unable make the necessary modification to read/write or even get chip id from MX25L12845E
Using standard SPI I'm able to read the chip id and random segment of memory address.

---------------------------- DO NOT DELETE OR EDIT anything above this line ----------------------------

Hi,
I have a library called SPIFlash which was created many years ago.
Since you created your own "SPIFlash" with the identical same name, Arduino is causing naming conflicts. I suspect your users have the same issue.
Could you possibly rename your library to something else?

Hello!

I'm submitting a bug report, I tried to identify what's going wrong but I cannot fix it...

Bug report :

Library version : 3.0.0
Branch : master
Arduino IDE : 1.8.4
OS : Windows
Microcontroller : Arduino-nano (Atmega328P)
Flash memory module : W25Q64 freshly erased
Current behaviour :
Writing Struct type is not reliable (8% of errors) while writing ByteArray is reliable (0% of error)
Expected behaviour :
Writing Struct should be as reliable as writing ByteArray

Code snippet that reproduces the bug :
First sketch (not reliable) : part of the code that loops for writing Struct

    for (long k=1;k<10000;k++) {
      addr = flash.getAddress(sizeof(configuration));

      Serial.println();
      Serial.print ("addr = ");
      Serial.println(addr);
      
      if (flash.writeAnything(addr, configuration)){
      }
      else {
        Serial.println ("Struct write failed");
      }
    }

I observe a strange thing on the serial output : "Struct write failed" tends to occur each 12 Struct writings...

Second sketch (reliable) : part of the code that loops for writing ByteArray

for (long k=1;k<10000;k++) { 
      addr = flash.getAddress(sizeof(pageBuffer));

      Serial.println();
      Serial.print ("addr = ");
      Serial.println(addr);
      
      if (flash.writeByteArray(addr, &pageBuffer[0], SPI_PAGESIZE)) {
       }
      else {
        Serial.println("Page write failed");
      }
      
    }

Library needs ability to check that data is not being written without erasing previously written data

Allow users to declare the WP & HOLD pins at initialization and then let them enable Write protection and Hold via functions.

---------------------------- DO NOT DELETE OR EDIT anything above this line ----------------------------

Add the ability to update_dataType_(..., ..., data, ...) for the Arduino due.

Updating can be achieved by reading the sector/32kb block of data into a buffer in the Due's SRAM, modifying the required data bits in the SRAM and then writing it back to the flash chip.

This is not a true update, but a workaround the limit that the flash memory can only erase one 4k sector at the minimum.

Two quick questions about erasing sectors:

1. If I call the eraseSector, pass a page to erase, and the offset different than zero, what will happen?

• Ex: eraseSector(page 0, offset 128) will erase the page 0, starting from byte 128 (maintaining the info in the 0-127 bytes) until the page 15;offset 128, (maintaining the info in the 128-255 bytes)? - Or it will just erase all data in the pages 0 to 15??

2. The erase sector erases all data in 16 pages, what happens if I try to erase starting on the last page? It will overflow and erase data in the 0-14 pages as well?

Thanks!

---------------------------- DO NOT DELETE OR EDIT anything above this line ----------------------------

Hello!

I'm submitting a bug report, I tried to identify what's going wrong but I cannot fix it...

Bug report :

Library version : 3.0.0
Branch : master
Arduino IDE : 1.8.4
OS : Windows
Microcontroller : Arduino-nano (Atmega328P)
Flash memory module : W25Q64 freshly erased
Current behaviour :
I edited the example sketch readWriteString in order to repeat the Write/Read/Erase functions of the sketch but after 3 iterations the Arduino nano freezes.
Here is the serial output :

Written string: This is a test String
To address: 7025326
Read string: This is a test String
From address: 7025326
Written string: This is a test String
To address: 4996536
Read string: This is a test Str�
To address: 4996536

Expected behaviour :
The Arduino should loop until the for condition is reached

Code snippet that reproduces the bug (edited readWriteString example sketch) :

#include<SPIFlash.h>

uint32_t strAddr;

#if defined(ARDUINO_SAMD_ZERO) && defined(SERIAL_PORT_USBVIRTUAL)
// Required for Serial on Zero based boards
#define Serial SERIAL_PORT_USBVIRTUAL
#endif

#if defined (SIMBLEE)
#define BAUD_RATE 250000
#define RANDPIN 1
#else
#define BAUD_RATE 115200
#define RANDPIN A0
#endif

//SPIFlash flash(SS1, &SPI1);       //Use this constructor if using an SPI bus other than the default SPI. Only works with chips with more than one hardware SPI bus
SPIFlash flash;

bool readSerialStr(String &inputStr);

void setup() {
  Serial.begin(BAUD_RATE);
#if defined (ARDUINO_SAMD_ZERO) || (__AVR_ATmega32U4__)
  while (!Serial) ; // Wait for Serial monitor to open
#endif

  flash.begin();
  randomSeed(analogRead(RANDPIN));

    for (int i=0;i<1000;i++) {  
    
      strAddr = random(0, flash.getCapacity());
      String inputString = "This is a test String";
      flash.writeStr(strAddr, inputString);
      Serial.print(F("Written string: "));
      Serial.println(inputString);
      Serial.print(F("To address: "));
      Serial.println(strAddr);
      String outputString = "";
      if (flash.readStr(strAddr, outputString)) {
        Serial.print(F("Read string: "));
        Serial.println(outputString);
        Serial.print(F("From address: "));
        Serial.println(strAddr);
      }
      while (!flash.eraseSector(strAddr));
    
      delay(1000);
    }  
}

void loop() {

}

//Reads a string from Serial
bool readSerialStr(String &inputStr) {
  if (!Serial)
    Serial.begin(115200);
  while (Serial.available()) {
    inputStr = Serial.readStringUntil('\n');
    Serial.println(inputStr);
    return true;
  }
  return false;
}

Other information :

• I tried to show memory usage during loops but this doesn't help
  Here is the serial output :

freeRam()=1585
Written string: This is a test String
To address: 6773221
Read string: This is a test String
From address: 6773221
freeRam()=1557
Written string: This is a test String
To address: 4114840
Read string: This is a test Str�
From address: 4114840
freeRam()=1557

• I moved the Write/Read/Erase functions into the main "loop" function and the problem is the same
• I uncommented #define RUNDIAGNOSTIC in SPIFlash.h with no effect
• I have better result by passing outputString and inputString as global variables but there is still unexpected behaviour. Here is the serial output with unprintable characters :
  
  
  The Arduino reboots at each "strange" line

I must be sure that the library is reliable before using it in projects so any help is welcome ! 😀

Best to probably to set a ludicrously high _notBusy() loop time limit default. Will have to explore further.

Hi,

These functions (readPage & writePage) are listed in the documentation (i.e. the README.md) but do not actually exist (at least in version 2.7.0)?

Thanks,

Matt