the raspberry PI at the core of a future machine?

[target]

at $25 for a complete linux computer the size of a credit card with everything on board ,seems tempting if you can get your hands on one .
so you start with a clean slate as current technology is up a cul de sac ,you then you add 3D imaging or GPR .
add your thoughts .it seems to good to ignore .
they will become more common as production gets into gear .

http://www.raspberrypi.org/faqs

[target]

http://singularityhub.com/2012/03/03...es-with-sales/

more info ,what a bargain !

[Qiaozhi]

If you pre-order the Raspberry PI, they will ship units in the order they're received, and there is a limit of one per household. Farnell are even giving away a free tee-shirt with each order while stocks last.

The model B seems to be the best option, if you want to be an early adopter. It might actually be better to wait until the summer, when the bugs have ironed out and versions with a case will be available. I expect there will also be some upgrades to the board as well.

[simonbaker]

A/D converter?

Any A/D converter on board?

-SB

[Davor]

Impressive. I generally detest micros in DIY electronics, but this one is interesting.

[apaulsalerno]

It has audio In/Out in some form at least, I imagine there must be some capabilities of the audio A/D, how many bit or the resolution I dont know.

[moodz]

embedded linux systems

....he he ... I was building embedded linux systems well over 10 years ago .... see

http://lwn.net/1998/0430/miniHOWTO.html

Has anyone heard the old story of the "stone soup" ...

If you have you will know what I am talking about so I wont go further here.

The rasberry PI is a good board ... however be aware it is severely lacking in peripherals that make Metal detector hardware easier to work with ....

like PWMs ( or OC output compares ) ... these are absolutely necessary for pulse generation ( IC or input compares ) are necessary for pulse measurement.
QEI .... quadrature encoders would be nice ... the Strawberry has none of these.
Standardised IO layout ....not on this one ....

You could do all the bit banging and timing in machine code but this is primitive and hard to maintain.

Its a good pick Davor but more for Linux hackers .... you would have to add some more hardware to get good pulse generation etc.

I would suggest the UNO32 for example and the protosheild boards for development .... it is 32 bit, runs at 80Mhz, on board programming, has stacks of IO and a high level SDK and best of all its available, open source hardware and software, standard arduino interface and is around the $25 mark. There are stacks of ther modules that can be stacked ( no pun intended ) on top.
moodz

[Qiaozhi]

Stone Soup

http://en.wikipedia.org/wiki/Stone_soup

[mario]

....he he ... I was building embedded linux systems well over 10 years ago .... see

http://lwn.net/1998/0430/miniHOWTO.html

Has anyone heard the old story of the "stone soup" ...

If you have you will know what I am talking about so I wont go further here.

The rasberry PI is a good board ... however be aware it is severely lacking in peripherals that make Metal detector hardware easier to work with ....

like PWMs ( or OC output compares ) ... these are absolutely necessary for pulse generation ( IC or input compares ) are necessary for pulse measurement.
QEI .... quadrature encoders would be nice ... the Strawberry has none of these.
Standardised IO layout ....not on this one ....

You could do all the bit banging and timing in machine code but this is primitive and hard to maintain.

Its a good pick Davor but more for Linux hackers .... you would have to add some more hardware to get good pulse generation etc.

I would suggest the UNO32 for example and the protosheild boards for development .... it is 32 bit, runs at 80Mhz, on board programming, has stacks of IO and a high level SDK and best of all its available, open source hardware and software, standard arduino interface and is around the $25 mark. There are stacks of ther modules that can be stacked ( no pun intended ) on top.
moodz
[ATTACH]18751[/ATTACH]
[ATTACH]18752[/ATTACH]

I agree.

This Raspberry Pi board is awesome for video/audio projects, but could never replace a full feature microprocessor. It is essentially lacking basic features that we have taken for granted like ADC and timers.

Although it does have a timer, it only has one, so your options of implementing multiple feature support for something useful (like quadrature decoding) is extremely limited to just one.

Also, if you are interested in getting started with microprocessors, and using the Arduino platform, I would never recommend the Uno32.

The Arduino platform is a closed system, so the "sketches" written for the true "Uno " will always have issues running on other platforms.

The bit banged Arduino IDE (development software) used in the UNO32 will always have compatibility issues with Arduino written programs - even more so now that the final Arduino IDE 1 has been released.

My advice, stick to the original Uno if you want Arduino. Otherwise, use the development tools that were written for your platform. Microchip development software is good enough to stand on its own.

[Davor]

So no micro is a better choice after all.
I got so pissed off by altera completely changing their specifications over a decade ago that I simply stroke off the micros from the menu altogether. The most of the hobby projects, and much more, are viable through garden variety components and no micros. Maybe it is a perfect moment to re-evaluate your goals when you think of too much programmable solutions.

[mario]

So no micro is a better choice after all.
I got so pissed off by altera completely changing their specifications over a decade ago that I simply stroke off the micros from the menu altogether. The most of the hobby projects, and much more, are viable through garden variety components and no micros. Maybe it is a perfect moment to re-evaluate your goals when you think of too much programmable solutions.

No Davor, don't let me be the one to discourage you. Even though the above mentioned platforms are not a one stop solution for every project imaginable, you should not be discouraged from learning more about microprocessors.

So many great projects have been created, and as newer technologies become available (like the Raspberry Pi), many more remain to be created.

moodz brought up the Arduino platform for a reason.

Most individuals are put off from learning something they are unsure about and are unwilling to tackle something they consider challenging. With this in mind, Arduino was realized, a highly customizable platform that makes the whole experience less stressful.

Before Arduino, people programmed their micros via an obscure and confusing syntax using assembly language, the lowest level of programming imaginable. With assembly, you are programming directly to the metal and is very difficult.

My issue with moodz's advice is that he has obviously not used Arduino, nor the platform he's recommending (uno32), which will ruin the whole microprocessor experience if you're a beginner and take his advice seriously. The two platforms are not equal. The support and communities (and ecosystems) built around either of the two are not the same.

As a beginner, stick to the original Arduino Uno platform. You don't want to be frustrated by basic compatibility issues and have things not work as they're suppose to work.

Once you have become more familiar with using Arduino, then you can move onto the Uno32 (a newer board) or the Raspberry Pi. I don't want to imply that the Uno32 is more difficult to code for, because it uses standardized languages easier than assembly. However, I do want emphasize that the two platforms are not the same.

Although I don't use the Arduino platform, I have used "Arduino compatibles", and they're not the same. As a beginner, I would not want to experience these differences first hand.

[moodz]

No Davor, don't let me be the one to discourage you. Even though the above mentioned platforms are not a one stop solution for every project imaginable, you should not be discouraged from learning more about microprocessors.

So many great projects have been created, and as newer technologies become available (like the Raspberry Pi), many more remain to be created.

moodz brought up the Arduino platform for a reason.

Most individuals are put off from learning something they are unsure about and are unwilling to tackle something they consider challenging. With this in mind, Arduino was realized, a highly customizable platform that makes the whole experience less stressful.

Before Arduino, people programmed their micros via an obscure and confusing syntax using assembly language, the lowest level of programming imaginable. With assembly, you are programming directly to the metal and is very difficult.

My issue with moodz's advice is that he has obviously not used Arduino, nor the platform he's recommending (uno32), which will ruin the whole microprocessor experience if you're a beginner and take his advice seriously. The two platforms are not equal. The support and communities (and ecosystems) built around either of the two are not the same.

As a beginner, stick to the original Arduino Uno platform. You don't want to be frustrated by basic compatibility issues and have things not work as they're suppose to work.

Once you have become more familiar with using Arduino, then you can move onto the Uno32 (a newer board) or the Raspberry Pi. I don't want to imply that the Uno32 is more difficult to code for, because it uses standardized languages easier than assembly. However, I do want emphasize that the two platforms are not the same.

Although I don't use the Arduino platform, I have used "Arduino compatibles", and they're not the same. As a beginner, I would not want to experience these differences first hand.

LOL .... oh so thats what that stuff littering my workbench is ... thanks Mario ..I was going to throw it out.

Moodzarduino

[moodz]

Seriously though ... if you read my orginal post carefully ... I never said to use the arduino SDK .... I meant use the Microchip IDE MPLAB .... Get yourself a good book on C .... look at the sample code on the microchip site ( theres heaps ).

The arduino SDK uses psuedo C or "sketches" .. it would be an OK platform if you could see the C code generated before it compiles. The trouble with sketches is they hide alot of the meaningful info with regard to real time programming which you will not notice as a biginner but will regret when you become more experienced.

The microchip pic32 ...absolutely creams the Atmega parts ( about 40 times faster ) used on the original arduino. When you are trying to get some 32 bit numbers multiplied before the next pulse arrives on your PI circuit ... the Atmega will not cut the mustard.

Mario, your advice is OK if you never want to get past blinking LEDs, reading some slow speed ADC or start stopping a motor but if you want to perform a realtime FFT in 24 bit math you will need the Microchip part ( and guess what ... there are prebuilt libraries to do all that DSP stuff ... you dont have to code it )

microchipmoodz

... (do you like the subliminal advertising).

[moodz]

So no micro is a better choice after all.
I got so pissed off by altera completely changing their specifications over a decade ago that I simply stroke off the micros from the menu altogether. The most of the hobby projects, and much more, are viable through garden variety components and no micros. Maybe it is a perfect moment to re-evaluate your goals when you think of too much programmable solutions.

Despite all the digital hype ..... I reckon you could build very good stuff using analogue techniques only.
Analogue computing used to be very big in control applications and sometimes could do things that digital cannot. I think the Russians were very good at it in some applications.

moodz

[mario]

moodz, you're killing me.

If you depend on using true Arduino sketches, your board will continue to "litter your workbench." Indeed, even if you are able to compile them, your board would not be able to drive the "blinking LEDs" programs you make so fun off (here's a hint: the uno32 can't drive the current -ironic that you should even bring that up!).

I will not get into a platform war with you considering that the Arduino platform is an established platform dating back several years now with more users and support than anything you can counter.

It's not fair to compare the newer chips found on the Uno32 with the ones used in the Arduino. Five years ago Microchip was making crap, and if you really want to be fair, how about finding a board from the same time period to challenge the hardware.

There will be a successor to the original Uno, and it's looking like it's going to bring an end to your comparisons considering that the likely hardware used in the successor will likely make anything from Microchip's DSP line like yesterday's hardware. Of course support from the Arduino's install base will follow, And this continue support will likely further enrage people like yourself (Microchip's aggressive PR machine is cut throat).

Anyway, according to your misconceptions about Arduino hardware, I would not be able to do what I would be doing in regards to DSP with the processor I currently use in my online projects, which incidentally is used in Arduino platform. Yes I will be using FFT, and no, the processor is not 24bit. However, it's certainly more than enough to tackle the difficult, processor-intensive calculations that will be required to do what needs to be done.

Grow up moodz, the Arduino platform is the most established platform for learning microprocessors. It's not meant to be cutting edge, the platform was conceived for learning. How about comparing the chip found on the Uno32 with Atmel's Cortex 9M line, the one with the integrated LCD controller, which also has a dedicated hardware HD video decoder (http://www.atmel.com/devices/SAM9M10.aspx). At 400Mhz, the Atmel SAM9M10 ...absolutely creams the Microchip DSP parts...