help choose ST Eval board -STM32072B- or -STM32091C- ?

[nquantum]

I'm choosing evaluation board for my beginning ARM world. I use to do thing with AVR , PIC 8bit before but ARM is my next new world.
My plan is to choose evaluation board coz it full set of thing I can learn hardware & software , well I'm more in hardware guy…
I plan to start M0 first then move on to the next M3 M4 …. so on , by Eclipse

Here is question : In ST evaluation board for M0 I found
-STM32072B-EVALEvaluation board with STM32F072VB MCU   $194.77
-STM32091C-EVALEvaluation board with STM32F091VC MCU   $194.77

Same price! Board & Function looks the same! Which one!?

For what I understand is :
-The STM32F0x1 line provides a high integration of functions and covers a wide range of memory sizes and packages, bringing flexibility to cost-sensitive applications.
-The STM32F0x2 line provides rich connectivity with crystal-less USB 2.0 and a CAN bus interface, making it the ideal choice for communication gateways, smart-energy devices or game terminals.

To be honest I don't know which to choose… F0x1 seem to more memory and without usb!? I don't sure , another is low memory with usb?
Can u guy help me some point in decision pls? Coz I want to make best use of it. (hobby)

Pls don't suggest me to go to Discovery or Nucleo coz that products I plan to play with them later... for now I love to try overview & learn full design of eval board kind of 1 board learn many thing , play many thing.
Unless there is strong point that I should these eval board for now...

[uncle_bob]

Hi

Both boards are Cortex M0 boards. Both have pretty limited flash and ram (for an ARM). You can get an M4 based board from ST (or others) with a lot more flash and ram on it. If you are developing for a product, the cost of an M4 vs an M0 is not super crazy. It is a lot easier to get started on a part that allows you to try out all of the cool tools (RTOS etc) than to start with one that dumps you into assembly language to squeeze everything into the part.

Basics:

72 has 128K flash, 16K ram, runs at 48 MHz. Has DMA, DAC, ADC
91 has 256K flash, 32K ram, runs at 48 MHz, Has more DMA, DAC, ADC.

You are getting twice as much of what you need (memory) on the 91 board. Of the two you have picked out, go for it.

Bob

[nquantum]

Thank for answer

I see your point about M4 have really more flash & sram. Seem 91 board is good way to go in these 2.
RTOS seem interesting to me too. Plan for learn FreeRTOS later also.

I also can see M4 board quite more advance than M0. Is it good idea to start familiar with M0 1st then just move on to M3 , M4?

[uncle_bob]

Hi

With the background you already have, start with the M4. You can learn the most "new stuff" on it. Going back to the M0 (if needed) will not be that much of a shock. It will be sort of like returning to the parts you already knew.

Things like FreeRTOS are cool. Learning it and at the same time trying to pack it as tight as you can ... not much fun at all. The packing part takes out stuff like debug features that are a big part of making learning easy.

------

You said not to mention the smaller boards .... I'll just say that money could be saved and still get you a very full featured M4 board.

-------

There are an ocean of people out there making ARM devices. They all have boards. They all have tool chains. Each and every one of them has absolute 100% proof that they are the best. If you doubt that, just visit each of their web sites. You will see "we are the best" somewhere in the first three sentences you read. If you are a corporate customer, you can have a few hundred sales guys through the lobby next week selling you different brands and flavors. There is nothing at all wrong with going with ST. If for some reason you get turned off by them, Freescale (now NXP) has a line of Kinetis parts that are worth looking at. Is this one or that one actually better? Good grief ....Can I construct a very artificial situation where this or that is best? I'd be a pretty poor engineer if I couldn't !

Have Fun !

Bob

[nquantum]

thank u!
Now I'm think toward to M4. I've check eval board for M4 quite wide range of them.
I narrow down to 7boards now these use cpu of:

Advanced lines
-STM32F469/479 – 180 MHz CPU/225 DMIPS, up to 2 Mbytes of dual-bank Flash with SDRAM and QSPI interface, Chrom-ART Accelerator™, LCD-TFT controller and MPI-DSI interface
-STM32F429/439 – 180 MHz CPU/225 DMIPS, up to 2 Mbytes of dual-bank Flash with SDRAM interface, Chrom-ART Accelerator™ and LCD-TFT controller

Foundation lines
-STM32F446 – 180 MHz/225 DMIPS, up to 512 Kbytes of Flash with dual Quad SPI and SDRAM interfaces
-STM32F407/417 – 168 MHz CPU/210 DMIPS, up to 1 Mbyteof Flash adding Ethernet MAC and camera interface

What is your suggest?
If u ask me what I want I want to jump into cortex-M ARM world , get use to it then maybe go to Linux ARM later… Again I'm hardware guy I'm more interest in design custom board in future.
So I want play around , some hobby project , robot , uav , LCD display etc.
Initial I want to jump to STM32F469/479 don't know it is too big step jump or not.
And I want board that I can experiment with for quite long run, like test-bench before into custom design.

STM3240G-EVALEvaluation board with STM32F407IG MCU
STM3241G-EVALEvaluation board with STM32F417IG MCU
STM32429I-EVALEvaluation board with STM32F429NI MCU
STM32439I-EVALEvaluation board with STM32F439NI MCU
STM32446E-EVALEvaluation board with STM32F446ZE MCU
STM32469I-EVALEvaluation board with STM32F469NI MCU
STM32479I-EVALEvaluation board with STM32F479NI MCU

[nquantum]

this is some info from ST web u might want to ref.
-Advanced lines:
STM32F469/479 – 180 MHz CPU/225 DMIPS, up to 2 Mbytes of dual-bank Flash with SDRAM and QSPI interface, Chrom-ART Accelerator™, LCD-TFT controller and MPI-DSI interface
STM32F429/439 – 180 MHz CPU/225 DMIPS, up to 2 Mbytes of dual-bank Flash with SDRAM interface, Chrom-ART Accelerator™ and LCD-TFT controller
STM32F427/437 – 180 MHz CPU/225 DMIPS, up to 2 Mbytes of dual-bank Flash with SDRAM interface, Chrom-ART Accelerator™, serial audio interface, more performance and lower static power consumption

Foundation lines:
STM32F446 – 180 MHz/225 DMIPS, up to 512 Kbytes of Flash with dual Quad SPI and SDRAM interfaces
STM32F407/417 – 168 MHz CPU/210 DMIPS, up to 1 Mbyteof Flash adding Ethernet MAC and camera interface
STM32F405/415 – 168 MHz CPU/210 DMIPS, up to 1 Mbyte of Flash with advanced connectivity and encryption

Access lines
STM32F411 – 100 MHz CPU/125 DMIPS, outstanding power efficiency with a new smart DMA optimizing power consumption for data batching (Dynamic Efficiency™ with Batch Acquisition Mode), high Flash memory and RAM density, USB, and enhanced peripheral set
STM32F410 - 100 MHz CPU/125 DMIPS, new milestone in outstanding power efficiency (89 µA/MHz and 6 µA in Stop mode) with a new smart DMA optimizing power consumption for data batching (Dynamic Efficiency™ with Batch Acquisition Mode), true random number generator, low-power timer, and DAC
STM32F401 – 84 MHz CPU/105 DMIPS, the smallest, cost-effective solution with outstanding power efficiency (Dynamic Efficiency™)

[dannyf]

What is your suggest?

It depends on what you want to do. A full-blown dev board has many interesting outboard devices on them so you save time wiring them into your experiment.

As to the choice of your mcu, they usually don't matter as long as they have a set of peripherals that you are likely to use and the core itself doesn't matter - assuming that you are going to code in a high level language like C.

Just so you know, code developed on one board / mcu may not be easily ported to a different board - it will depend on how you write your code. So if you ever think about retargetting, even in a one-off project, pick a board that has a corresponding discover or nucleo board.

[Kilrah]

Initial I want to jump to STM32F469/479 don't know it is too big step jump or not.

The "core" of controller is very similar in all cases, the difference is in the peripherals that can stay unused so it doesn't hurt or make it a "bigger step" to take a part with more features - it's not because they're in that you can't just ignore them until you're ready to tackle them.
I.e it won't take more effort to make a LED blink on an F479 than on an F0. The 479 will just have a megaton of other things that you'll be able to fiddle with when you're ready for it.

[uncle_bob]

Hi

Ok, A MCU with an M4, >= 1 meg flash (on chip), >= 64K ram (on chip), and >= 100 MHz clock puts you in a different environment. It does not quite get you to a comfortable Linux install. It does get you a lot of space for FreeRTOS and similar tools. The idea is to get exposure to those tools and to see what they can do. You are not quite into the "hardware does not matter" world of Linux with FreeRTOS. You are a lot further away than banging every single bit into each register on a PIC16. Moving code around is not trivial, but if you learn the right techniques, it's not a start from scratch thing either.

Stop and take a look at the chips these boards are based on. Can you get an "easy" 100 pin LQFP package or is the smallest part in a 484 pad 0.4 mm BGA? One can be used at home, the other ... forget it. Is the chip a $5 to $10 item in low quantities from a distributor who stocks them? If you are a hardware guy, trust me, pretty soon you will be making up your own board !!!

How much of your budget goes into this board? Consider that "stuff happens" and boards get fried. My boards seem to get a forest of wires soldered to them pretty quick. After a few cycles of wires on / wires off they aren't very pretty anymore. A board that spends the entire budget up front (as opposed to two or three cheaper boards) is not the way I would go.

What do I have sitting here? Well ... umm .... errr. .. a very large pile of Freescale boards. Most of them cost me less  than $40. The expensive ones were $100. I also have a couple of ST boards. . They are not the same sort of board you are looking at. One is a 32F746G DISCOVERY  board with an M7 on it. It is for a software defined radio project. The other is the earlier version of the same board for the same project. The new one is $50 and the older one was $30. For a project that did not involve porting code from one to another, I'd just get the newer one. The M7 board has a crazy BGA on it ... I will not be building one at home ...

Lots of choices. Plan on buying several rather than buying only one. The M4 is a good place to start. You can go up or down from there.

Bob

[nquantum]

That lot of info clear me more! thank guys!
Happy to see that core is just similar , I just start learn peripherals of it.

AND Hey! I forget about I could fried my big board! (I use to fried a lot IGBT when I do thesis!)
Even I love to have a big board. (Maybe sometime later) I'd rather begin (to fried) with small board 1st , the better. 

So I consider this one:
-32F429IDISCOVERY Discovery kit with STM32F429ZI MCU $29.9
-32F469IDISCOVERY Discovery kit with STM32F469NI MCU $60

1st or maybe the 2nd one.

[uncle_bob]

Hi

The more expensive kit has a much fancier display than the cheaper one. If you can use the display, that alone justifies the price delta. Most of what I do (other than the radio project) does not use a display. To me / for my stuff / my money the display is tempting but not a slam dunk decision. You also get a couple of mic's and a micro SD card slot on the higher end board.

The only obvious advantage to the less expensive board (other than half the price) is the LQFP package. That's great if you decide to do aboard of your own with the same part. It might be helpful if you need to get to a pin they "forgot" to bring out to a header.

I'd probably go for the $60 board.

Bob

[nquantum]

I can see more peripherals in the $60 one so I've ordered that already.

Actually I though of design custom board to learning ARM programming. Even choose some chip already haha. But don't know how some peripherals connect each other so I research some boards so that I post this. Soon I make some custom.. 

I've read in web: programming c/c++ bare metal also look for some book. ex.
-Mastering STM32 by Carmine Noviello
-The Definitive Guide to ARM Cortex-M3 and Cortex®-M4 Processors by Joseph You
Some useful recommend?

Further more! To be honest I've no idea how to program to drive TFT LCD. Even graphic character LCD (monochrome) I've but never try yet.
Only simple character LCD like 20x2 I've try.
For this kind of LCD to program I don't know where to start. Can u quid line me something?

[Kilrah]

You'll have all you need in the examples/demo code you can download for the board.

[uncle_bob]

You'll have all you need in the examples/demo code you can download for the board.

+1

Start with the demos and see how they do it. They will have the drivers, *and* show how you actually can get the drivers to do something useful. The documentation often just points you to the driver and is worthless for how to use it. The demo fills that very important gap.

[nquantum]

Thank you!
I'll play the stuff then. I'll head back when I can't figure it out on nxt topic.

[uncle_bob]

Thank you!
I'll play the stuff then. I'll head back when I can't figure it out on nxt topic.

Hi

Consider that if things branch off to a new topic, at least giving us a heads up when that happens. There is a *lot* of background in this thread that likely will  be relevant in that one as well.

Good luck !!!

Bob

[nquantum]

Consider that if things branch off to a new topic, at least giving us a heads up when that happens.

I'm not sure I'm understand right.
U mean:
1. Better continue in this topic since is cover background. To be not branch too much.
2. start new topic and put link to this topic coz it cover some background about ARM. So another people can reference to this for useful.

Which one?

[Kilrah]

2

[dannyf]

"Some useful recommend"

The data sheet and the library manual? The data sheet is really the most useful document for any programmer.

The You bok isn't that useful to a practioner. There is a PDF book from someone at Indiana university. I think it is on a M3 bjt very hands on.

[nquantum]

2

okay 

The data sheet and the library manual?

It's this book:
https://leanpub.com/mastering-stm32

There is a PDF book from someone at Indiana university. I think it is on a M3 bjt very hands on.

I think u refer to this one , look good&free! thank u!
http://www.cs.indiana.edu/~geobrown/book.pdf