nice, i thought the onboard (not on-chip) peripherals have different pins than the netmf uses (many pins can have different types of usage)

leds and button is kinda easy.

a-d is also, the serial ports are also in the netmf port (except one isnt working becouse of the micro usb port having the same pins)

all 3 spi modules work, one i2c (netmf limit)

and the 2 da converters are also implemented.

wich spi module and cs pin it uses? and whats about the interrupt pin? (the accelerometer)

also, what you think about the audio ic? that would be cool to have too, but from what ive read its i2s

i cant answer your question, but iam sure soldering wires will make way more noise/interference than just using 2 devices on one spi port.

i mean thats what it is made for.

also, there isnt any interferance, each spi device has a chipselect pin, wich activates it. when not activate it does nothing.

edit: also, when you think you can use 2 at the same time, thats not possible, you can read/write on one port, than the other, not at the same time.

that would require dma, wich iam fairly sure is not in netmf. netmf's threading is also no real threds, never run 2 at the same time.

my opinoin: yes, it lacks some documentation, especially for the netmf core and porting kit.

than: there is no nice ide for compiling the firmware itself, and gcc support is limited (at least right now for stm32f4's)

i found it easiert to write c/c++ code for my stm than messing around with the firmware.

the compiler secretlabs uses is very expensive, there are some ports of the firmware, but well, i personally havent found it easwy to use.

right now iam playing a lot with 8 bit avr's - they have a brilliant (and for me well known) ide - atmel studio is just visual studio with a different compiler. they are very fast - a 16mhz avr beats my 168mhz stm32f4 a lot - thats becouse netmf is interpreted. there are lots of tutorials and code.

tuts and samples is something i havent found a lot of the stm32's.

dont get me wrong - netduino/netmf is kinda nice, but my opinion is it will take some more years of work to make it really good/outstanding.

thats no exception.

thats just a message that a thread stopped, wich is perfectly fine (except its a thread you dont wanna be stopped).

i guess its from the webserver, since netmf doesent offer async there are threads used for the blocking methods.

as long everything works id not be worried.

yeah id test it.

iam not sure if that will work without firmware modifications.

edit: id also make sure before if the pins not used otherwise, and all is correct (capacitors, vbat not connected anywhere else...)

or it will burn

another solution would be just adding an external rtc ic. well, that adds some extra costs (wich doesent matter for low run stuff, only when you really gonna make a lot and maybe sell them)

the benefit would be: the newer and better ones have a temerature compensated crystal inside.

so they only need a battery connected. aviable in both spi and i2c

they would also add some battery backed sram as storage space (some bytes, not that much, but also not really bad)

cpu.pins. is only the hardware pin enumeration. means: pin 1 there is pin1 on the _uC_

you need to use Pins.yourpin (without cpu)

 (0->~4000)

doesent fit in a byte, so you have to send/read 2, and shift it right before/after.

like

byte[] ReadBuffer = new byte[2];

this._Device.ReadRegister(addr, ReadBuffer);

int value = shift the byte correct here;

return value;

great

sounds like the stuff  http://www.wvshare.com offers. (except they dont offer netmf, and netmf Drivers for their modules)

yes you could speed it up, but not in netmf. you would have to add that to the firmware, in c/c++.

netmf is interpreted and therefore not as fast as c/c++ code.

that with the pwm is a bad idea, it wount work, and even if, i wouldnt increase any speed.

calling the spi write/read often is just not very efficient, it has some overhead.

not even that doesent help you, but ill have to add: writing bigger junks is very fast compared to writing a lot of small ones.

you have to flash the uart firmware first.

found here (with some instructions)

http://forums.netdui...-v420-update-1/

it has 2 firmware, one for rs232 deployment, one for uart deployment.

1: yes (well i honestly dont 100% understand your question, but iam sure its yes)

2: yes

3: yes (edit: well i dont know if _you_ can, but its possible)

(notice you didnt ask how todo it, so all my answers are correct   )

My general practice is to have the constructor of the main class create an instance of it's self and run from there. Can I assume the examples run as static to keep it easy to understand

exactly this.  i find it ugly to make all static, and for me it looks odd and confusing.

about gc isnt much to say: it cares about it on its own. memory management: dont push more items in a queue than you remove and your fine

that question also gets asked every month

its what ziggurat said, its normal, so dont worry. an error looks different.

(windows apps show the same message when a thread ends)

http://blog.codeblac...with-RS232.aspx

maybe Verdris wanna lik the product.

ive never saw anything with minimal amp requipment, maybe it means the

supply will at least be able to output 1 amp.

my power supply has 7 amps max, all fine with netduino.

http://at.mouser.com...ualkey967-LS355

its very small also - but can be dangerous when not handled correctly.

btw: i only know that some step up regulators may need an consumer to not blow up, becouse of the voltage sensing.

(the lm from ebay work fine, even when nothing is connected)

its looks ok, what does tghe who am i register answer?

ill add my basic code for this module, its i2c, but maybe it helps

also try changing your spi speed, start with 1-2 mhz (idk the specs for the device now, but 10mhz sounds high)

maybe Manual reset Event may help you

http://msdn.microsof...resetevent.aspx

depends on the power consumption of the motor and the voltage you out into the netduino (more volts in = more heat for the regulator = less power output)

id suggest using external power supply for motors.

i see.

ok, i assume most Motor shields use an h-Bridge.

a shield would be the easierst ofc (maybe one that accepts external power supply)

or just build an h-Bridge yourself, wich requies a little more work.

Transistors i dont suggest, they have a voltage drop and are only capeable of some milliamps (like 200 the Standard ones).

mosfets would be way better, but still, and h-Bridge is the way to go.

//edit: removed, guess i understood that wrong.

read write with usb port: short: no, it can print debug data that you can read, but not write.

but you can use a ftdi cable or cp2102 adapter, that make a virtual serial port over usb - a great thing

client /ser: well, i do that not via ethernet, ill use the cp2102 uart solution for that. seems to be way more lightweight -

but the principle is the same. the gui runs in a windows app and reads/writes data to the serial port.

btw: i think you can hook up an enc28j60 the the normal n2 and make ethernet work, but iam not sure.

if that doesent work you can use mIP - a non secretlabs enc28j60 driver -

but with that you cant use the netmf socket classes.

do you power the ds1307 with 3.3 or 5v?

it needs 5v. or it wount work correctly.

(the 3.3v i2c levels are ok, but the supply has the be 5v)

- it works with 3.3v when you ground v-bat (and remove the battery before ofc...), but than theres no timekeeping when power is off ...

//edit: btw in the link from my signature is a ds1307 class, and the multi i2c i use also. it also shows how to use the ram from the ds1307

interesting. i just used a max232 for this, but your version is smaller in size wich is nice!