cancel
Showing results for 
Search instead for 
Did you mean: 

STM32 New Year wish list

gdp123a
Associate II
Posted on May 07, 2009 at 13:34

STM32 New Year wish list

45 REPLIES 45
janek
Associate II
Posted on May 17, 2011 at 12:58

One more thing.

I would like to see the 64 pin and even 100 pin devices in QFN packages.

Jan

mdeneen
Associate II
Posted on May 17, 2011 at 12:58

I would love a true 32 bit timer. Chaining 2 16 bit timers works, but you have to worry about the timer rolling over when you read it and getting invalid results.

There are also a few things which are poorly documented, or assume you are very familiar with ARM. Personally, I had some trouble finding information about SYSTICK and what exactly the various buses and clocks were. For example, what are APB1, APB2, SYSCLK and HCLK... and how they are impacted by the PLL settings. Honestly, the whole PLL setup is still a little confusing to me, and I wasn't able to find a lot of documentation on it.

Bit-banding is also very very cool, but you would almost not know it existed unless you went out of your way to look for it. Some more visibility would be great. You can replace GPIO_SetBits / GPIO_ResetBits with some macros which compile to just a few instructions. This is great for when you need to quickly toggle a bit in an ISR.

mcuisp
Associate II
Posted on May 17, 2011 at 12:58

My wish:

The trim step/resolution of HSI down to 10kHz(0.1%)/step.

Able to work under USB full speed with out a 8MHz crystal.

Note:the USB full speed allow a +-0.25% frequency error.

gdp123a
Associate II
Posted on May 17, 2011 at 12:58

Hi JJ Sprague

did you have a wish list? It was your suggestion that prompted this thread. Perhaps you were on holidays when it was active.

It's been very interesting hearing the wishes of others - and hopefully ST will find them valuable as well.

Greg.

jj
Associate II
Posted on May 17, 2011 at 12:58

Hi Greg-

Thanks - no holidays for us - worked/working just about everyday.

My goal was to make it easier for STOne-32 to see/know what his forum felt was ''hot.'' Believe that's been pretty much accomplished - thanks to all who participated.

My display firm - and most of our clients seek these:

a) ''automatic/hardware'' generation of SPI-CS - perhaps even with multiple CS so that we don't have to add hardware or ''bit-bang'' STM pins

b) improved I2C operation - both master & slave - especially a more ''graceful'' recovery should the process ''hang.'' (perhaps a reasonable, programmable time-out could achieve without sacrificing our normal timers)

c) we'd love 2-3 programmable ''strobe'' outputs (like /RD or /WR) for the non-FSMC devices. It is in-elegant to drop bit then raise it (or vice-versa) - as it now stands we must do this in a cpld to achieve strobe & other hardware functions. Believe someone else mentioned, too.

d) wish that CAN & USB were ''not'' self-annihilating.

e) welcome a detailed Ap-Note on ''double-buffering'' USB. (enables higher through-put) If the STM32 requires a ''tweak'' believe this is worthwhile.

f) AND - many would LOVE a semi-automated ''smart'' header file (or similar)which would ease, speed & enhance our use of Alternate or Remapped Function Pins. We all love the multi-function ''options'' - however with this flexibility comes (imho) ''undue & often unforgiving'' complexity. I would love to see the appropriate clocks being automatically enabled (as one example). Perhaps even something as simple as a spreadsheet Ap could achieve much of this process.

Hope that others find this of value - thanks to all who post and participate. We've learned much and there is a great spirit of friendliness & helpfulness here - best wishes to all...

st3
Associate II
Posted on May 17, 2011 at 12:58

Ethernet, and a TCP/IP stack to use wuth it; eg:

http://www.st.com/mcu/inchtml.php?fdir=pages&fnam=nichelite

relaxe
Associate II
Posted on May 17, 2011 at 12:58

@ Vince:

You should check out UM0427: Firmware LIbrary Reference at chapter 9: Flash memory.

You can actually FLASH_ProgramWord() data into unused part of your memory flash, then read them back. Since flash is non-volatile, you can have up to
16-32micros
Associate III
Posted on May 17, 2011 at 12:58

Dear all, jj & greg {especially}

All thanks from ST Marketing & Designers for these valuable wish-list,

I will be back to you with a complete list in one single message by the end of the week so we can analyze and share with you our comments.

Great Thanks for all !! again & again...

Cheers,

STone-32.

lil-vince
Associate II
Posted on May 17, 2011 at 12:58

Hi,

STOne, have you news about your list?

I have a wish: a small size (like 128/256 bytes) eeprom memory, I think it´s sufficient for many applications and when we need only 128/256 bytes (or less) to save somes data, we must use a 1kB or 2kB page (and it´s Flash memory, not eeprom).

(I´ve no idea about the cost or other things to implement it, but I think it may be good)

regards,

Vince

[ This message was edited by: lil-vince on 21-02-2009 09:50 ]

lil-vince
Associate II
Posted on May 17, 2011 at 12:58

Hi relaxe,

Thanks for your answer, but I´ve already a Flash save application which run (see my post

http://www.st.com/mcu/forums-cat-8026-23.html

, I´ve just a little question but nobody answer :-[ ) so, I know how to do.

Quote:

you can have up to 512K of ''better than eeprom''

Why is it better than eeprom? The number of erease cycle is only 10.000 cycles, so for somes applications (not for mine but...) it´s too small.

And especially the smaller part of memory we can erease is 1 page (1kB or 2 kB), so, when we have an application which need only a few word to be saved, we must erase an entire page... (For example my application recquire only 80 bytes to be write in flash and two separate place is better (2 differents parameters not save with the same command)(I was using MPS430 from TI and we can erase only 64 bytes segment, and we have access to an information memory of 256 bytes ''separate'' from main Flash) )

And the number of erase/write is smaller than somes eeprom...

Do you see what I mean?