Looking for unusual microcontrollers.

[Citizen]

Hello humans.
I am currently  doing some  hobby projects with FPGA and very  impressed  with it. Right now it  seems like any problem can  be solved with an FPGA
But i am also looking forward on learning how to use  an  advanced microcontroller in near future.

 
 I study Nanooptoelectronics ( Master of Science), we deal more with  semiconductor technology and  physics behind it. But i still think hardware design and programming skills are always a good thing to mention in your Curriculum Vitae. Moreover i like to program hardware!

Right now  i am looking for some interesting µC , nothing like Atmel or  PiC.  As i started to programm  my  atmega 2560, i got bored, couse every schooler can prog it:)
Something  unique, modern, interesting and  at the same time useful  for  my further industy career.
 The program language should be C (or like C) , no midage assembler please:)

What i found is:

 Cypress PsoC 5 :  FPGA like  designing process , but more analog oriented( you have ADC,DAC's, OPAMPS etc). And it has ARM M3. So it is like CPLD and ARm in one CHIP. But still  i read that PSOC is not that great as it seems, because  you will rarely  find a good design solution with its on-chip hardware. So PsoC arent really  as useful as it seems.

Analog Devices BlackFin: Has a cool name and logo:)))  But it is also a very  powerfull device. It seems to me it  is more for  Video/audio/touch processing. Not   really something for use in science:) And it doesnt have any ADC/DAC at all. Pretty weak. But maybe i am  just wrong, if so correct me:)

Analog Devices Precision Analog Microcontrollers :"....The ADuCM36x family extends the precision data acquisition systems with the first fully integrated, 4 kSPS, 24-bit data acquisition system incorporating dual, high performance multi-channel sigma-delta (?-?) analog-to-digital converters (ADCs), 32-bit ARM Cortex™-M3 processor, and Flash/EE memory on a single chip designed for direct interfacing to external precision sensors in both wired and battery powered applications."

http://www.analog.com/en/processors-dsp/analog-microcontrollers/products/index.html
The last device seems to me pretty interesting. I can imagine  using this device to measure/controll some   badass precise process in a lab:))

 I am sure there should be some other interesting devices i didnt find yet.
Any other suggestion?=)

[AndyC_772]

Something  unique, modern, interesting and  at the same time useful  for  my further industy career.

Something useful for your career isn't going to be "unique", it's going to be common. Nobody will hire you because you can program something obscure, unless they happen to need that particular skill - and, by definition, that's highly unlikely.

The parts most obviously missing from your list are the general purpose ARM Cortex microcontrollers from any of a wide range of vendors. The world + dog are using them because they're fast, cheap, low power, and available in a huge range of variants with different peripherals attached.

Forget an idea of learning something just because it's different or unusual, and focus on learning something that's appropriate for the task in hand, and for which there is demand. STM32 is probably as good a place to start as any.

[C]

I agree with andy,

You could build a board  that did _____.
If you were going in to production you might combine to of those boards on one board.
But for a lot of reasons, it would be great if the board you build was built to work in parallel with more of the same board.

Think of an scope for example, one board does a channel. Want a 16 channel scope run 16 boards in parallel. You might build a board to  take two of the channels in and output a combined channel. If you did not hit speed limits of the combiner, you could put some in parallel and some in serial.  You would be building more universal building blocks.

C   

[mikeselectricstuff]

Xmos.

[millerb]

Here's something unusual:

http://www.greenarraychips.com/home/products/index.html

[Citizen]

Thanks for replies.

XMOS looks really interesting:)

So does the Greenarray, but 450 Dollar ( i will pay 200 Euro max)  for a Dev Board is a bit too much for me:))


The unique factor is not only for industry career, but more for  pioneer experience:)

[MyCo]

http://www.xilinx.com/content/xilinx/en/products/silicon-devices/soc/zynq-7000.html

[millerb]

The MaxQ line is another weird microcontroller in that it essentially only has one instruction.

[Citizen]

Zynq from Xilinx is like a big Brother of Cypress PSOC. Very attraktive and  unseful. But cÄmon, about 1000 Euro for  Dev. Board...
Maybe i will still device for Cypress Psoc 5.
Anyone knows where is the diference between  PsoC 5 and 5 Lp.  Besides the LP is  a low power one. I cant find any Psoc5 Dev boards, only 5LP on cypress webpage.

[scarrier]

Not sure if this qualify as a mcu, but I know people that use gumstix for embeded project.

[Metalcrowe]

Maybe parallax propeller

[free_electron]

try to implement an F8 core , or a SC/MP system in the fpga. an SC/MP with 4 processor could be fun ...

[Hypernova]

Zynq from Xilinx is like a big Brother of Cypress PSOC. Very attraktive and  unseful. But cÄmon, about 1000 Euro for  Dev. Board...
Maybe i will still device for Cypress Psoc 5.
Anyone knows where is the diference between  PsoC 5 and 5 Lp.  Besides the LP is  a low power one. I cant find any Psoc5 Dev boards, only 5LP on cypress webpage.

PSoC5LP is PSoC5 with most of the silicon bugs ironed out, you'll find that they have better specs on their analog sub blocks and CAN bus. Basically what the PSoC5 was supposed to be but couldn't be ready in time. The CAN bus for example was in the 5 ES versions but couldn't pass certification.

[lorth]

Epiphany from Adapteva... A parallel processor...

Can't wait to get my Zynq + Epiphany by the end of the summer.....

[AndyC_772]

So, why might I ever choose to use one of these unusual processors instead of using, say, an ARM or PIC for a given job? Better price/performance ratio? Wider range of useful integrated peripherals? Availability of a really excellent, cost-effective tool chain? Ability to do something practical and worthwhile that a common CPU from a tier 1 manufacturer simply can't do?

[millerb]

So, why might I ever choose to use one of these unusual processors instead of using, say, an ARM or PIC for a given job? Better price/performance ratio? Wider range of useful integrated peripherals? Availability of a really excellent, cost-effective tool chain? Ability to do something practical and worthwhile that a common CPU from a tier 1 manufacturer simply can't do?

Everything has its strengths and weaknesses. The multicore microcontrollers have pretty sparse built in hw modules for things, but make up for it by implementing those features in software. Need three SPI interfaces and a UART? No problem, just add the library to your project. No need for complex interrupt schemes either, so you can often save a lot of development time.

The down side is that if some unusual micro doesn't sell well, it might end up being pulled from the market leaving you having to redesign your product. You have to spend valuable time learning something new which might not even leave the prototype stage. Also could have significant trouble finding folks to hire to maintain or do additional work on your product. Management types might think you are more interested in playing with toys instead of coming up with solutions if you propose to use something unusual in the requirements phase .

[nctnico]

Going small might be an option. Maybe try something ultra-low power like the MSP430 combined with power harvesting. One of my recent MSP430 designs exceeded the customer's spec for runtime from a battery by a factor of 3  O0

If you want to combine with an FPGA you can look into Xilinx' Picoblaze. You have to program it in assembly but it only has 5 instructions or so. You can create your own peripherals to do powerful stuff. I have used this for DSP applications. I put sine tables, a-law to linear conversions, multipliers, etc into peripherals and let the Picoblaze move data around and make decissions. I'm rather sad it has been a while since I had a real FPGA project.

[kiss_eng]

Actel/Microsemi SmartFusion - FPGA + Analog + uC

http://www.actel.com/products/smartfusion/default.aspx

[kiss_eng]

Going small might be an option. Maybe try something ultra-low power like the MSP430 combined with power harvesting. One of my recent MSP430 designs exceeded the customer's spec for runtime from a battery by a factor of 3  O0

Have you seen the MSP430 appnote on powering one off the energy harvested from a fiber optic line? They're pretty awesome little low power micros.

[Jeroen3]

Have you seen STM32F373 and STM32F4 series?
Or the NXP LPC43xx (dual core) ?
Or the complete opposite of those, Atmel ATTiny10.

There are many choices.

[Citizen]

Thank you very much for replies. I already have   choosen the Cypress PsoC

Already purchased this kit:

http://www.cypress.com/?rID=51577

Looks pretty interesting, but right now i am  looking some   tutorial videos  and trying to programm my own blinking LED;)

[lorth]

Thank you very much for replies. I already have   choosen the Cypress PsoC

Already purchased this kit:

http://www.cypress.com/?rID=51577

Looks pretty interesting, but right now i am  looking some   tutorial videos  and trying to programm my own blinking LED;)

Look at their app notes and extensive example tutorials that come with the Creator...

[daqq]

Unusual? Well, MAXQ is a one instruction computer, Atmel used to have until recently a 4 bit architecture.

Then there are mixed systems: on a single chip you get a DSP and an ARM (or different) processor ( http://www.ti.com/lsds/ti/dsp/overview.page ) or even several of each: http://www.ti.com/product/66ak2h12

I think there was even a device which had an ARM9 and an ARM7 on a single chip.

FPGAs often slap a core or two of PowerPC or ARM onto a chip.

Unusual enough for you? 

[westfw]

I don't suppose anyone has some tools for the Toshiba TMP47P443VN that they're not using any more?
I bought some via eBay on a lark, figuring "I can figure out a 4-bit micro; no problem!" (the chip is a 28pin SDIP 4bit micro with 4k of OTP program memoryl a member of the TMP47 family.)  As I looked into it in more detail, I discovered...  No programmers, no compiler, no assembler, not very much documentation  :-(

[amyk]

I don't suppose anyone has some tools for the Toshiba TMP47P443VN that they're not using any more?
I bought some via eBay on a lark, figuring "I can figure out a 4-bit micro; no problem!" (the chip is a 28pin SDIP 4bit micro with 4k of OTP program memoryl a member of the TMP47 family.)  As I looked into it in more detail, I discovered...  No programmers, no compiler, no assembler, not very much documentation  :-(

Here is an assembler I found for the TLCS-47 family.

Maybe try the Internet Archive for older versions of the Toshiba site as there was probably something there 16 years ago...