Forums

Your Favorite ARM?

Started by Martin Riddle December 26, 2011
"Werner" <wdahn@netfront.net> wrote in message 
news:jdd45r$2ee2$1@adenine.netfront.net...
> On 27/12/11 11:36, Martin Riddle wrote: >> "John Larkin"<jjlarkin@highNOTlandTHIStechnologyPART.com> wrote in >> message news:5fcif792ljl9morgkae9kq27stpenoc5dj@4ax.com... >>> On Mon, 26 Dec 2011 20:19:27 -0500, "Martin Riddle" >>> <martin_rid@verizon.net> wrote: >>> >>>> Which Mfg has the best options for their ARM processors? >>>> I've never worked with a ARM, and I was going to play with a Atmel >>>> AVR32, but the ARM seems to have more features for a buck or two >>>> more. >>>> Preferable package is LQFP. And not so pricey tools. >>>> 12 bit DAC and ADC a plus. >>>> >>>> Cheers >>>> >>>> >>>> >>> >>> Depends on how much horespower you need. We have one board with a >>> dozen LPC1754s on board, and they work fine. It has a 12-bit ADC and >>> a >>> 10-bit DAC, on-chip flash and ram, 100 MHz. We pay about $3.50 each >>> for them, programmed by Arrow with our code. >>> >>> The 12-bit ADC is accurate to 9 bits maybe. We wound up doing a >>> curve >>> fit to every one at cal time, to get it to 0.1%. uP ADCs tend to be >>> like that. >>> >>> We also use a bunch of LPC3250s, much more serious parts: BGA, >>> vector >>> floating point, DRAM controller, 260 MHz or so, about $7.50. >>> >>> The NXP LPCs are OK, but the manuals are sort of pitiful. >>> >>> John >>> >>> >> >> Ok, I can live with that. I never seen really good performance from >> on >> board ADC's anyway, always some errata associated with them. DAC's >> were >> always external. >> The LPC17xx looks like it would fit the bill. Even has a Ethernet >> which >> I will need. >> This must be ARMv7 = Cortex-M3... >> >> Ok, I'm off to find a demo board. >> >> Thanks! >> >> Cheers > Before you run off consider this one: > > http://search.digikey.com/hk/en/products/STM32F4DISCOVERY/497-11455-ND/2711743 > > The eval board costs roughly the same as a chip, about 20 USD. > Cortex M4 with floating point (and potentially ethernet). > > Really sexy! >
Yes, That has 2 x 12bit DAC's in it. For $20 can't go wrong. I can add the PHY for the ethernet. STM32F405 is about $11 in qty, not exactly cheap, but if the DAC and ADC's work well then it pays off. Thanks for pointing this one out. Cheers
On Mon, 26 Dec 2011 19:35:54 -0600, jw wrote:

> On Mon, 26 Dec 2011 20:19:27 -0500, "Martin Riddle" > <martin_rid@verizon.net> wrote: > >>Which Mfg has the best options for their ARM processors? I've never >>worked with a ARM, and I was going to play with a Atmel AVR32, but the >>ARM seems to have more features for a buck or two more. Preferable >>package is LQFP. And not so pricey tools. 12 bit DAC and ADC a plus. >> >>Cheers >> >> > Waht is an ARM? (Dont say the thing attached to my shoulder :) ).
You don't KNOW?!?!?!?! Or are you trolling? At any rate, its a processor made to a design sold by these guys: http:// www.arm.com/. Think of it as occupying the equivalent ecological niche as the 8051, only with 32 bits and a much better instruction set. -- My liberal friends think I'm a conservative kook. My conservative friends think I'm a liberal kook. Why am I not happy that they have found common ground? Tim Wescott, Communications, Control, Circuits & Software http://www.wescottdesign.com
On Mon, 26 Dec 2011 20:19:27 -0500, Martin Riddle wrote:

> Which Mfg has the best options for their ARM processors? I've never > worked with a ARM, and I was going to play with a Atmel AVR32, but the > ARM seems to have more features for a buck or two more. Preferable > package is LQFP. And not so pricey tools. 12 bit DAC and ADC a plus. > > Cheers
As for tools -- as long as the processor JTAG port is compatible, tools are pretty much interchangeable. I prefer the Gnu tools, but that's for idealogical reasons as much as anything. -- My liberal friends think I'm a conservative kook. My conservative friends think I'm a liberal kook. Why am I not happy that they have found common ground? Tim Wescott, Communications, Control, Circuits & Software http://www.wescottdesign.com
On Tue, 27 Dec 2011 08:57:35 -0800, John Larkin
<jjlarkin@highNOTlandTHIStechnologyPART.com> wrote:

>On 27 Dec 2011 02:07:00 -0600, boB wrote: > >>On Mon, 26 Dec 2011 18:57:49 -0800, John Larkin >><jjlarkin@highNOTlandTHIStechnologyPART.com> wrote: >> >>>On Mon, 26 Dec 2011 20:19:27 -0500, "Martin Riddle" >>><martin_rid@verizon.net> wrote: >>> >>>>Which Mfg has the best options for their ARM processors? >>>>I've never worked with a ARM, and I was going to play with a Atmel >>>>AVR32, but the ARM seems to have more features for a buck or two more. >>>>Preferable package is LQFP. And not so pricey tools. >>>>12 bit DAC and ADC a plus. >>>> >>>>Cheers >>>> >>>> >>>> >>> >>>Depends on how much horespower you need. We have one board with a >>>dozen LPC1754s on board, and they work fine. It has a 12-bit ADC and a >>>10-bit DAC, on-chip flash and ram, 100 MHz. We pay about $3.50 each >>>for them, programmed by Arrow with our code. >>> >>>The 12-bit ADC is accurate to 9 bits maybe. We wound up doing a curve >>>fit to every one at cal time, to get it to 0.1%. uP ADCs tend to be >>>like that. >>> >>>We also use a bunch of LPC3250s, much more serious parts: BGA, vector >>>floating point, DRAM controller, 260 MHz or so, about $7.50. >>> >>>The NXP LPCs are OK, but the manuals are sort of pitiful. >>> >>>John >>> >> >>Interesting ! >> >>We use NXP LPC176x and LPC2366 parts and I like them. >>I was going to do some kind of A/D calibration tables to help the 10 >>bit converters. >> >>Do you do a 1:1 (4096 point) table mapping or do you just do a >>limited amount of points and interpolate or something else ? > >On the board I mentioned, we did a 64-point lookup table with >interpolation. The ADC transfer function had some jumps, so that was >good. Sometimes we do polynomials for stuff like this, if the curves >are softer. > >> >>How long does it take to calibrate one part in production > >I recall times like 20 minutes to test/cal a board, all automated. >This is it: > >http://www.highlandtechnology.com/DSS/V220DS.shtml > >The uPs are each electrically isolated and each one runs a local PID >loop to do each 4-20 mA channel. Sort of a nuisance to design. > > > and is it >>repeatable as far as drift etc out in the field ?? > >I sure hope so! Seems OK so far. But I guess asking a uP ADC to do >0.1% is pushing things. > > > I would guess >>that maybe it is but how well does it really "stick" ?? How different >>is one part's A/D from another ? Are they all over the place ? > >Enough that we can't spec 0.1% without a lookup table for each one. >The cal table also picks up all the resistor tolerances, opamp >offsets, whatever, so it's all lumped together. > >In retrospect, we could have used a real ADC, ADS7866 maybe, and saved >a lot of hassle. > >John
Very COoL to hear it works OK. We certainly don't need 0.1% but anything that makes it like, 1% or less would be nice. So, how good do you find the linearity without doing anything at all ?? Seems that I did this with the 10 bit converter in the LPC2366 once but can't remember for sure. I hope the A/D mux doesn't cause any problems. None that I know of though. I just know that the whole A/D system, including resistors and A/D Vcc etc needs to be taking into account and a table should certainly do it. Thanks John ! boB K7IQ
This is an OpenPGP/MIME signed message (RFC 2440 and 3156)
--------------enigE5AB85C6CB8BA0EBFC9AA4EA
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

Tim Wescott wrote:

> On Mon, 26 Dec 2011 20:19:27 -0500, Martin Riddle wrote: >=20 >> Which Mfg has the best options for their ARM processors? I've >> never worked with a ARM, and I was going to play with a Atmel >> AVR32, but the ARM seems to have more features for a buck or two >> more. Preferable package is LQFP. And not so pricey tools. 12 bit >> DAC and ADC a plus. >>=20 >> Cheers >=20 > As for tools -- as long as the processor JTAG port is compatible, > tools are pretty much interchangeable. >=20 > I prefer the Gnu tools, but that's for idealogical reasons as much as > anything.
86 general input/output lines ought to be enough for anyone... =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D The high-performance, low-power Atmel 8-bit AVR RISC-based microcontroller combines 256KB ISP flash memory, 8KB SRAM, 4KB EEPROM, 86 general purpose I/O lines, 32 general purpose working registerhttp://atmel.com/dyn/products/product_card.asp?part_id=3D3632s, real time counter, six flexible timer/counters with compare modes, PWM, 4 USARTs, byte oriented 2-wire serial interface, 16-channel 10-bit A/D converter, and a JTAG interface for on-chip debugging. The device achieves a throughput of 16 MIPS at 16 MHz and operates between 4.5-5.5 volts. http://atmel.com/dyn/products/product_card.asp?part_id=3D3632 =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D http://www.arduino.cc/en/Main/ArduinoBoardMega2560 mike --------------enigE5AB85C6CB8BA0EBFC9AA4EA Content-Type: application/pgp-signature; name="signature.asc" Content-Description: OpenPGP digital signature Content-Disposition: attachment; filename="signature.asc" -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.18 (GNU/Linux) iQEcBAEBAgAGBQJO+odNAAoJECGtZwCbtgwXWg4IAJ2kjhHGF/E9gaQc2ve0I2oX Oy2eoKyydrbD0gICjoxVM7OIGNnSFPqRi1GwNpuiHI/hNTPIN09qBayhoL1w1UWU Qagykwuqt4a7P3+t8sX9/AN3VRplF3iq2IeUTuowJS43u7zYgHXW+Hvz90QjlOtQ NsB0bE0XwKCMC1Wu+CeVNNRVh60ZYR9bJf7HdPLGTzaaA2weNmJw8jNZtMC/CfCI BjSmuwRSaP30sRNGzD4I0mzUPgQrvX64poJERmVP5nCJkor35RAs36tVBfJdQkN3 nEXUQ78SwnViULfNkPllW2/6ylNEgnpqnhBZ77aef9hkm5sgCzLh653h88WwbwM= =3rs3 -----END PGP SIGNATURE----- --------------enigE5AB85C6CB8BA0EBFC9AA4EA--
On 12/27/2011 7:21 PM, boB wrote:
> On Tue, 27 Dec 2011 08:57:35 -0800, John Larkin > <jjlarkin@highNOTlandTHIStechnologyPART.com> wrote: > >> On 27 Dec 2011 02:07:00 -0600, boB wrote: >> >>> On Mon, 26 Dec 2011 18:57:49 -0800, John Larkin >>> <jjlarkin@highNOTlandTHIStechnologyPART.com> wrote: >>> >>>> On Mon, 26 Dec 2011 20:19:27 -0500, "Martin Riddle" >>>> <martin_rid@verizon.net> wrote: >>>> >>>>> Which Mfg has the best options for their ARM processors? >>>>> I've never worked with a ARM, and I was going to play with a Atmel >>>>> AVR32, but the ARM seems to have more features for a buck or two more. >>>>> Preferable package is LQFP. And not so pricey tools. >>>>> 12 bit DAC and ADC a plus. >>>>> >>>>> Cheers >>>>> >>>>> >>>>> >>>> >>>> Depends on how much horespower you need. We have one board with a >>>> dozen LPC1754s on board, and they work fine. It has a 12-bit ADC and a >>>> 10-bit DAC, on-chip flash and ram, 100 MHz. We pay about $3.50 each >>>> for them, programmed by Arrow with our code. >>>> >>>> The 12-bit ADC is accurate to 9 bits maybe. We wound up doing a curve >>>> fit to every one at cal time, to get it to 0.1%. uP ADCs tend to be >>>> like that. >>>> >>>> We also use a bunch of LPC3250s, much more serious parts: BGA, vector >>>> floating point, DRAM controller, 260 MHz or so, about $7.50. >>>> >>>> The NXP LPCs are OK, but the manuals are sort of pitiful. >>>> >>>> John >>>> >>> >>> Interesting ! >>> >>> We use NXP LPC176x and LPC2366 parts and I like them. >>> I was going to do some kind of A/D calibration tables to help the 10 >>> bit converters. >>> >>> Do you do a 1:1 (4096 point) table mapping or do you just do a >>> limited amount of points and interpolate or something else ? >> >> On the board I mentioned, we did a 64-point lookup table with >> interpolation. The ADC transfer function had some jumps, so that was >> good. Sometimes we do polynomials for stuff like this, if the curves >> are softer. >> >>> >>> How long does it take to calibrate one part in production >> >> I recall times like 20 minutes to test/cal a board, all automated. >> This is it: >> >> http://www.highlandtechnology.com/DSS/V220DS.shtml >> >> The uPs are each electrically isolated and each one runs a local PID >> loop to do each 4-20 mA channel. Sort of a nuisance to design. >> >> >> and is it >>> repeatable as far as drift etc out in the field ?? >> >> I sure hope so! Seems OK so far. But I guess asking a uP ADC to do >> 0.1% is pushing things. >> >> >> I would guess >>> that maybe it is but how well does it really "stick" ?? How different >>> is one part's A/D from another ? Are they all over the place ? >> >> Enough that we can't spec 0.1% without a lookup table for each one. >> The cal table also picks up all the resistor tolerances, opamp >> offsets, whatever, so it's all lumped together. >> >> In retrospect, we could have used a real ADC, ADS7866 maybe, and saved >> a lot of hassle. >> >> John > > Very COoL to hear it works OK. We certainly don't need 0.1% but > anything that makes it like, 1% or less would be nice. > > So, how good do you find the linearity without doing anything at all > ?? Seems that I did this with the 10 bit converter in the LPC2366 > once but can't remember for sure.
I don't know how John does it, but if you have a stable clock, you can do this sort of thing pretty well with an RC time constant, suitably buffered. You start it off at some known multiple of Vref and watch it decay--nine or ten time constants will get you to zero within the resolution of your ADC. You do need a decent capacitor, or else its voltage coefficient will cause errors. Cheers Phil Hobbs -- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 845-480-2058 hobbs at electrooptical dot net http://electrooptical.net
On Wed, 28 Dec 2011 03:03:16 -0500, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

>On 12/27/2011 7:21 PM, boB wrote: >> On Tue, 27 Dec 2011 08:57:35 -0800, John Larkin >> <jjlarkin@highNOTlandTHIStechnologyPART.com> wrote: >> >>> On 27 Dec 2011 02:07:00 -0600, boB wrote: >>> >>>> On Mon, 26 Dec 2011 18:57:49 -0800, John Larkin >>>> <jjlarkin@highNOTlandTHIStechnologyPART.com> wrote: >>>> >>>>> On Mon, 26 Dec 2011 20:19:27 -0500, "Martin Riddle" >>>>> <martin_rid@verizon.net> wrote: >>>>> >>>>>> Which Mfg has the best options for their ARM processors? >>>>>> I've never worked with a ARM, and I was going to play with a Atmel >>>>>> AVR32, but the ARM seems to have more features for a buck or two more. >>>>>> Preferable package is LQFP. And not so pricey tools. >>>>>> 12 bit DAC and ADC a plus. >>>>>> >>>>>> Cheers >>>>>> >>>>>> >>>>>> >>>>> >>>>> Depends on how much horespower you need. We have one board with a >>>>> dozen LPC1754s on board, and they work fine. It has a 12-bit ADC and a >>>>> 10-bit DAC, on-chip flash and ram, 100 MHz. We pay about $3.50 each >>>>> for them, programmed by Arrow with our code. >>>>> >>>>> The 12-bit ADC is accurate to 9 bits maybe. We wound up doing a curve >>>>> fit to every one at cal time, to get it to 0.1%. uP ADCs tend to be >>>>> like that. >>>>> >>>>> We also use a bunch of LPC3250s, much more serious parts: BGA, vector >>>>> floating point, DRAM controller, 260 MHz or so, about $7.50. >>>>> >>>>> The NXP LPCs are OK, but the manuals are sort of pitiful. >>>>> >>>>> John >>>>> >>>> >>>> Interesting ! >>>> >>>> We use NXP LPC176x and LPC2366 parts and I like them. >>>> I was going to do some kind of A/D calibration tables to help the 10 >>>> bit converters. >>>> >>>> Do you do a 1:1 (4096 point) table mapping or do you just do a >>>> limited amount of points and interpolate or something else ? >>> >>> On the board I mentioned, we did a 64-point lookup table with >>> interpolation. The ADC transfer function had some jumps, so that was >>> good. Sometimes we do polynomials for stuff like this, if the curves >>> are softer. >>> >>>> >>>> How long does it take to calibrate one part in production >>> >>> I recall times like 20 minutes to test/cal a board, all automated. >>> This is it: >>> >>> http://www.highlandtechnology.com/DSS/V220DS.shtml >>> >>> The uPs are each electrically isolated and each one runs a local PID >>> loop to do each 4-20 mA channel. Sort of a nuisance to design. >>> >>> >>> and is it >>>> repeatable as far as drift etc out in the field ?? >>> >>> I sure hope so! Seems OK so far. But I guess asking a uP ADC to do >>> 0.1% is pushing things. >>> >>> >>> I would guess >>>> that maybe it is but how well does it really "stick" ?? How different >>>> is one part's A/D from another ? Are they all over the place ? >>> >>> Enough that we can't spec 0.1% without a lookup table for each one. >>> The cal table also picks up all the resistor tolerances, opamp >>> offsets, whatever, so it's all lumped together. >>> >>> In retrospect, we could have used a real ADC, ADS7866 maybe, and saved >>> a lot of hassle. >>> >>> John >> >> Very COoL to hear it works OK. We certainly don't need 0.1% but >> anything that makes it like, 1% or less would be nice. >> >> So, how good do you find the linearity without doing anything at all >> ?? Seems that I did this with the 10 bit converter in the LPC2366 >> once but can't remember for sure. > >I don't know how John does it, but if you have a stable clock, you can >do this sort of thing pretty well with an RC time constant, suitably >buffered. You start it off at some known multiple of Vref and watch it >decay--nine or ten time constants will get you to zero within the >resolution of your ADC. You do need a decent capacitor, or else its >voltage coefficient will cause errors. > >Cheers > >Phil Hobbs
Hmmm... That sounds like a neat idea. And have the micro calculate its own 1-e^-t/rc while it's calibrating itself. Not sure of the best way of switching the cal circuit in and out though. I'm out of A/D inputs myself. Maybe a decent cap could be plugged into a 2 pin socket of some sort so each unit doesn't have to have the possibly expensive cap stuffed in each board, unless it needs to be calibrated in the field. I don't need 0.1% like John does. It's gotta take care of its own V-regulator (reference) too so has to be in the target PCB. boB K7IQ
On Dec 26, 7:22=A0pm, Spehro Pefhany <speffS...@interlogDOTyou.knowwhat>
wrote:
> On Mon, 26 Dec 2011 18:57:49 -0800, the renowned John Larkin > > > > > > <jjlar...@highNOTlandTHIStechnologyPART.com> wrote: > >On Mon, 26 Dec 2011 20:19:27 -0500, "Martin Riddle" > ><martin_...@verizon.net> wrote: > > >>Which Mfg has the best options for their ARM processors? > >>I've never worked with a ARM, and I was going to play with a Atmel > >>AVR32, but the ARM seems to have more features for a buck or two more. > >>Preferable package is LQFP. =A0And not so pricey tools. > >>12 bit DAC and ADC a plus. > > >>Cheers > > >Depends on how much horespower you need. We have one board with a > >dozen LPC1754s on board, and they work fine. It has a 12-bit ADC and a > >10-bit DAC, on-chip flash and ram, 100 MHz. We pay about $3.50 each > >for them, programmed by Arrow with our code. > > >The 12-bit ADC is accurate to 9 bits maybe. We wound up doing a curve > >fit to every one at cal time, to get it to 0.1%. uP ADCs tend to be > >like that. > > >We also use a bunch of LPC3250s, much more serious parts: BGA, vector > >floating point, DRAM controller, 260 MHz or so, about $7.50. > > >The NXP LPCs are OK, but the manuals are sort of pitiful. > > >John > > The erstwhile Luminary, now a part of Texas Instruments, Stellaris > Cortex M3 parts are pretty nice too, esp. for their built-in ethernet > PHY. > > If you are serious about on-chip analog peripherals, you can look at > relevant AD parts which contain pretty decent ADC/DAC converters and > an ARM7TDMI (32 and 16 bit modes) as sort of a peripheral. > > Best regards, > Spehro Pefhany > -- > "it's the network..." =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0=
"The Journey is the reward"
> sp...@interlog.com =A0 =A0 =A0 =A0 =A0 =A0 Info for manufacturers:http://=
www.trexon.com
> Embedded software/hardware/analog =A0Info for designers: =A0http://www.sp=
eff.com- Hide quoted text -
> > - Show quoted text -
Did you actually use the ethernet code that comes with the Stellaris development board? We had a consultant renege on a contract because she claimed that the Luminary/TI libraries were incomplete and unuseable. I didn't have time to validate her claim - which is why we were using a consultant - but It sounded a little suspicious.
On Wed, 28 Dec 2011 08:06:11 -0800 (PST), the renowned Bob
<stephensdigital@gmail.com> wrote:

> >Did you actually use the ethernet code that comes with the Stellaris >development board? We had a consultant renege on a contract because >she claimed that the Luminary/TI libraries were incomplete and >unuseable. I didn't have time to validate her claim - which is why we >were using a consultant - but It sounded a little suspicious.
The dev boards came with a choice of one of three development systems- we bought all three and then purchased full versions of the one we liked the best for the price. IIRC, they're all based on BSD or something like that and I don't recall any major complaints about the driver- and two or three different people have created working devices with Ethernet as the main communications. I can ask them for more details if it's important to you. Sadly, I've not had a chance to play with it much myself. Best regards, Spehro Pefhany -- "it's the network..." "The Journey is the reward" speff@interlog.com Info for manufacturers: http://www.trexon.com Embedded software/hardware/analog Info for designers: http://www.speff.com
On 28 Dec., 04:04, m II <C...@in.the.hat> wrote:
> Tim Wescott wrote: > > On Mon, 26 Dec 2011 20:19:27 -0500, Martin Riddle wrote: > > >> Which Mfg has the best options for their ARM processors? I've > >> never worked with a ARM, and I was going to play with a Atmel > >> AVR32, but the ARM seems to have more features for a buck or two > >> more. Preferable package is LQFP. =A0And not so pricey tools. 12 bit > >> DAC and ADC a plus. > > >> Cheers > > > As for tools -- as long as the processor JTAG port is compatible, > > tools are pretty much interchangeable. > > > I prefer the Gnu tools, but that's for idealogical reasons as much as > > =A0anything. > > 86 general input/output lines ought to be enough for anyone... > > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
> The high-performance, low-power Atmel 8-bit AVR RISC-based > microcontroller combines 256KB ISP flash memory, 8KB SRAM, 4KB EEPROM, > 86 general purpose I/O lines, 32 general purpose working > registerhttp://atmel.com/dyn/products/product_card.asp?part_id=3D3632s, > real time counter, six flexible timer/counters with compare modes, PWM, > 4 USARTs, byte oriented 2-wire serial interface, 16-channel 10-bit A/D > converter, and a JTAG interface for on-chip debugging. The device > achieves a throughput of 16 MIPS at 16 MHz and operates between 4.5-5.5 > volts. > > http://atmel.com/dyn/products/product_card.asp?part_id=3D3632 > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
> > http://www.arduino.cc/en/Main/ArduinoBoardMega2560 >
for the same price you can get something like a stm32f407 - 168MHz cortex4, 1Mbyte flash, 192kbyte ram, 16ch. 12 bit adc, 4xUART, usb, ethernet, 2xCAN, etc. and theres plenty of cortex mcus that are cheaper with a bit less memory that are cheaper unless you absolutly need 5V io, why choose the slow 8bitter? -Lasse