Highly anticipated low-end MCUs

[dannyf]

What are the low-end MCUs that you look forward to playing with?

For me, it is STM32F030F, a 20pin CM0 chip. For its form factor, performance, versatility and low price.

It is my replacement for the equally low-end 16F684.

[dannyf]

Not to mention it's isp-capable.

[Laurynas]

http://www.cypress.com/?rid=92146

MCU is available for some time and i already used about a dozen of them, but this prototyping kit is great when i want something simple like arduino, but with "proper" IDE and chip

[HackedFridgeMagnet]

Just designed a board using the STM32F030F, finished and ordered yesterday. But only RS components seem to have the chips in stock now. Hopefully they will make more soon.

By ISP you mean the SWD on the STM32s?

[dannyf]

MCU is available for some time

Wow! That's really interesting.

[miceuz]

http://www.cypress.com/?rid=92146

wow, this looks interesting, ordered

[nuhamind2]

By ISP you mean the SWD on the STM32s?

I think he means the uart bootloader

[Psi]

I have a STM32F030  discovery kit here somewhere but ive not played with it yet.

Definitely seems like a good replacement for the larger AVRs which can get up to $10 each

[miguelvp]

http://www.cypress.com/?rid=92146

Nice!! 24MHz or 48MHz with programmable logic blocks? decisions decisions decisions, might as well get both! or many of each, but more of the 4200s.

Nah I'll start with one of each for now, Thank you VERY much!

[andyturk]

... i want something simple like arduino, but with "proper" IDE and chip

A "proper" IDE would actually let you write C++ code. :-(

[miguelvp]

... i want something simple like arduino, but with "proper" IDE and chip

A "proper" IDE would actually let you write C++ code. :-(

Not with 32kB of Flash RAM, 4kB of SRAM and 4kB of ROM

[miguelvp]

More pricey PSOC 4200 at $25 but with Arduino header and capacitance slider, leds etc
http://www.element14.com/community/docs/DOC-53992

Edit:
Features
PSoC® 4200 SoC
- ARM Cortex™-M0
- 48MHz CPU
- Supports an extremely low-leakage hibernate mode consuming only 150nA, and also 20nA best-in-class stop mode
Arduino® Shields and Pmod™ Compatible interface
CapSense® Slider with SmartSense™ Auto Tuning
RGB LED
User button and more!

[true]

Waiting for general availability and general pricing for STM32F042...

Also using STM32F030F4P6, nice little chip.

[Laurynas]

That pioneer board also has a PSoC5 along with PSoC4 (M3 core, much more regular stuff and programmable digital blocks).
I got a couple of those boards, but I'm more excited about small, cheap, basic dev boards

[Seg]

... i want something simple like arduino, but with "proper" IDE and chip

A "proper" IDE would actually let you write C++ code. :-(

Not with 32kB of Flash RAM, 4kB of SRAM and 4kB of ROM

By "C++" do you mean "STL"? Because STL on a micro is insanity.

[mikeselectricstuff]

Most interesting thing for me at the moment, more mid than low end,  is PIC32MZ - 512K RAM, hi-speed USB, QSPI and external memory interface, in packages down to a 64QFP.
 

[hans]

I've been looking at a PIC16F1455, as a $1 chip capable of USB without a crystal, in a 14-pin package.

But seeing all those interesting STM32F0 chips makes me think whether it's not more productive to change to ARM all together

The PIC32MZ is an interesting.. should be capable of running Linux.
Also, finally, they have got a chip with proper external memory integration. However, with 512kB integrated RAM I don't see when you will need that anytime soon
Not too impressed with PPS on PIC32 though. I tried to design in a PIC32MX470 on a project last month, but it felt more like a clutter than a feature. SPI SCK was not on PPS matrix, not every pin supports SPI SDO/SDI at the same time. Maybe I'm spoiled with the PIC24 PPS where it is like "oh I need timer/UART/SPI, I just throw it on any RPx pin and sort it out when I write firmware"

[dannyf]

I've been looking at a PIC16F1455, as a $1 chip capable of USB without a crystal, in a 14-pin package.

Yeah. Both 1455 and 1454 are quite interesting and competitive, particularly vs. older / other PIC16F chips.

PIC24 PPS

That remappable pin assignment implementation is absolutely the best, in my view - actually the whole pic24 family is very well done. No one has come close to Microchip in terms of PPS.

[dannyf]

A "proper" IDE would actually let you write C++ code. :-(

A "proper" IDE would actually let you get your job done. Period.

In this particular case, it actually allows you to write whatever code you wish to write, as long as you are able to set up your tool chain.

So if you cannot write C++ code on it, you have to blame yourself for that.

[gmb42]

http://www.cypress.com/?rid=92146

Does anyone know where I can get these in (or shipped at non-exorbitant cost to) the UK, Farnell don't have stock, Mouser have a 15 week lead time?  I can't see any shipping costs on the Cypress site.

[dannyf]

You will, once you go through the checkout process.

[Laurynas]

cypress online shop is best for chips, as they ship them for free and they keep lower costs than most distributors (i ordered 20 of these 4200 chips for 1$ each a couple of days ago).
for these prototype kits they have 15$ shipping so it's quite high for one device but not that exorbital when you're getting 5+.
for all other kits(like pioneer) they charge 25$ so it's usually more economical to get them from farnell or other distributors.

[poorchava]

I like PIC10F200: best replacement for 555 timer on dense boards.

[dannyf]

For that, I use 12F675 - you gain two extra pins.

[miguelvp]

cypress online shop is best for chips, as they ship them for free and they keep lower costs than most distributors (i ordered 20 of these 4200 chips for 1$ each a couple of days ago).
for these prototype kits they have 15$ shipping so it's quite high for one device but not that exorbital when you're getting 5+.
for all other kits(like pioneer) they charge 25$ so it's usually more economical to get them from farnell or other distributors.

I ordered one of each last night and shipping to Chicago was $7.50 not terrible at all. Only thing is that you have to register and provide a web page address.

[andyturk]

A "proper" IDE would actually let you write C++ code. :-(

A "proper" IDE would actually let you get your job done. Period.

In this particular case, it actually allows you to write whatever code you wish to write, as long as you are able to set up your tool chain.

So if you cannot write C++ code on it, you have to blame yourself for that.

What are you talking about? Have you ever tried to use Cypress' tools?

PSoC Creator does *not* support C++. I've gone several rounds with their tech support people on the issue. It's not a problem with the toolchain itself, which is based on GCC. It's a deficiency in their IDE which wants to be very aware of what's going on in the source. It simply fails to deal with C++. Going around PSoC Creator is difficult because the IDE generates low-level C code that wires up all your analog and digital blocks.

I was hoping to port a project from STM32 to PSoC5LP. It would have been perfect for Cypress' chip, but all my existing code was in C++. It was a choice of going with PSoC and re-writing a lot of code, or using a more conventional mcu with some external analog parts. The keeping the existing code was worth more than removing some op amps from the BOM.

[mikeselectricstuff]

Recently I've been heavily going back to PICs (8,16 and 32 bit) after a few detours into NXP ARMs in recent years.
The main reasons being, in no particular order :
* same programmer for all devices, and Pickit 3's standalone programmer-to-go mode is absolutely ideal for production programming, either on a jig or handheld - I can just give it to a subcontractor and tell them to connect it, press a button and check the light goes green.
* Same IDE for all devices, and very similar compilers
* Similar peripherals from 8 to 32 bit devices
* I can order preprogrammed parts from Microchipdirect for a few pennies and a couple of days extra.
* wide range pf packages from DIP to QFN
* Pin mapping, especially on PIC24 - pity this is a lot more limited on the PIC32MX
* Internal oscillators that are accurate enough to do UARTs - it has taken some manufacturers a while to catch up with this.
* always available.

 

[Corporate666]

http://www.cypress.com/?rid=92146

MCU is available for some time and i already used about a dozen of them, but this prototyping kit is great when i want something simple like arduino, but with "proper" IDE and chip

Wow, what a price point!  I ordered a handful of them - THANK YOU!

For anyone doing any serious development, you can also get touch with a Cypress FAE and they can get you samples of pretty much any of the chips you want - they are awesome.

[mariush]

Oh well, they charge 15$ shipping.  Shame, I would have ordered 2 boards.

[miguelvp]

Oh well, they charge 15$ shipping.  Shame, I would have ordered 2 boards.

They charge $7.50 here in the US an extra $7.50 to Europe is not that bad I think.

You can always buy the chips (PSoC 4200) at $1 free shipping worldwide with no minimum, your choice of 40-QFN 28-SSOP or 44-QFP

http://www.cypress.com/?id=2218

but then you have to program them, although it's well documented how to, it won't be as easy as with the $4 board.

http://www.cypress.com/?docID=48133

Mouser mentions shipping by 4/21 but their shipping to Europe might be about the same not sure, it's about $6 in the us.

[Laurynas]

shipping will be less of an issue when farnell and others gets them in stock.

As for programming chips in your devices the cheapest way i found so far is desoldering 3 resistors from a Pioneer kit and using either a fancy 10 pin cable or soldering 3 wires where those resistors were (my choice).

[dannyf]

The UDBs in those  chips are very interesting: kind of fpga-on-a-chip. Not sure how capable they are. Making a CRC or encryption engine out of them would be interesting.

Otherwise, the rest of the peripherals not as impressive. I do think their focus on rich (and interesting / unique) peripherals not the core itself is a good one. Most of the times, we don't need fast cores, just fast and rich peripherals.

My old Keil has those device in the database, the start-up file but no header files. Weird.

[dannyf]

Or you could buy one of those $4 kit and break off the target board and and solder four wires to your new chips.

I think they provide a header on the programmer / debugger side.

But no Keil support for the programmer - not sure about IAR. So for me J-link is the only way to go.

[Laurynas]

those 4$ kits are bootloader/serial based, not SWD, so they can't replace 25$ pioneer yet

not sure what you meant by keil/IAR, but you can export Creator projects:
http://www.cypress.com/?rID=92387&cache=0&source=creatorStartPage
and you can setup your custom toolchains as well

oh, and as for UDBs, CRC is one of standard UDB components

[electrodacus]

I use the STM32F050 in my Solar BMS.
Is 64KB Flash and 8KB RAM.
I use CooCox + gcc
My choice was mostly based on low power consumption and capacitive buttons.
Here is a 3D render of the solar BMS currently on Kickstarter

[poorchava]

Mouser mentions shipping by 4/21 but their shipping to Europe might be about the same not sure, it's about $6 in the us.

yaaa....

I use the STM32F050 in my Solar BMS.
Is 64KB Flash and 8KB RAM.
I use CooCox + gcc

I use the same combination. I like it alot.

[neslekkim]

Why not order from cypress direct? I ordered these now and paid only $15 in shipping, to Norway.

[mariush]

The board fits into a cardboard CD case, and then in a basic envelope.  This can be send anywhere in the world for 2-3$ (air mail priority whatever). 

It's just stupid to ask for 15 dollars to ship a board that can be shipped like a CD anywhere in the world.

I get it that they maybe have a contract and use bigger boxes to ship anything so they have a minimum shipping fee of 15$ (I can probably put 100 at quantity and get 100 boards with 15$ shipping)... I just don't feel like paying 15$ to ship a 4$ board so they lost me - I'd rather spend that 15$ to get 5-10 different pic chips from farnell with my next order. That's how you lose potential new developers for your chips.

[neslekkim]

well, $15 is better than €40... and I ordered 4 of them to compensate for the shippingcost
but yes, i agree, if you get someone in US to order, and ship, it will be much cheaper though.

Same with Atmel, I think it was $30 in shipping for the new programmer atmel-ice, so for now, I'm on hold on that one..

[gmb42]

If they get to Farnell and I can get my usual free shipping for orders over £20, then I'll get some.  I'll just have to be patient.

[granz]

Recently I've been heavily going back to PICs (8,16 and 32 bit) after a few detours into NXP ARMs in recent years.
The main reasons being, in no particular order :
* same programmer for all devices, and Pickit 3's standalone programmer-to-go mode is absolutely ideal for production programming, either on a jig or handheld - I can just give it to a subcontractor and tell them to connect it, press a button and check the light goes green.
* Same IDE for all devices, and very similar compilers
* Similar peripherals from 8 to 32 bit devices
* I can order preprogrammed parts from Microchipdirect for a few pennies and a couple of days extra.
* wide range pf packages from DIP to QFN
* Pin mapping, especially on PIC24 - pity this is a lot more limited on the PIC32MX
* Internal oscillators that are accurate enough to do UARTs - it has taken some manufacturers a while to catch up with this.
* always available.

I often find myself going back to PICs for those same reasons, after several detours.  The PIC24s are very handy because of PPS, and it is super quick to layout a cheap 2-layer board because routing becomes easier.  Internal oscillator for USB is also great.  USB+PPS+No crystal is a great combo.  Recently I used the PIC24FJxxGCxx series (not really low-end) and was quite happy with VBAT mode to keep the RTCC alive at something like 350nA.

I also am interested in the PIC32MZ, but it seems like they have been struggling with silicon bugs (last I looked at least).  Long-term availability is generally very important to me, and I've never had problems with PICs in that regard.

[dannyf]

PIC32MZ

I would put PIC32 in the same bag as AVR32 (the Atmel homebrew version), XMEGA and XMOS/Propeller etc. as things that are hopelessly dead before they are even born.

PIC24, on the other hand, is severely under marketed and under appreciated: Think what Microchip could have been had they invested a little more on that a few years earlier.

[westfw]

if you get someone in US to order, and ship, it will be much cheaper though.

I order 20 chips, a Pioneer board, and 4 of the "prototyping kits", direct from Cypress, to the US, and was charged $12 shipping (it doesn't look like it will get sent in one package, though.)

A "hobbyist retailer" may optimize their packaging and shipping for small orders, but it's rather unlikely that an electronics distributor or manufacturer would do so.  (that said, I was impressed by the XMOS "Startkit" mailer/packaging...)

[granz]

PIC32MZ

I would put PIC32 in the same bag as AVR32 (the Atmel homebrew version), XMEGA and XMOS/Propeller etc. as things that are hopelessly dead before they are even born.

PIC24, on the other hand, is severely under marketed and under appreciated: Think what Microchip could have been had they invested a little more on that a few years earlier.

Yes, I agree, I have concerns about the long-term life of the PIC32 as well (new parts + support).   It's too bad though, MIPS is quite a nice architecture.  I guess it remains to be seen.

I think I've used PIC24 varieties more then any other microcontroller.

[senso]

I might get one of those 4$(5€ at Mouser) in a collective order so we don't have to pay for shipping from Mouser.

[dannyf]

Microchip is a victim of its own (super) success: PICs have been a great story for them, so much so that they wanted to milk that cow a little bit too long. Before they realized it, technology turned against them.

A sad story.

[hli]

The UDBs in those  chips are very interesting: kind of fpga-on-a-chip. Not sure how capable they are. Making a CRC or encryption engine out of them would be interesting.

There is already a CRC component available, no need to do this by yourself.

Otherwise, the rest of the peripherals not as impressive.

Did you ever look at the list of available components? Just because a peripheral is not available in hardware directly, it doesn't mean its not available as UDB solution...

[dannyf]

Just because a peripheral is not available in hardware directly, it doesn't mean its not available as UDB solution...

How many of them can you have concurrently?

Just because you can create many peripherals doesn't mean you can have them all at the same time.

[Laurynas]

Just because a peripheral is not available in hardware directly, it doesn't mean its not available as UDB solution...

How many of them can you have concurrently?

Just because you can create many peripherals doesn't mean you can have them all at the same time.

This depends on a chip. They have PSoC4 with more UDBs on their roadmap and there is always an option to go PSoC5 for much more UDBs and analog stuff.

I'm not sure how often you need ALL peripherals at the same time, but I think you can reconfigure UDBs in software to use them for different purposes.

[dannyf]

Rarely I used ALL peripherals. But it is common for me to have a few SPIs, a couple UART, 4+ PWM channels, and a couple timers + 1-2 adc channels.

The incorporation of CPLD/FPGA into a mcu is certainly quite interesting. I was simply trying to get a sense as to how many of those peripherals you can create with 4 UDBs.

[granz]

Microchip is a victim of its own (super) success: PICs have been a great story for them, so much so that they wanted to milk that cow a little bit too long. Before they realized it, technology turned against them.

A sad story.

Well, I'm not sure it's that sad of a story.  I'm sure there are lots of things behind it, but Microchip stock has been on an overall rise over the last few years.  Compared to ST for example, they are doing much better.  For example, according to yahoo finance, last reported income for Microchip was 350M, where as it was -500M for ST, and -22M for Atmel.

[dannyf]

That's why looking forward is important. Going to a pic seminar and then a st or lpc ... Seminar and look around you. I went st conference in shanghai last year and saw a sea of young faces.

if I were running microchip, I would be very scared. But then, that's no my problem,

[Royce]

What are you talking about? Have you ever tried to use Cypress' tools?

PSoC Creator does *not* support C++. I've gone several rounds with their tech support people on the issue. It's not a problem with the toolchain itself, which is based on GCC. It's a deficiency in their IDE which wants to be very aware of what's going on in the source. It simply fails to deal with C++. Going around PSoC Creator is difficult because the IDE generates low-level C code that wires up all your analog and digital blocks.

I was hoping to port a project from STM32 to PSoC5LP. It would have been perfect for Cypress' chip, but all my existing code was in C++. It was a choice of going with PSoC and re-writing a lot of code, or using a more conventional mcu with some external analog parts. The keeping the existing code was worth more than removing some op amps from the BOM.

I can't vouch for it, my $4 kits are in the mail yet, and I've only just noticed PSoC4 as of yesterday on account of their $1/chip sale. So apologies if this is just noise. That said, I did run across this paper from a fellow that claims to have ported the Arduino (C++) libraries to Creator.

I will have a go at replicating his result once my boards arrive and I have a little bit of time.

It is disappointing that C++ doesn't work out of the box. I agree that C++ is handy and can make for cleaner clearer code. Even STL is not out of the question for small MCU's like this. Check this guy's work with STL and the AVR. You just have to be mindful of what the code is doing under the hood, but when do you not have to be mindful in resource constrained situations?

[hli]

Rarely I used ALL peripherals. But it is common for me to have a few SPIs, a couple UART, 4+ PWM channels, and a couple timers + 1-2 adc channels.

Getting _all_ of that in a PSoC4 will be a stretch. The 4245 has 2 SCB blocks (fixed function for SPI, I2C or UART), 4 TCPWM blocks (for PWM or as timers), two OpAmps, two comparators and one SAR ADC (with analog mux). Then you have 4 UDBs, each with 2 Macro-Cells and one data path. One UDB can do e.g. an 8-bit PWM (use two for 16 bit), or SPI, or a timer, or something completely different (there is an example doing square root in UDBs...).  Timers can be, sometimes, replaced by the very flexible clocking system (you can get basically multiple different frequencies down to single Hz or lower).
OTOH, PSoC3/5 come with up to 24 UDBs, dual SAR-ADC and DelSig-ADC, and SC/CT blocks for doing interesting anal,og stuff (like PGAs, TIAs or mixers)

[westfw]

My cypress order arrived today, and I need to comment on packaging optimization.  The "prototyping kits" have lovely packaging; it looks designed for handing out at trade shows (or MAYBE mailing; I don't know how that transparent plastic film window would hold up in the real mail.)  Not particularly scalable (mailing two isn't going to be cheaper than mailing one, probably), but a really well-done alternative to traditional samples, IMO.

They arrived in a FedEx padded envelope, together with the pioneer kit.  Presumably, this accounted for the $12 shipping.

The bare chips, OTOH...  OMG.  One full chip-tube-sized box containing a tube with the 10 SSOP chips.  A SECOND, tray-sized box, with tray, containing the 10 LQFP chips.  This was "free shipping" supposedly.  Ouch.

Here's a picture of the prototype modules.  The upper-left is the actual board, peeking through its transparent window...

[miguelvp]

Are full size chip tubes still one meter long?

Mine shipped on Friday and have been holding at FedEx since Monday, they had them all ready and sorted but it's not scheduled for delivery until tomorrow. Yeah, don't try to make me happy by delivering on Monday and make me wait until the 3rd business day instead!

On other news I got an Intel Galileo dev board today, I know poor GPIO performance because it's driven by I2C via a cypress chip (CY8C9540A-24PVXI), but I'm more interested in the mPCIe connector and the 400MHz Pentium class SoC Quark processor, ethernet and 256MB of DDR3 ram. ($14 per chip).

I might play with the duino stuff but I doubt it, I'm more looking into using linux for now to test some things for now.

[Q-Kernel]

My favorite MCU is the PIC24F128GC010. (128kb Flash / 8kb Ram) Most important it has PPS which simplifies the board layout and gives you the ability to fix board mistakes in software.  It has a lot of analog peripherals like 2 channel 16-bit delta-sigma ADC, 50 channel 12-bit pipelined ADC with 10M samples per second, 2 DAC's, 3 OPAMP's etc. It also has of course SPI's, UARTS's, I2C's, PWM's, IC's, LCD controller and USB on the go. Very low power (deep-sleep 75nA)

Very nice core architecture and the speed is very good. Compared to other processors in the EEMBC coremark benchmark the PIC24 get 1.88 coremark/MHz. Other small MCU's get:
- ATMega (8-bit) 0.54
- MSP430 (16-bit) 1.1
- PIC24F (16-bit) 1.88
- STM32F0xxx (32-bit) 2.2

I did several designs with the PIC24F128GA306 which is its predecessor without the sigma-delta ADC and USB on the go. Working now on a battery powered design where I use almost all the peripherals.

[Laurynas]

Rarely I used ALL peripherals. But it is common for me to have a few SPIs, a couple UART, 4+ PWM channels, and a couple timers + 1-2 adc channels.

The incorporation of CPLD/FPGA into a mcu is certainly quite interesting. I was simply trying to get a sense as to how many of those peripherals you can create with 4 UDBs.

Here is an example from a device of mine.
This device has 9 8-bit PWMs, 4 16-bit PWMs, I2C port. Report of used resources:

Resource Type                 : Used : Free :  Max :  % Used
============================================================
Digital clock dividers        :    1 :    3 :    4 :  25.00%
Pins                          :   36 :    0 :   36 : 100.00%
UDB Macrocells                :    9 :   23 :   32 :  28.13%
UDB Unique Pterms             :   15 :   49 :   64 :  23.44%
UDB Total Pterms              :   21 :      :      :
UDB Datapath Cells            :    3 :    1 :    4 :  75.00%
UDB Status Cells              :    0 :    4 :    4 :   0.00%
UDB Control Cells             :    0 :    4 :    4 :   0.00%
Interrupts                    :    1 :   31 :   32 :   3.13%
Comparator/Opamp Fixed Blocks :    0 :    2 :    2 :   0.00%
SAR Fixed Blocks              :    0 :    1 :    1 :   0.00%
CSD Fixed Blocks              :    0 :    1 :    1 :   0.00%
8-bit CapSense IDACs          :    0 :    1 :    1 :   0.00%
7-bit CapSense IDACs          :    0 :    1 :    1 :   0.00%
Temperature Sensor            :    0 :    1 :    1 :   0.00%
Low Power Comparator          :    0 :    2 :    2 :   0.00%
TCPWM Blocks                  :    4 :    0 :    4 : 100.00%
Serial Communication Blocks   :    1 :    1 :    2 :  50.00%
Segment LCD Blocks            :    0 :    1 :    1 :   0.00%


Although there are a lot of UDB macrocells left, but since 3/4 datapath cells are used, it'd say 3/4 of logic resources is used.
Opamps, ADC, CSD, IDACs, Comparators, LCD, and a communication block(I2c/serial/SPI...) are still free.

Note that sometimes there are more efficient user components than the ones that are in Creator by default. In this case those 9 8-bit PWMs are built from 3 UDB blocks while 4 16-bit PWMs are using dedicated TCPWM modules.

So in this case i needed a lot of PWMs and got them by using dedicated blocks + UDBs. in annother device those UDBs could be used for additional serial/i2c stuff or those dedicated TCPWM blocks could be used as counters.

Not sure if this is a good enough example to show PSoC flexibility, but i just started with PSoC and don't have anything more interesting.

[dannyf]

Most important it has PPS

Another fan of PIC24's PPS,

I also liked the dedicated timers for OC and PWM: I have been using them as general timers and loved that.

Microchip could have done so much for PIC24 yet they did so little.

[granz]

My favorite MCU is the PIC24F128GC010. (128kb Flash / 8kb Ram) ...

That is an excellent chip, I used that exact one recently, and I'm very happy with it in a battery-powered data-logger application. 

On another note: I noticed a while back that some PICs have configurable logic cells also.  It looks like they started adding them in 2011 to some parts.  Does anyone have any experience with these?  It looks like they are mostly intended to interconnect peripherals, rather then be used to add extra peripherals.

http://www.microchip.com/pagehandler/en-us/press-release/microchip-launches-8-bit-mcus-with-configurable-lo.html

[Q-Kernel]

They only have that in the 8-bit pic's. The functionality is great but I find that the 8-bitters are very limited. It would be nice if they put those logic cells into the PIC24/dsPic and PIC32 line.

On another note: From all PIC's (8, 16 and 32-bit) I prefer the 16-bit line (PIC24/dsPIC) because they have the most versatile peripherals and work in a wide range (16-70) MIPS.
The 8-bit are in a lot of cases not powerful enough for my projects while the PIC32MX doesn't give that much more performance and their PPS is not that good.
I always use PIC24/dsPIC if I need less then 70 MIPS,  512kb Flash and 32kB of RAM. That is more or less what you get from the smaller ARM M3.

Above that I use ARM M4. Haven't tried the PIC32MZ yet because of the silicon bugs but it is faster than most ARM M4's

[miguelvp]

I got my PSoC® 4100 & PSoC® 4200 Protoyping Kits Other than I need a good long USB Extender cable an a USB HUB to be able to access it better.
Only tried the PSoC 4100 for now but after encountering this link, I wished I got the Pioneer Kit along with both.

@westfw, check this out if you didn't find it already:

http://www.element14.com/community/thread/23736/l/100-projects-in-100-days

For those that don't have the Pioneer kit, you can always adapt the code for your needs.

Edit: Also you can program the prototyping kits with Android? didn't try, but I could use my transformer prime with keyboard and dock to program these things? No IDE for development just to configure them via the USB UART, the SDK only takes care of been able to communicate with the USB-Serial Bridge Controller.

[Corporate666]

I bought 5 of the PSoC4 4200 kits - I agree with the previous poster, very cool packaging, super light weight and BS-free.  I haven't had much time to play with them yet but what a fantastic concept by Cypress... hell, the whole kit is barely above the cost of the chip + what a small PCB would cost... and it comes with USB connectivity that you can just program and snap it off afterwards.  Wonderful!

I also got the PSoC Pioneer kit before and I highly recommend it.  I think it's only about $15-20 and I use it for fleshing out tons of ideas before committing resources to them - definitely worthwhile.

Counter-intuitive as it sounds, the PSoC3 is actually above the PSoC4 in power/perpipherals, and the PSoC5 is the top of the line... so if anyone gets hooked on the PSoC concept but needs more breathing room, PSoC3's are cheap and can have tons of peripherals instantiated in them.  I have one in a commercial product running 26 PWM's, an ADC, lots of interrupts and lots more - all in a chip that cost about $3.

[Zeta]

My cypress order arrived today,

They arrived in a FedEx padded envelope, together with the pioneer kit.  Presumably, this accounted for the $12 shipping.

Ouch this is what I was afraid of. $15 shipping is ok for me but Fedex customs agent fees will probably be over $50. I wish they offered USPS shipping.

[dannyf]

This device has 9 8-bit PWMs, 4 16-bit PWMs, I2C port. Report of used resources:

That's a pretty big device.

Having read the datasheet the psoc4 datasheet, their capabilities are quite limited, given 4 udb. If I am not mistaken, out of the 4 udb, you can make 4 8-bit timers / pwm, OR 2 16-bit timers / pwm, or 1 32-bit timer / pwm.

Still quite useful but seems those devices are geared to be more like mcus than the psoc3/5 devices.

Still impressive. Just not as impressive as they are made out to be.

[miguelvp]

Ouch this is what I was afraid of. $15 shipping is ok for me but Fedex customs agent fees will probably be over $50. I wish they offered USPS shipping.

If you purchase 2 $4 boards they'll charge you $50 taxes? I don't get it.
I did pay $7.50 shipping (no pioneer kit) and came from Canada  I guess that warehouse is closer to Chicago.

Maybe you can contact the closer sales office from you and inquire.
http://www.cypress.com/?id=1062

I did end up getting the Pioneer from element 14 and they offer DHL shipping, I figured since they too the effort to drive the 100 projects in 100 days, I give them my business for that kit instead of ordering directly from cypress.

[vvanders]

Just got my prototyping kits and pioneer, pretty impressed so far(I've been using the IDE while waiting for the hardware). I'll echo what other people are saying about the packaging, pretty slick and the "PSoC rocsk!" sticker behind the board is cute.

Plan on putting some more time into these a bit later(including the PSoC5 on the pioneer at some point).

[Corporate666]

This device has 9 8-bit PWMs, 4 16-bit PWMs, I2C port. Report of used resources:

That's a pretty big device.

Having read the datasheet the psoc4 datasheet, their capabilities are quite limited, given 4 udb. If I am not mistaken, out of the 4 udb, you can make 4 8-bit timers / pwm, OR 2 16-bit timers / pwm, or 1 32-bit timer / pwm.

Still quite useful but seems those devices are geared to be more like mcus than the psoc3/5 devices.

Still impressive. Just not as impressive as they are made out to be.

The PSoC4 also has four 16-bit "timer/counter/PWM" blocks, so you wouldn't likely be using the UDB's as timers/PWM's.  There are also on-chip op-amps and comparators, as well as a couple of communication blocks (usable for UART or I2C).  There's furthermore on-chip LCD drive and capacitive touch sensing, which greatly reduces the amount of things you would need to use the UDB's for. 

[Zeta]

Ouch this is what I was afraid of. $15 shipping is ok for me but Fedex customs agent fees will probably be over $50. I wish they offered USPS shipping.

If you purchase 2 $4 boards they'll charge you $50 taxes? I don't get it.
I did pay $7.50 shipping (no pioneer kit) and came from Canada  I guess that warehouse is closer to Chicago.

It's not taxes (Actually I don't pay import taxes), but rather Fedex fee for storage, opening the box and filling the customs form. Sounds crazy but thats how things work with Fedex here. :/

[Corporate666]

Ouch this is what I was afraid of. $15 shipping is ok for me but Fedex customs agent fees will probably be over $50. I wish they offered USPS shipping.

If you purchase 2 $4 boards they'll charge you $50 taxes? I don't get it.
I did pay $7.50 shipping (no pioneer kit) and came from Canada  I guess that warehouse is closer to Chicago.

It's not taxes (Actually I don't pay import taxes), but rather Fedex fee for storage, opening the box and filling the customs form. Sounds crazy but thats how things work with Fedex here. :/

What country are you in?  If you want one of these PSoC kits, I don't mind sending you one via USPS if you pay shipping plus the cost of the kit. 

FWIW, Cypress just sends the orders over to distributors - most of my samples and orders from the Cypress website are actually shipped from Digikey.  It doesn't even say Digikey on the package, but something like "Cypress fulfillment", but then is in Thief River Falls, MN - which is Digikey.  My point being, these kids should be in stock soon with Digikey and other distributors, so another option would be to hang in there and get it from one of those guys (who will ship to you via USPS) when stock shows up.

[Laurynas]

This device has 9 8-bit PWMs, 4 16-bit PWMs, I2C port. Report of used resources:

That's a pretty big device.

Having read the datasheet the psoc4 datasheet, their capabilities are quite limited, given 4 udb. If I am not mistaken, out of the 4 udb, you can make 4 8-bit timers / pwm, OR 2 16-bit timers / pwm, or 1 32-bit timer / pwm.

Still quite useful but seems those devices are geared to be more like mcus than the psoc3/5 devices.

Still impressive. Just not as impressive as they are made out to be.

Think of UDBs as a Joker card. You have opamps/PWMs/Counters/Comparators and some other stuff as standalone components, and UDBs are there to fill in what's missing for your particular project.
And you don't have to use Cypress UDB components - I'm using a community version PWM component that turns 1 UDB into 3 PWM outputs.

Anyway, this is a lowend MCU, PSoC5 is much more impressive but it's no longer a 1-2$ chip.

[westfw]

The range of "programmable logic" available in microcontrollers is getting interesting.  Low-end PICs have a small block or two as part of their "CLC" peripherals.  PSoC 3 anf 5 (and the ill-fated Atmel FPSLIC) have quite a lot of configurable logic, and I guess the PSoC 4 is in-between.  All sorts of attempts to find a "sweet spot" of usefulness.

I attended a Cypress/Vendor seminar back when the PSoC-5 was relatively new, and one of the other attendees pointed out that sharing the software development tool with the hardware configuration tool (single tool, single set of design files, etc) BREAKS the traditional development model where the hardware engineers and the software engineers barely talk to each other.  An interesting observation; perhaps not applicable to the sort of do-it-all embedded microcontroller people, but more true of those large corporate entities that buy large quantities of chips.  I guess we'll see what happens.

PSoC-1 has the Cypress proprietary core, that I haven't heard anything good about.
PSoC-3 has an (advanced, single-cycle) 8051 core.
PSoC-4 has an ARM CM0
PSoc-5 has an ARM CM3

(Note that the Pioneer board is still PSoC-4 based.  There's an interesting looking "FreeSoC" board that has a PSoC-5: http://freesoc.myshopify.com/products/freesoc-development-kit - Open Source and all that.  It's a bit beyond the "impulse buy" price point, though...)

[Laurynas]

(Note that the Pioneer board is still PSoC-4 based.  There's an interesting looking "FreeSoC" board that has a PSoC-5: http://freesoc.myshopify.com/products/freesoc-development-kit - Open Source and all that.  It's a bit beyond the "impulse buy" price point, though...)

Pioneer has PSoC4 as it's main chip, and PSoC5 as it's programmer, however PSoC5 can be programmed and has some headers for easy access. There are demo projects using PSoC5 on Pioneer board in element14 forums, but i haven't tried it myself.

[westfw]

Pioneer has PSoC4 as it's main chip, and PSoC5 as it's programmer, however PSoC5 can be programmed

Cool.  I missed that!

[miguelvp]

Pioneer has PSoC4 as it's main chip, and PSoC5 as it's programmer, however PSoC5 can be programmed

Cool.  I missed that!

It's part of the 100 projects in 100 days I linked above:
http://www.element14.com/community/thread/25084/l/psoc-4-pioneer-kit-community-project036-what-i-can-use-the-psoc-5lp-too

[Royce]

I got my ProtoKits (2x42 1x41) yesterday and was able to tink with them a little after my wife and child went to bed. I did have to read and execute the ProtoKit guide's bootloading section, linked at the bottom of the kit product page, carefully to get the thing to work. I got weird results before that.

Anyway, it is an interesting mixture. I haven't worked with UDB blocks before. I'm a little hesitant about all the generated code, but at least you can see it all. The links to the datasheets for the modules are pretty handy too. And apparently you can even make your own UDBs.

I'm sure it will take some getting used to. The first thing I tried, putting a clock straight out on a pin, was a no go, for example. It has to go thru something to get to a pin. But I did get a compile error and the message was enough to clue me in to the fact that I couldn't push the clock directly to a pin even though the schematic drawing thing would like me hook that up.

I definitely think this chip will be worth the time investment to learn. (Time is one of my more valuable assets these days)

[hli]

I'm sure it will take some getting used to. The first thing I tried, putting a clock straight out on a pin, was a no go, for example. It has to go thru something to get to a pin. But I did get a compile error and the message was enough to clue me in to the fact that I couldn't push the clock directly to a pin even though the schematic drawing thing would like me hook that up.

That used to work with PSoC3/5, but with PSoC4 there were some changes in the clocking system. Fortunately, there is a nice explanation in the PSoC Sensei Blog. (The follow-ups to that explain the fractional clock dividers, and the chained clock divider system, which give you 12 different clocks if needed).

I definitely think this chip will be worth the time investment to learn. (Time is one of my more valuable assets these days)

I think it is. Whenever I am about to prototype something in hardware, I take out a PSoC5 board. Its so much easier when you just draw a schematic, configure some components and are ready. I once did a 2-channel-scope+8-bit-logic-analyzer just in a single chip, together with the whole triggering stuff, PGAs and so.

[vvanders]

That was a big part in my buying the pioneer. You still have to solder on headers and get an external programmer but its the highest end psoc5 as well.

spent some more time with the idea last night, love the mixed schematic\source editing. Very C#ish which is a large plus in my book. Had logic driven pwm up and working in less than 5 minutes. I only wish the psoc4 had all the analog goodies that psoc5 has(PGA, vdac, etc).

[vvanders]

You know, now that I think about it the Prototype boards might make a nice platform for a cheap hobby logic analyzer/oscope. I think I remember reading the ADC on the 42xx can do ~60kHz and the UDB can run logic at 48Mhz. Shame there's no DMA but I might see how quickly the CPU can offload data from UDB to SRAM. Even 24Mhz would be enough to capture up to a 10Mhz SPI bus.

[miguelvp]

You know, now that I think about it the Prototype boards might make a nice platform for a cheap hobby logic analyzer/oscope. I think I remember reading the ADC on the 42xx can do ~60kHz and the UDB can run logic at 48Mhz. Shame there's no DMA but I might see how quickly the CPU can offload data from UDB to SRAM. Even 24Mhz would be enough to capture up to a 10Mhz SPI bus.

From: http://www.cypress.com/?docID=46322

12-bit SAR ADC
The 12-bit 1 MSample/second SAR ADC can operate at a
maximum clock rate of 18 MHz and requires a minimum of 18
clocks at that frequency to do a 12-bit conversion.

So 8 datapoints for the oscilloscope, gives you 200KHz for the scope part of the device
4 datapoints 400KHz

Couple of pioneer (PSoC 4200 based) projects:
http://www.element14.com/community/thread/24287/l/psoc-4-pioneer-kit-community-project17--2-channel-oscope-with-graphicslcd
http://www.element14.com/community/thread/24101/l/psoc-4-pioneer-kit-community-project012-pioneer-board-oscilloscope

Main link to the 100 projects
http://www.element14.com/community/thread/23736/l/100-projects-in-100-days
Those can be adapted to the protoboard fairly easy (other than the one that uses the PSoC5 chip)

[tjaeger]

STM32L0x2 looks awesome:  Cortex M0+, Super-low power, EEPROM, crystall-less USB, 1MSPS ADC/DAC, LCD.  Hopefully cheaper than the STM32L1.

http://www.st.com/web/en/cn/catalog/mmc/SC1169/SS1817

Check the documentation now ! The devices will be on the shelves within April 2014 !

I find this very hard to believe...

[Kjelt]

Given that the STM32L0 start at 32kB they will probably not be super cheap.

[dannyf]

ST track record in meeting their delivery deadline isn't that great. The 32F030 chips are supposed to be widely available early 1Q14, for example. But they are nowhere to be found.

[HackedFridgeMagnet]

ST track record in meeting their delivery deadline isn't that great. The 32F030 chips are supposed to be widely available early 1Q14, for example. But they are nowhere to be found.

Seems like mouser, Digikey, RS, and others have some.

but the 20 pin TSSOP versions are very scarce.
I want a few now, element 14 had 1500 recently now they have 0. Should've been quicker.

[Crazy Ape]

ST track record in meeting their delivery deadline isn't that great. The 32F030 chips are supposed to be widely available early 1Q14, for example. But they are nowhere to be found.

Seems like mouser, Digikey, RS, and others have some.

but the 20 pin TSSOP versions are very scarce.
I want a few now, element 14 had 1500 recently now they have 0. Should've been quicker.

Seems to be a few of the 20 pin TSSOP versions kicking about on Aliexpress, just grabbed 5 of them to have a play with.
http://www.aliexpress.com/wholesale?SearchText=STM32F030F4P6

[HackedFridgeMagnet]

Thanks for the tip Crazy ape.

Got some too.

Never bought chips of Ali before, but had good experiences with my other purchases.

[Crazy Ape]

Thanks for the tip Crazy ape.

Got some too.

Never bought chips of Ali before, but had good experiences with my other purchases.

You hear of fake chips floating about on Chinese sites from time to time but I'm yet to come across any personally, though I do stick to sellers that have feedback and a number of previous sales, seems to be fairly safe. My guess is that most of them are excess from production runs and likely already been paid for once, so prices are often very good.

You'll find crazily cheap prices on some of the more expensive chips from time to time, great for hobbyist/prototyping related purchases.
Example (I personally bought and used these):
$11 - Cy7c1061av33-10zxc  (16Mbit 10ns SRAM)
$5   - Adv7123jstz330 (Triple 330MHZ Video DAC)

[kert]

I have a hobby of scanning Octopart and Digikey for the cheapest ARM cores now and then. STM32F030 and Freescale Kinetis KE04 line are in my "absolutely have to play with one of these as soon as time allows" list. Kinetis KL03 and KL02 plus LPC1102 "chip scale package" or UFBGA ones are there too, just because they are so tiny, so i want to see if i can  fit them in very small servos and such.

[westfw]

The chip scale packages are depressing.  Not only do I think I can't solder them, I think I can't get a PCB made at reasonable cost.  It looks like things should be 4/4/4mil and six layers (although perhaps the smaller 0.4mm WSCP BGA packages could get away with only 4 layer?)

[neslekkim]

If it is only for prototyping etc, you can use proto-advantage.com, find the appropriate board, something like this:
http://www.proto-advantage.com/store/product_info.php?products_id=4000011
enter the digikey number for the part, and you get pcb with the part ready soldered..

[Crazy Ape]

If it is only for prototyping etc, you can use proto-advantage.com, find the appropriate board, something like this:
http://www.proto-advantage.com/store/product_info.php?products_id=4000011
enter the digikey number for the part, and you get pcb with the part ready soldered..

I don't think proto-advantage have a suitable board just yet, at least I couldn't see any.
There are folks playing with these tiny chips though, here are some random (though interesting) LPC1102 links from around the internet.

So small it's scary!
http://static.electro-tech-online.com/customimages/2011/07/1701_hi-1.jpg

DIP adapter layout (via's directly under center four balls).
http://www.electro-tech-online.com/attachments/jason-01-gif.55164/

A D.I.Y. etched single sided board.
http://320volt.com/en/lpc1102-arm-icin-cevirici-karti-header-board-pcb/

An interesting (hack) method from Russia (re-route the die interconnects of a different chip)
http://habrahabr.ru/post/218233/
http://beta.hstor.org/getpro/habr/post_images/17f/3ab/75d/17f3ab75d5f9169808049a7c59fdcf6b.jpg

[neslekkim]

I can guess that it's also by demand ?, i got some qfn10 and qfn20 packages soldered like that, way to tiny for my solderingskillz..

[kert]

The chip scale packages are depressing.  Not only do I think I can't solder them, I think I can't get a PCB made at reasonable cost.  It looks like things should be 4/4/4mil and six layers (although perhaps the smaller 0.4mm WSCP BGA packages could get away with only 4 layer?)

Agreed, CSPs and things like LGA pcb modules ( some newer wireless modules do that ) can be pretty depressing, but still an interesting challenge.
I was searching for pretty much any prototyping solution for these tiny bumped WLCSP modules and didnt actually find any. I would think that something along the SchmartBoardEZ BGA soldering approach could be made to work, though.

If it is only for prototyping etc, you can use proto-advantage.com, find the appropriate board, something like this:

Thanks for the protoadvantage tip, this will become handy.

[kert]

I don't think proto-advantage have a suitable board just yet, at least I couldn't see any.
There are folks playing with these tiny chips though, here are some random (though interesting) LPC1102 links from around the internet.

I sent them a request through their "give me a part number, we'll suggest a board" link and they recommended this
http://www.proto-advantage.com/store/product_info.php?products_id=4000011

They'll just place the part in the corner. I ordered both MKL02Z32CAF4R and LPC1102UK,118 populated, will see how it works.

[Crazy Ape]

I don't think proto-advantage have a suitable board just yet, at least I couldn't see any.
There are folks playing with these tiny chips though, here are some random (though interesting) LPC1102 links from around the internet.

I sent them a request through their "give me a part number, we'll suggest a board" link and they recommended this
http://www.proto-advantage.com/store/product_info.php?products_id=4000011

They'll just place the part in the corner. I ordered both MKL02Z32CAF4R and LPC1102UK,118 populated, will see how it works.

The LPC1102UK is 0.5mm pitch so should work, but the MKL02Z32CAF4R is a 0.4mm pitch part.

[paulie]

I sent them a request through their "give me a part number, we'll suggest a board" link and they recommended this
http://www.proto-advantage.com/store/product_info.php?products_id=4000011

$21 plus $5 sh. lol.

[neslekkim]

$21 plus $5 sh. lol.

Do it better yourself then.. without equipment it's cheap and easy, but ofcourse, if you find soldering this stuff easy, why not?

[paulie]

i have done it myself, and its VERY easy. 2 minutes with a $3 reball kit and embossing heat gun. i just find it amusing that anyone would consider paying that to play with a $1 chip. specially when a short session on eagle and 1 week turnaround with osh would get a few made for about 1/20th that cost. of course places like that will always take advantage of those with money to burn and little sense of value.

[westfw]

a short session on eagle and 1 week turnaround with osh would get a few made for about 1/20th that cost.

If you actually try that, let us know how it goes.  I got as far as the layout, and found that I couldn't come CLOSE to the published oshpark design rules:

The minimum specs for 2 layer orders are 6 mil traces with 6 mil spacing, and 13 mil drills with 7 mil annular rings.
The minimum specs for 4 layer orders are 5 mil traces with 5 mil spacing, and 10 mil drills with 4 mil annular rings.

The trace/spacing doesn't allow for running tracks between balls, and the drill size doesn't allow for a multi-layer breakout, even with vias in the pads.  It's not just a little beyond the design rules, either: it's about 40% (I can get it to work with 3/3mil width/clearance.)  Oh, it MIGHT work.  But that's not the way to design a PCB.

And that was for the NXP (.5mm BGA) part.  The freescale part has 0.4mm ball spacing.

And no, I'm not considering "paying $20 for a PCB to use a $1 chip"; I'm just finding those chips unusable.  (There are presumably PCB companies than can do 3/3mil boards.  But none that I know of in the "pay by the square inch" or "<$20 for 5 boards" category.  So I'd be looking at significantly MORE than $20 to put the chips on a home-designed PCB as well.)

[poorchava]

TI recommends 3.1mil track/gap for 0.5mm pitch WCSPs, via size 0.254mm, sole size 0.127mm. I would call that deeeeep in hardcore smarphone-level HDI field.

[Corporate666]

i have done it myself, and its VERY easy. 2 minutes with a $3 reball kit and embossing heat gun. i just find it amusing that anyone would consider paying that to play with a $1 chip. specially when a short session on eagle and 1 week turnaround with osh would get a few made for about 1/20th that cost. of course places like that will always take advantage of those with money to burn and little sense of value.

When it comes to value, you have to consider the big picture.  If I was paying an EE to develop a device, and he proudly told me he spent just a half hour and whipped up his own BGA/DIP adapter and it would be here next week, I'd be pretty upset.  $20 for something that you can buy off the shelf, know it works, and get it right away is really a very small price to pay for anyone but a pure hobbyist, but I doubt most hobbyists are using fine pitch BGA's anyway.

I do wish the manufacturers would make "proto packages" though - I can understand not wanting to manufacture DIP/QFP or such for hobbyists, but it would be nice if they had some BGA's reflowed into DIP adapter boards in stock and ready to go for the hobbyist.  After all, even quite complex designs have engineers working on them in their office/lab before BGA production chips are ever used in pick/placed volume production.

[tjaeger]

When it comes to value, you have to consider the big picture.  If I was paying an EE to develop a device, and he proudly told me he spent just a half hour and whipped up his own BGA/DIP adapter and it would be here next week, I'd be pretty upset.  $20 for something that you can buy off the shelf, know it works (emphasis mine), and get it right away is really a very small price to pay for anyone but a pure hobbyist, but I doubt most hobbyists are using fine pitch BGA's anyway.

But how do you know it's going to work without proper decoupling?

[dannyf]

But how do you know it's going to work without proper decoupling?

We know for sure that it's not going to work without proper XYZ - XYZ being anything you care to put there.

[kert]

If I was paying an EE to develop a device, and he proudly told me he spent just a half hour and whipped up his own BGA/DIP adapter and it would be here next week, I'd be pretty upset.  ...

Agreed. In general, i have noticed that for most prototyping tools if it costs less than 50 bucks or so, my own time spent recreating them will be more expensive - especially if its just a one off curiosity investigation.
An ATmega328P in a DIP socket for $25 is LOL expensive too, nevertheless i have a bunch of Arduinos ( and a myriad of other quick proto kits ) around here.

[tjaeger]

But how do you know it's going to work without proper decoupling?

We know for sure that it's not going to work without proper XYZ - XYZ being anything you care to put there.

The point being that it's going to be tough to implement proper decoupling on a breakout board -- there's no guarantee that the traces from the power pins are short or even stay close together (although you might get lucky here because there are 16 ways to place the chip, but then again, you also have separate digital and analog supply, so good luck).

[Kjelt]

When using these boards I always glue some smd caps on the chip and connect them with very thin teflon wire to decouple the chips.

[tjaeger]

When using these boards I always glue some smd caps on the chip and connect them with very thin teflon wire to decouple the chips.

I still don't see how this would work with a BGA chip.  You'd have to scrape off the solder mask or something.

[Kjelt]

Ah sorry for BGA indeed you have to clean pcb traces where available and hope your close enough to the pins. For prototyping it is not very critical in my experience, a hang or fault once a day is not a dealbreaker. The purpose is to gain experience with the chip. But in that case a smaller breakout pcb with pins close to the chip would be better.

[westfw]

I have to say that I really like the trend toward "little" eval boards that plug into USB and sell for 'cheap.'  I think TI started it with the 430F2013 USB stick, and the Cypress boards that have been mentioned are an excellent example.  All pins available, usable but minimalistic programming interface, and easy to play with.  If NXP put their LPC1102 on a DIP18 adapter with some decoupling and a serial bootloader, and sold it for $5, I'd be pretty happy.  The price-point bar is pretty low, though.  The LPCXPresso boards are only about $20, and so are for-profit boards like the "Teensy 3."  $30 bucks for a mount-it-yourself bga adapter+chip isn't horrible if you really want to work with that chip, but it's not an enticing toy or impulse purchase, either.

I generally observe that if you want to catch engineers' imaginations and lure them to being interested in your products, you almost need to treat them like hobbyists.  The number of occasions where a company or engineer will actually "evaluate" a wide range of chips/vendors using their "eval board" type offerings is very small.  Usually you get a decision made based on someone's PRIOR or company legacy experiences (ranging from "I used an arduino in HS; let's use an AVR" to "our code base is 100k lines of code written for a 68000 that no one paid much attention to making portable.  We're for damned sure stuck with a big-endian 32bit word chip and some new 68xxx compatible like CPU32 sounds a lot better than something more different."

(now, the BIG eval boards come in handy prototyping your software while your HW design is percolating through its cycle.   The aforehinted 68k product had as much of its core as possible brought up on a 68331 eval board before the HW was done, and took about two days to get running on the actual HW when it came in.)

[kert]

The LPCXPresso boards are only about $20, and so are for-profit boards like the "Teensy 3."  $30 bucks for a mount-it-yourself bga adapter+chip isn't horrible if you really want to work with that chip, but it's not an enticing toy or impulse purchase, either.

Actually if someone tried to make a new Lilypad with micro connectors and a Cortex mcu,  sort of like Digistump Digispark or SquareWear 2.0 but even smaller, i'd be all for it. The ultimate micro proto board.

[dannyf]

I suspect that you can simply compile it as other F4 chips and load the code to the chip.

Worst comes to work, compile it as a generic m4 chip, use the f4 start-up code and the F4 stdperiph library.

BTW, my Keil supports it.

[chris0822]

I got one of those STM32F429 Discovery boards for free at a ST Seminar.  I would be careful ordering it since a decent number of them at the seminar had faulty LCD displays.  Seems like it was a manufacturing issue so it's possible they have been corrected by now.

We used a time limited version of Keil to compile code for it, but I will try CoIDE when I get a chance as that is what I've used in the past with other ST Discovery boards.  The debugger seems to work really well with built in ST Link programmer.

[kert]

The point being that it's going to be tough to implement proper decoupling on a breakout board -- there's no guarantee that the traces from the power pins are short or even stay close together (although you might get lucky here because there are 16 ways to place the chip, but then again, you also have separate digital and analog supply, so good luck).

Got the demos from protoboard, pic attached. They couldnt find a 0.4mm pitch freescale part, so had to substitute for qfn.
The NXP one looks super cool and tiny though. Possibly could solder this with a hotair rework station on my own board, but it would be a bit of crapshoot i guess.
Unfortunately it will have to wait several weeks before i can spend time attempting to power these up though.

[BloodyCactus]

I'm looking forward to the Propeller2!