Parallel flash programming for STM32 ?

[pandit]

Posted on May 19, 2013 at 14:21

Hello,

         I was just going through some old datasheets for my AT89S51 microcontrollers based on the 8051. I came across parallel flash programming i.e. directly programming the flash through hardware, without going through the CPU as is done with the ICP/IAP routines. Any idea how this is done with the current generation of ARM micros or is it just an outdated method of programming ? I couldn't find any other method for the STM32 except ICP/IAP. Comments......??

#flash-programming #parallel

[Tesla DeLorean]

Posted on May 19, 2013 at 14:31

Given the number of bits associated with the address and data lines it's rather impractical to expose these. Let along the issues with multiplexing the pin functions.

JTAG would permit access to internal buses, but the most practical method is to place code in RAM, along with data blocks, and run that code to deliver the data to the flash array.

[pandit]

Posted on May 19, 2013 at 14:41

That's certainly a problem (the vast number of bits...) which you don't encounter in the 8051. But placing data blocks in the RAM would also be done THROUGH the CPU, if you get what I mean, wouldn't it? I was wondering if there's a way to bypass the CPU altogether. JTAG is probably one way to do that.

[Tesla DeLorean]

Posted on May 19, 2013 at 15:31

I wouldn't spend too much time worrying about what is in the ''CPU'' and what is outside. The JTAG chains are ultimately going through flip-flops on the die and can be used to set and sequence activity both inside and outside the core. There is a trade-off here in the number of cycles required to do different things, and executing code fragments at operational speed to effect the result you want. The goal however is to do this with optimal speeds, not how the ''CPU'' fits into the equation, or is separated from the process.

Remember also that things within the die are able to operate at speeds at least an order of magnitude faster than something like an 8051, and at least twice as fast as the external pin interface

[pandit]

Posted on May 19, 2013 at 17:39

Hmmmm.........

Thanks a lot. I know that any way of programming the STM32 will be faster than the 8051. Just was curious to know about the different methods available.

[pandit]

Posted on May 19, 2013 at 17:41

Oh by the way, is this process the same for ALL ARM based microcontrollers, like say the Fujitsu series, where there is no mention of any kind of bootloader?